Preferred Citation: Ooms, Herman. Tokugawa Village Practice: Class, Status, Power, Law.

Berkeley: University of California Press, c1996 1996.

http://ark.cdlib.org/ark:/13030/ft0000034x/

Tokugawa Village Practice

Class, Status, Power, Law

Herman Ooms

UNIVERSITY OF CALIFORNIA PRESS

Berkeley · Los Angeles · Oxford

© 1996 The Regents of the University of California

In memoriam Maria De Buyser, boerendochter

Preferred Citation: Ooms, Herman. Tokugawa Village Practice: Class, Status, Power, Law. Berkeley: University of California Press, c1996 1996.
http://ark.cdlib.org/ark:/13030/ft0000034x/

In memoriam Maria De Buyser, boerendochter

― xv ―

Acknowledgments
A number of people have assisted me in giving shape to this book, personally through comments, unknowingly through
publications, or anonymously through institutions. I wish here to gratefully acknowledge their contributions.

My greatest intellectual debt is to Pierre Bourdieu for helping me to think my way through a number of important subjects
taken up here. His writings, of great help in orienting my interpretations in Tokugawa Ideology , provided me with fresh
angles of analytical vision in the present work as well. A second scholar to whom I owe my thanks, first for his writings and
then for his personal assistance, is the local historian Ozaki Yukiya from the city of Ueda in Nagano prefecture. The heavy
concentration in this book of material from Kita-Saku district in Nagano is due in part to his publications. I am grateful for his
generous assistance in gathering data in Mochizuki during a brief visit in the summer of 1991 and his readiness to reduce my
ignorance by answering numerous queries.

Once the manuscript had taken form, Tom Keirstead and Anne Walthall generously offered their comments at a critical
moment when, simply happy that it was all on paper, I had lost all perspective on the project. Also of great help were those
who read and commented on particular chapters: my wife Emily Groszos Ooms, Mark Ramseyer, and graduate students at
UCLA. I would like to mention especially Christine Schoppe, who was always putting things together in unexpected ways, and
Jason Creigh, a close reader of texts. Others I have to thank collectively for engaging with my ideas when I presented some of
the chapters as papers: at Harvard University; at a mini-conference

― xvi ―

called "Learning the Rules: Schooling, Law, and the Reproduction of Social Order in Early Modern Eurasia" at the University of
Minnesota; at two international workshops on the topic "Social Embeddedness of Japanese Law" at the University of
California, Los Angeles; at a seminar at Otani University in Kyoto organized by Professor Okuwa[*] Hitoshi, where Maeda
Ichiro's[*] critique changed my thinking on some important points. I also owe my thanks to Yoshiko Kainuma, Eiko Ikegami,
and Kurozumi Makoto for assistance in deciphering and translating some recalcitrant material, to Frank Upham for some
helpful suggestions, and to Chase Langford for producing the map of the Kita-Saku area.

Finally, this study was made possible by financial support from a number of institutions over almost a decade: a fellowship
from the Japan Foundation, a research fellowship from the National Endowment for the Humanities, a research grant from
the Social Science Research Council, a research travel grant from the U.S.-Japan Friendship Commission, travel and research
funds from the UCLA Center for Japanese Studies, and UCLA Academic Senate research funds.

― xvii ―

Abbreviations
KDJ Kokushi daijiten . Edited by Kokushi daijiten henshu[*] iinkai . 15 vols. Yoshikawa kobunkan[*] ,
1979-94.
MK Mibun to kakushiki . Nihon no kinsei , 7. Edited by Asao Naohiro . Chuokoronsha[*] , 1992.

NAK-KS2 (1) Nagano-kenshi: Kinsei shiryo-hen[*] , Vol. 2, Toshin[*]chiho[*] : , 2: . Edited by Nagano-ken

. Pt. 1. Nagano-shi: Nagano-kenshi kankokai[*] , 1978.

NAK-KS2 (2) Nagano-kenshi: Kinsei shiryo-hen[*] , Vol. 2, Toshin[*]chiho[*] : , 2: . Edited by Nagano-ken

. Pt. 2. Nagano-shi: Nagano-kenshi kankokai[*] , 1979.

NAK-T 4 Nagano-kenshi: Tsushi[*]4 (Kinsei 1 ) : 4( 1). Edited by Nagano-ken . Nagano-shi: Nagano-kenshi

kankokai[*] , 1987.

NAK-T 5 Nagano-kenshi: Tsushi[*]5 (Kinsei 2) : 5( 2). Edited by Nagano-ken . Nagano-shi: Nagano-kenshi

kankokai[*] , 1988.

NAK-T 6 Nagano-kenshi: Tsushi[*]6 (Kinsei 3 ) : 6( 3). Edited by Nagano-ken . Nagano-shi: Nagano-kenshi

kankokai[*] , 1989.

NKR 12 Kinsei no jikata-machikata monjo · . Nihon komonjogaku ronshu[*] , 12 (Kinsei 2).

Edited by Nihon komonjo gakkai . Yoshikawa kobunkan[*] , 1987.

NNS 3 Ken'i to shihai . Nihon no shakaishi , 3. Iwanami shoten, 1987.

― xviii ―

NNS 4 Futan to zoyo[*] . Nihon no shakaishi , 4. Iwanami shoten, 1986.

NNS 5 Saiban to kihan . Nihon no shakaishi , 5. Iwanami shoten, 1987.

NRT 3 Kinsei . Nihon rekishi taikei , 3. Edited by Kodama Kota[*] , Nagahara Keiji , et al. Yamakawa
shuppan, 1988.

NSSS 14 Buraku . Nihon shomin seikatsu shiryo[*] shusei[*] , 14. Edited by Harada Tomohiko et al.
[*]
San'ichi shobo , 1971.

NST 27 Kinsei buke shiso[*] . Nihon shiso[*] taikei , 27. Iwanami shoten, 1974.

NST 36 Ogyu[*]Sorai . Nihon shiso[*] taikei , 36. Iwanami shoten, 1973.

TKKk Tokugawa kinreiko[*]koshu[*] . Edited by Ishii Ryosuke[*] . 4 vols. Sobunsha[*] , 1959-61.

TKKz Tokugawa kinreiko[*]zenshu[*] . Edited by Ishii Ryosuke[*] . 6 vols. Sobunsha[*] , 1959-61.

― xix ―
Map 1.
Provinces of Tokugawa Japan

Source : Conrad Totman, Green Archipelago:

Forestry in Pre-Industrial Japan (Berkeley:
University of California Press, 1989).

― xx ―
Map 2.
Kita-Saku District, Shinano

―1―

Introduction
Class, status, power, and law are not new topics to non-Japanese macrostructural treatments of Tokugawa society (1600-1868).
Scholars have described well their historical specificity, though often without discussing their interrelationship. Together, as a
cluster of general questions, they constitute a Marxian-Weberian problematic perhaps so familiar that it seems disingenuous
to persist in treating them still as questions. Indeed, one might ask what the point is of returning to them if they are in reality
answers given long ago posturing posthumously as questions—an "answeramatic" disguised as a problematic. How is this
book situated within the fields of Tokugawa history and general sociological theory? A second question a reader acquainted
with my previous interest in Tokugawa ideology might raise concerns my new orientation.[1] Why this shift from intellectual
history and representations of reality to social history and institutional practice?

Post factum justifications for what one has gotten into in the first place without clearly knowing why, except that it looked
interesting, partake of retrospective fiction. This is especially true for questions regarding one's intellectual trajectory. With
the proviso that reflections of this genre never get at real reasons, I can offer some rationale that one may expect from an
introduction.

Previous scholarship, including my own analysis of ideological constructs, approached the issues of class, status, domination,
and so on, from a position that was located in the political center. The scholar's

[1] My book Tokugawa Ideology: Early Constructs, 1570-1680 was published by Princeton University Press in 1985.

―2―

totalizing perspective—"modernist" some would call it now—from such a vantage point is the cognitive equivalent of the
rulers' reach of dominion. Past works described how the system of governance was set up, how it functioned, reproduced
itself, collapsed—and begat modernity. Even when I argued that the ideology of the times was not produced or even
commissioned by the rulers, my emphasis was clearly that it served their interests. Perhaps less explicitly, I also implied that
ideas or social representations mattered as much as they did because of their direct link to practice.

I formulated some of those structuralist and functionalist interpretations on the basis of insights gained from Pierre
Bourdieu's writings. His writings have further led me to turn my attention to ways in which institutions, like laws, not only
constitute sites of ideological embodiment but also define fields of practice and power contests. To a considerable extent,
institutions circumscribe, directly, with less mediation by ideology than one might think perhaps, what people are, do, can be,
can do, or feel instinctively to be unthinkable or impracticable. This simple truth calls for a kind of history that focuses on
what people do within and with institutions. In regard to Tokugawa Japan, this insight translated for me into a desire to
examine what peasants did and could do in their villages. How did the institutional setting of a military regime influence
intravillage power distributions? Conversely, did village politics affect institutions?

Questions of this genre have received hardly any attention in Western scholarship, and although Japanese historians have
produced a staggering amount of work about local rural history, very little of it has crossed the language barrier. Thomas C.
Smith's admirable classic The Agrarian Origins of Modern Japan touches only briefly on intravillage politics. Scholars writing
about peasant resistance (Bix, Kelley, Vlastos, and Walthall) deal mainly with what peasants did when they left their villages to
protest extravillage authority and only occasionally with the allocation, use, and contestation of power within villages. These
studies are also limited to the mid and late Tokugawa period, when the most dramatic collective protests took place.

Class has been discussed far more than status by these scholars, who either emphasize its growing importance or argue its
irrelevance. Either way, the assumption has been that status may explain collective action in the first Tokugawa century but
not later, when class in-

―3―

terests prevailed (or, according to an anticlass argument, when less stable group alliances cut across various social strata).
Schematically, class either replaced status as the motor of social stratification or failed to do so contrary to expectations. In
either case the issue was class, not status.

My critique of this kind of diachronic dichotomizing is inspired by Bourdieu's reworking of the Weberian categories of class,
status, and power through his flexible model of convertible capital (economic, social, and symbolic). Within this perspective,
class and status, far from being mutually exclusive, constitute forms of power that are complementary and occasionally
substitutable for each other. Chapters 2 and 3 of this book address this class-status question and, like the two concepts, are
hard to separate. In chapter 2 I concentrate on the role of class-based rural power even in the early period, following in this
principally the lead taken by the Japanese historian Mizumoto Kunihiko. In chapter 3 I examine the manipulation of power
through status positions in village political struggles throughout the period.

Some time ago class ceased to be an almost obligatory subject for social historians, many of whom were attracted to insights
promised by discourse analysis or the new cultural history. Bourdieu's work, however, has opened up possibilities to engage
the topic anew by "bringing in" status. In an interesting turnabout, in 1992 the well-known British historian David Cannadine
also called for a reengagement with class analysis, via the study of status.[2] I share this point of view.

When discussing class, it is important to maintain the distinction, however elementary, between objective socioeconomic
conditions shared by a number of people or groups and subjective awareness or practice (not to be taken for granted) based
upon a recognition of those conditions. Too often, either a subjective class is postulated automatically on the basis of the
existence of objective conditions (in doctrinaire Marxism) or the significance of an objective class is belittled because of the
absence of class agents (in anti-Marxist arguments). An objective class is an analytical construct that is simultaneously the
initial impetus for a query and its final result. Although classes in seventeenth-century Japan (or France) obviously do not look
the same as modern classes—a

[2] See David Cannadine, "Cutting Classes," New York Review of Books , Dec. 17, 1992, 52-57, esp. 57.

―4―

fact that is often forgotten in class analyses of Tokugawa Japan—their absence clearly is hard to argue.[3] An example of an
objective class in this study is the dogo[*] , or rural magnates, which Japanese historians, long interested in class, have
identified as existing as an identifiable socioeconomic stratum from the sixteenth through the eighteenth century. Yet even the
term that names and thus constitutes this group as a group is a modern scholarly creation. When, however, dogo[*] organized
armed resistance against land surveys in certain regions toward the end of the sixteenth century they acted subjectively as a
class.

In the seventeenth century there are numerous instances of peasants agitating around issues of village governance. Analysis
identifies these peasants as belonging successively to different socioeconomic strata and acting to further shared interests by
submitting group petitions. In this instance also, it is justified to speak both of an objective property class and a subjective
social class, although their membership is not identical. The objective class may be identifiable across whole provinces, while
peasant demands were always limited to the immediate environment; thus theirs was a social class consciousness expressed
in practice but embedded in and circumscribed by the village.

Often these village disputes concerned issues of status. At a macro and official level, Tokugawa Japan was, of course, a society
of "orders" (Stände ), legally defined status groups. Thus, all villagers were commoners, and most of them belonged to the
status order of peasants. Yet this commonality, sanctioned legally by the state (from a distant center), should not be
reconfigured into an equally distant image of village solidarity. On closer examination, it is clear that villagers always
differentiated among themselves within a clearly marked hierarchy comprising multiple status gradations. This status was in
part indirectly sanctioned legally by overlords (and thus an "order"), in part locally generated (and thus closer to the modern
meaning of social standing).

This intravillage status hierarchy, expressed in prescribed forms of social interaction and normative gradations of material
consumption, was closely related to economic power as either an addition to or a

[3] For a cogent defense of the need for a class analysis of seventeenth-century France, a "society of orders" like Tokugawa
Japan, see William Beik, Absolutism and Society in Seventeenth-Century France: State Power and Provincial Aristocracy in
Languedoc (Cambridge: Cambridge University Press, 1985), chap. 1, esp. 7-9.

―5―

substitute for it. That economic power translates into political control, which it supplements with privileges (often based on
tales of ancestry), thus creating and securing rare status positions, is not difficult to understand. Once the dog grew a tail,
however, the tale could whip the dog. Status could be made to function as a first line of defense of political privilege, even
when economic underpinnings had given way, by holding off new claimants who, pointing to their new wealth, insisted on a
share in the control of village affairs. This tug of war for political power between peasants differentiated by economic and
social status can be documented throughout the Tokugawa period.

The discussion of status and power in villages is limited in this book to those aspects of this dyad that were the subject of
explicit written regulations and contestations. This means that the players discussed here are households, not individuals. It
goes without saying that within households too power was distributed and exercised according to status, in this case along
gender and age differentials. This is the area of (mostly un- or undercodified) dispositional habits, which does not belong to a
study of explicit social and economic stratification that, moreover, extends beyond intrafamilial roles.

This limitation does not mean that I hold a monolithic view of human agency. Individual identities are plural, not singular,
because they are constituted by multiple differences in kin, gender, age, lineage, neighborhood, and regional systems. In the
public political realm, however, such multiple identities are simplified and channeled to create single group agents in
accordance with the needs of policymakers—or their challengers. In Tokugawa Japan this was especially true of households,
neighborhoods, and lineages. They were politically overdeter-mined by extravillage lordly authority. Many a village dispute
centered around the way power was distributed through these institutions.

Status is a system of stratification through socially recognized honor, esteem, and privilege. The positive connotation of these
terms, however, is misleading since such a description of a status system is the work of an outside observer who perceives
grades of more or less honor, esteem, and privilege and relates to the whole hierarchy at once. This structuralist gaze is of a
different quality than the experience of those who occupy a position within the status system. Status holders always relate to
the system differentially, from a specific position that determines the scope and intensity of their interest in other positions,
especially neighboring ones. One can at least say that, while they enjoyed public social

―6―

honor and esteem from those below them (and private envy and resentment as well), status holders must look down upon
occupants of lower positions (with fear perhaps) as undeserving of such recognition, as embodiments of dishonor. Things are
obviously more complicated than this dichotomy indicates. Even in a rigidly stratified society like Tokugawa Japan not all of
the aspects of social interaction were patterned on status prescriptions. For instance, the "outcastes," too often still referred to
by the pejorative term eta , were not treated as polluted in every social encounter. The point is that a status system is always
also a system of socially sanctioned dishonor.

This is particularly true in the case of the outcastes, victims not only of social disrespect but also of stigmatizing ostracism and
outright discrimination. For the last two decades, Japanese historians have produced a monumental body of research about
the Tokugawa history of today's burakumin (in part descendants of the outcastes), a subject virtually untouched by non-
Japanese scholars. This record, however, is not only a collection of particular histories of individuals, families, or communities.
It also reveals changes and developments in the status of outcastes as such over the whole Tokugawa period and suggests that
other factors besides cultural prejudice were at work.

Obviously, a study of status cannot neglect the subject of outcastes, although I almost did. Only after the other chapters were
written did I become aware of the need to address this difficult issue. This near oversight is not without its own significance.
The outcastes lived with their own headmen in hamlets near but apart from peasant villages and, in eastern Japan since the
1720s, under the direct jurisdiction of their seminational leader, or Danzaemon (an office name), in Edo. Since they formed a
"society outside society," as Japanese historians often put it, they have usually been treated separately in modern
historiography as well. Their serious study has been a ghettoized subdiscipline until very recently. Thus, it is possible to read
endlessly about status, class, and power within Tokugawa villages without encountering them. And yet, as I argue in chapter 5,
the treatment of these outcastes in Tokugawa and modern times constitutes perhaps the clearest and ugliest example of the
operation of status.

The topography of power within villages was not shaped solely by intravillage factors. Extravillage lordly authority helped
draw some of the lines over which confrontations took place. This authority interfered as a source of prestige and power,
directly by creating "titled

―7―

peasants" and the village offices of headman and neighborhood heads and by issuing village laws, indirectly by allowing the
officialization of village-generated statuses as they appeared on documents drawn up locally and by sanctioning through
noninterference separate, "autonomous" village codes.

Nevertheless, it has often been said that villages were self-governing to a very high degree. A headman and the neighborhood
heads who assisted him constituted the core of the village council (where only titled peasants had a seat), which took care of
all public affairs, foremost the delivery of tribute and corvée to the overlord's district intendant but also many judicial and
penal matters. In domains of the daimyo and shogun a district intendant (a bannerman in the shogunal houseland, a special
port, on of the shogunal domain) was in charge of from a dozen to over a hundred villages. Often there were also village group
headmen, who channeled communication between the headmen of a number of villages and the intendant. Tribute and
population registers were kept at both the village and the district level. Matters that could not be settled by the neighborhood
heads or by the village council or headman could be taken to the district intendant (via the village group headman) in the form
of petitions or suits or, ultimately, to shogunal or domainal higher-ups, usually in Edo. Defendants were routinely invited to
respond to the accusations in writing, thus completing the record of the plaintiff's lawsuit. The written testimonies of
witnesses or interested parties would be taken into consideration by the authorities as they tried to reach a solution—a verdict
or an ordered settlement. It should be understood that at all these levels administrative and judicial functions overlapped;
Tokugawa Japan did not have an independent judiciary and was a society without lawyers.

Although the sketch above may blur distinctions between locality and center, the degree to which Tokugawa villages enjoyed
autonomy has long been debated among Japanese historians. This debate has centered mainly on the structural position of
village regulations within the overall power distribution of society. Did these codes constitute limits to, or were they vehicles
for, the penetration of supravillage authority within the village? I engage this question in chapter 4, where I supplement the
traditional approach, which is based almost exclusively on the study of legal texts, with an analysis of actual practice. This
perspective emphasizes the articulation within a single juridical domain of two specific legal authorities, a supra- and an
intravillage authority,

―8―

which created a distinct space that allowed each one to choose whether to use the other explicitly or implicitly.

This strategic area, where village headmen or district intendants exercise some discretion in deciding which laws to apply,
overlook, or subvert (and to what extent) is, of course, hard to explore. We usually know only the outcomes of such
deliberations, not the processes by which they were reached. Local historians, however, have discovered diaries, notes, and
queries by village headmen that reveal strategies they weighed or used in applying, interpreting, or circumventing existing
regulations. All too often one feels obliged to choose between a "naive" view that equates written law with actual practice and
a skepticism that questions the effectiveness of laws. Headmen's private papers provide an escape from this simplistic
dilemma. They reveal judicious appraisals by historical actors of the relative power of law and other circumstantial
constellations of forces.

The approach to the study of Tokugawa law followed here is twofold and aims at the interplay between the normative power
of laws and actual social practice. On the one hand, in the final chapter I attempt to sketch a general macro picture, not of
Tokugawa laws, but of the juridical field and the specific power generated by laws. On the other hand, most of the other
chapters are constructed to a large extent around lawsuits and law cases, where one can witness how laws figured in various
local power struggles and contestations. To change unwanted situations, the peasants relied far more frequently on suits and
petitions than on mass protests or uprisings. One cannot avoid the impression that lawyerless Tokugawa Japan was far more
litigious than the Japan of today.

Even though contention often took the form of litigation, one should keep in mind that the distinction between petitions or
requests and what we would call suits or legal action is not at all clear. There are several reasons for this. Suits were always
drawn up according to an official format in a formal language of entreaty ("with fear and trembling") as supplications
professing in advance willing subjection to the good will of the authorities.[4] Furthermore, there were no attorneys

[4] The phrases "formal language of entreaty" and "willing subjection" are taken from Geoffrey Koziol, Begging Pardon and
Favor: Ritual and Political Order in Early Medieval France (Ithaca: Cornell University Press, 1992), 8, 45 His description of tenth-
and eleventh-century France applies well to Tokugawa Japan, perhaps not surprisingly given the "feudal" character of these
two societies: "Petition and concession, humility and grace, dependence and domination; these were the categories that
expressed the fundamental political relations of the tenth and eleventh centuries" (81).

―9―

available, and the governing authorities (village councils or overlords) were the "judges." The outcomes of suits/petitions,
rather than being "verdicts," were often conciliations or compromises without clearcut winners or losers, reached with the
help of appointed ad hoc mediators. Because restoring social order was more important than conforming to an abstract rule of
law or justice, apologies were often the decisive moment in legal closure.

Since petitioning or litigating is only one way for individuals or groups to exercise power, a regional study of such action
viewed against a detailed background of other forms of social and economic power would be ideal. I was led to explore that
possibility after coming across a particularly well documented case of a poor peasant woman engaged in a protracted
confrontation against all levels of village and even supravillage authority. The woman was from a village in Kita-Saku district
in eastern Shinano (Nagano prefecture). I begin the book with her story.

Ultimately, however, my goal was a more comprehensive statement than an area study would allow. Although much is to be
gained from regionally controlled research, I abandoned this way of organizing my project. Nevertheless, I was able to use
much of the material thus gathered, often in the form of other cases presented throughout this book, giving it a strong regional
emphasis. Thus, in good part, this is a study in rather than of a region.

A legal case usually yields a narrative, a fragment revealing what people cared about and why. Knowledge of circumstantial
detail is often essential for an adequate understanding of the issue. An analysis of practice through a number of case studies
with ever-changing actors and settings, however, can be taxing to the reader. The detail can at times be bewildering and
overwhelming. Yet, the alternative of a quick summary of the issues to please the hurried reader sacrifices too much of what
makes history—an old-fashioned idea—interesting and what reminds us that cases are about people.

It is just as appropriate to remind oneself that we get to know those

― 10 ―
people because others, namely, local historians, have retrieved their stories. Without their assiduous efforts, a study like this,
which touches upon the history of many locales, would be impossible. With this in mind, I have chosen not to relegate their
names to footnotes in order to protect a seamless narrative and uninterrupted analysis. The reader will have to bear with this
reminder in the text that it is the work of these historians that carries the study forward.

To a considerable extent, this book is a collection of essays that can be read independently and in almost any sequence.
Linking the different topics and chapters, however, is the fact (and I use the term without hesitation) of inequality and often
the reality (again, no apologies) of social suffering as they were institutionalized in the communities of Tokugawa villages. The
frequent description of villages as harmonious and consensual is a misrecognition if not an outright denial of these realities.

― 11 ―

1
"Mountains of Resentment"
One Woman' s Struggle against Tokugawa Authority
Follow your judgment.
Ken, 1757

In this study I often rely on data from Nagano prefecture, formerly Shinano province, especially from the Saku district in the
east (since 1624 split into a northern, Kita-, and a southern, Minami-, half). This is in part accidental and in part by design and
deserves an explanation. (For provinces, see map 1; for Kita-Saku, see map 2.)

Learning of my interest in Tokugawa village law, my colleague Anne Walthall recommended an article about a peasant
woman who sued officials in Makibuse, a village in Kita-Saku. As it turned out, this case provided an unusually well
documented entry into village politics. Additional reading drew me further into a study of the region, which, I soon found, had
been explored by many prominent historians of the early Tokugawa period, including Kodama Kota[*] , Miyagawa Mitsuru,
Oishi[*] Shinzaburo[*] , and the pathbreaking local historian Ichikawa Yuichiro[*] . Independent scholars continue this
tradition of local history today. Ozaki Yukiya is one of them.

Like Ichikawa before him, Ozaki until recently supported his passion for local history by his work as a high school teacher. He
wanted to become a professional historian, but in the immediate postwar period his father, convinced that science was the
wave of the future, steered his son in the direction of physics and chemistry. Fortunately, Ozaki was able to squeeze some
courses on Tokugawa documents into his schedule at the university. When I established contact with him in 1991, he was
teaching high school science by day and researching and writing local history by night and on weekends.

Initially interested in questions of discrimination against burakumin , Japan's indigenous minority, Ozaki organized a club in
the mid sev-

― 12 ―

enties to explore the history of Mochizuki, the town into which Makibuse has been incorporated. He taught this small group of
retired history buffs how to read the difficult Tokugawa texts, and together they began collecting, cataloging, and reading tens
of thousands of documents from the storehouses of descendants of Tokugawa village headmen in the area.[1]

In the storehouse of the descendant of Makibuse's headman, Ozaki came across dozens of documents referring to a woman
named Ken. Intrigued, he devoted a great deal of time to piecing together the fragments of her story. I retell it here, following
the trail of documents, supplemented with relevant historical data from other sources.[2]

The Registration Incident: 1761-1763

Our point of entry into Ken's story is an incident involving the population register of Makibuse that is neither the beginning
nor the end of Ken's story (see table 6 for a chronology of Ken's life). I chose this incident because often long-simmering
questions regarding status surface in the historical record as disputes around the proper entry in village population registers,
which were updated every year. These registers recorded not only a person's name but also his or her social and

[1] Currently the group is writing a multivolume history of the town and its surrounding villages. In the fall of 1992 Ozaki
resigned from his high school teaching position to accept an appointment with Nagano's new prefectural museum, scheduled
m open near Nagano city at the time of the 1998 Winter Olympics. As director of historical acquisitions for the Tokugawa
period, he is at last pursuing his passion for local history on a full-time basis.

[2] Today Makibuse is part of Mochizuki, a township with some 3,100 households and a population of 11,300 that since 1959
has included a number of surrounding villages, each of which in turn comprises what were once several Tokugawa villages.
For the materials on Mochizuki, see Mochizuki-machi kyoiku[*] iinkai, Mochizuki no burakushi , nos. 1-5 (1976-79), and
Mochizuki no chomin[*]no rekishi , nos. 6-15 (1980-91). The documents related to Ken are still in the possession of the
descendant of the Tokugawa headman of Makibuse village, who was reluctant to allow me access to them on my visit there in
the summer of 1991. Fortunately, Ozaki had made copies, which he graciously put at my disposal. Ozaki published his findings
on Ken in two partly overlapping but mostly complementary articles: "Kenjo oboegaki: Kinsei noson[*] joseishi e no kokoromi,"
Mochizuki no chomin[*]no rekishi 8 (1983): 61-100; and "Shinshu[*] Saku-gun Makibuse-mura Kenjo ikken—Kinsei noson[*]
joseishi shikiron m shite," Rekishi Hyoron[*] 419 (1985): 45-66. References to sources other than those directly pertaining to
Ken's story are my own. The story has become famous locally through a play based on it.
 ― 13 ―

legal status. Individuals contesting their status might cause delays in the village headman's submission of the registers to the
higher authorities, or they might escalate the confrontation into a lawsuit, in either case leaving evidence for historians to
examine. In Ken's case, we are well informed about the immediate issue she was raising and the pressures applied to her to
conform. Behind this incident, however, lies the story of a peasant family, which adds a number of other dimensions to this
incident.

Population registers, along with village laws and land cadasters (chronologically the earliest of the three), were the most
important documents the supravillage authorities ordered the village headmen to keep up-to-date. Titled peasants had to
signal agreement with the registers and laws for themselves and their dependents by affixing their seals.[3] In principle, the
laws were read at least once a year to all the villagers, sometimes, as in Makibuse, at every formal meeting of the village
council.[4] The list of signatures attached to these laws eventually evolved into separate, complete population registers
(jinbetsu aratamecho[*] ). The earliest such register in Kita-Saku dates from 1642. A record of Buddhist affiliation, required
nationwide since 1665, was first added to these population registers but came to be drawn up separately and known as
shumon[*]aratamecho[*] .[5] These lists often functioned as population registers with increasingly detailed information on the
age, provenance, and whereabouts of each co-resident of each household and his or her relationship to the household head. By
the 1720s a full-fledged, nationwide population census was being conducted every six years. The headman, assisted by the
kumi heads (the heads of the official neigh-

[3] On the process of diffusion of these laws from the lord to the peasants and their ratification by the village, see Yokota
Fuyuhiko, "Kinsei sonraku ni okeru ho[*] to okite," (Kobe[*]Daigaku daigakuin bunkagaku kenkyuka[*] ) Bunkagaku nenpo[*] 5
(1986): 150-55.

[4] Ichikawa Yuichiro[*] , Saku chiho[*]Edo jidai no nomin[*]seikatsu (Nagano-ken Minami Saku-gun Nozawa-machi: Saku
Insatsujo, 1955), 24. As is so often the case with Tokugawa institutions, however, some of these laws antedated the bakufu
order: the oldest one in Shinano is from Kyowa[*] , a village five kilometers west of Makibuse, and dates from 1628 (ibid., 15).

[5] For an introduction to the shumon[*]aratamecho[*] , see L. L. Cornell and Akira Hayami, "The shumon[*]aratame cho[*] :
Japan's Population Registers," Journal of Family History 11, no. 4 (1986): 311-28. For a recent discussion of the development of
population registers, see Yokota Fuyuhiko, "Kinseiteki mibun seido no seiritsu," in Mibun to kakushiki , ed. Asao Naohiro,
Nihon no kinsei, 7 (Chuokoronsha[*] , 1992), 41-78, esp. 55-78 (hereafter MK).

― 14 ―

borhood groups, the goningumi ), was responsible for drawing up a new register every year and having each entry properly
verified and certified with each household head's seal. (For special terms, please consult the Glossary.)

In Kita-Saku district, as elsewhere in bakufu , or shogunal, territories, the yearly population check took place in the third
month.[6] A copy of the register remained with the headman, the original being forwarded to the lord's chargé d'affaires, who
for Makibuse in 1763 was quartered at an office in Shimogata village, the administrative center for a shogunal bannerman's
(hatamoto ) small fief by the same name.[7]

Thus, on 1763/2/8,[8] a woman named Ken refused to certify her household's entry with her seal. A yearly routine had taken on
the dimensions of an event. No doubt greatly annoyed by this flagrant case of insubordination by a mere woman, which would
reflect badly on his own governance and certainly trigger an investigation from above, the headman, one Chuemon[*] , had no
way of concealing the matter, so he reported it to the Shimogata office.

Why was a woman's seal required? Usually the male head of the household held legal authority for all its members, whether
affines, cognates, or other co-residents. And Ken was neither single nor a widow. In 1756, then thirty-seven, Ken had married a
man a few years her elder. He was from Mochizuki, a way station on the Nakasendo[*] inland highway between Edo and
Kyoto, located across a low range of hills

[6] Ichikawa Yuichiro[*] , Saku chiho[*] , 70.

[7] Between 1702 and 1765, Makibuse was one of thirteen villages in the area that together constituted the five-thousand-koku
fief of Shimogata (see Nagano-ken, ed., Nagano-kenshi: Kinsei shiryo-hen[*] , Vol . 2, Toshin[*]chiho[*] , pt. 1 [Nagano-shi: Nagano-
kenshi kankokai[*] , 1978], 724 [hereafter NAK-KS2 (1)]). Its overlord history, like that of most villages in the fertile Kita-Saku
plain, is a complex one. Villages were shifted numerous times, sometimes even divided, among various lords. Makibuse
started as a village of the fudai domain of Komoro (first under the Sengoku house, 1590-1622; then under the shogun's brother
Tadanaga, 1622-24; then under a Matsudaira, 1624-47). Between 1648 and 1661 it was entrusted as a shogunal possession
(azukaridokoro ) to the Aoyama. Subsequently it was part of the Kofu[*] Tokugawa domain until 1700. After two years of
incorporation in the shogunal domain proper (tenryo[*] ) under an intendant (daikan ), it was transferred to the jurisdiction of
the Shimogata bannerman's fief until 1765. From that year until the end of the Tokugawa period it was again part of the
tenryo[*] .

[8] Dates are referred to by year first, followed by the lunar month and the day according to the traditional Japanese calendar.

― 15 ―

some two and a half kilometers due west of Makibuse. His name was Rokuemon, formerly Yohachi. Two years before the
incident, however, in the winter of 1761, he had left Ken, and his whereabouts were unknown.
At the time of the first registration after her husband's disappearance, in the spring of 1762, Ken requested to keep him on the
rosters as the in absentia household head. The following year, however, she officially petitioned the headman, in a document
legalized with the seals of her relatives and kumi members (the same persons, since in Makibuse lineages overlapped with
neighborhood kumi),[9] to take him off the rosters, leaving her the head of a single-member household, for her parents were
no longer alive and she had no heir.

This Ken did on 2/6, that is, only two days before she surprised the headman by refusing to affix her seal to the population
register. Had her relatives suspected the trouble she was to cause, they certainly would never have signed the petition.
Instead, they would have placed her under the fiduciary authority of one of them, as she had been before her marriage. Did
they not know Ken well enough to foresee this? Was this why she had sought legal authority? What lay behind this surprise
move?

In his report to the Shimogata magistrate, the headman carefully explained the futile efforts by the village authorities (himself
and the four kumi heads) to make this recalcitrant woman comply with shogunal law and also recorded Ken's reason for her
behavior:

Ken did not affix her seal to this year's population register. What was the reason for this delinquency? We asked her
to please affix her seal, there being regulations concerning the population register. Even if I did it, [she said,] come
fall I intend to leave the village, so there is no need for my seal. In any case, [we said,] it is an offense not to affix
your seal this year. All the village officials urged her strongly to quickly affix her seal, but no matter what, she did
not listen and [just] went home.[10]

The headman also mobilized the members of the next authority level down, namely, Ken's kumi members/relatives, asking
them to persuade

[9] Kumi is short for goningumi , the term for a five-household neighborhood association, an official administrative subdivision
of a village. Their membership was supposed to cut across, rather than overlap with, lineage membership. Makibuse had four
kumi.

[10] Ozaki, "Kenjo ikken," 47-48.

― 16 ―

her, but they were equally powerless. He finally ordered her to appear in person in Shimogata, about seven kilometers to the
east of Makibuse, across the hills, in the open plateau of Kita-Saku, on the other side of Gorobe-shinden, near the Chikuma
River.

But Ken did not budge. "If this is an offense, then it is one I have been looking for that suits me well (ochido ni ainarisoro[*]mo,
saiwai nozomu tokoro ni gozaro[*] )," she replied, according to the headman's report, "and even if I receive the death penalty or
banishment, no matter what punishment [might await me], I have thought this over very well [and taken all such possibilities
into consideration]." Thus Ken not only categorically refused to appear before the Shimogata magistrate—she interpreted the
order to do so as a threat—but used the occasion to impress upon the headman again the depth of her determination,
countering his order/threat by warning him that threats would be of no use. At least this is the headman's view of the
confrontation, aimed at impressing his superior with the particular difficulties he was having with this woman.

Ken's leaving the village would result in the disappearance of a household, which was always an extremely grave matter, for it
affected the distribution of tribute that the village as a whole owed the lord. The lord, who "owned" the land, which "owned"
the people on it, did not let them go easily. Yet the land under Ken's name was minuscule, a mere 0.19 koku , three-quarters of
it a dry plot attached to the homestead.[11] The reassignment of her land to another peasant in order to produce the
apportioned part of the village tribute quota could not have been the only thing at stake. As we shall see, there were in the
village poverty-stricken peasants in need of land who presumably would have gladly taken over the holding. It seems
particularly nasty of the headman to deny this woman her request to leave the village. Was he getting back at her for some
unspecified reason?

When moving to a domicile out of one's locale of registration (village or township), one had first to secure a certificate of leave
(okurijo[*] , or "sending document") in order to properly register in one's new resi-

[11] The term koku , a measure of volume equivalent to 180 liters or 44.8 gallons, refers to the putative rice yield of fields, the
basis for the computation of tribute. All property, including dry fields and homesteads, was converted into koku equivalents
for tribute purposes. Ken's property was thus registered as equivalent to a mere 34 liters of rice.

― 17 ―

dence. Ken knew the procedure; she had followed it a year earlier, without, however, being granted her wish. This time she
seemed to have changed strategy, hoping, perhaps, to force the hand of the village authorities or at least to cause trouble by
exploiting the jurisdictional division between village and district.[12]

In 1762/3, only a year earlier and thus around the registration time when she had kept her absent husband recorded as
household head, Ken had petitioned to leave the village but the petition had not been granted, possibly because her husband
still held legal authority for the household. In 1763, however, she held the title and seal of household head. The petition of 1762
read as follows:

In the winter I was asked [by her lineage and kumi members and possibly also by the village officials] m establish a
successor for the household [through adoption of an heir or remarriage], but I do not want to do that. I want to
work the fields but cannot even become a tenant. So I have no alternative because I have trouble paying the tribute
 and performing corvée. Thus I request to donate the residence land, paddies, dry fields and woodland, all of it, to
my family temple, the Baikei-in [in the adjoining village of Iribuse]. If this request is honored, then I and Rokuemon
[my husband] will have to be taken from the population register. I want to leave Makibuse. By making this request,
[I hope] I am putting an end to your order [to establish a successor]. There is no need to call me and talk to me about
this again and again. Would you please examine my proposal.[13]

This petition throws some light on the headman's ill will. He refused to grant Ken's request to leave after commending her
property to a temple because she had balked at his order to establish an heir. There is evidence here of the same
uncompromising, blunt determination that comes through in her refusal to sign the register a year later, when she defiantly
told the headman that "I do not care for my life, no matter

[12] Peasants crossing domain boundaries en masse resorted to a similar strategy on a larger scale. Perhaps the most famous
such mass exodus was an 1853 crossing from Nanbu to Sendai domain led by Miura Meisuke (see Herbert Bix, "Miura Meisuke,
or Peasant Rebellion under the Banner of 'Distress,'" Bulletin of Concerned Asian Scholars 10, no. 2 [1978]: 18-28).

[13] Ozaki, "Kenjo ikken," 48. Makibuse did not have a temple. The village may have split off at some time from Iribuse, where,
according to the 1671 population register, 74 of the 133 inhabitants of Makibuse were registered with the Baikei-in; 49 were
registered with a temple in Mochizuki (doc. 351 in NAK-KS2 [1]: 779-80).

― 18 ―

what happens to me; I am not in the least afraid; there is no need for me to go to the [Shimogata] office."[14]

This was an angry woman who, all by herself, challenged every level of authority: the shogunal magistrate, the village
headman, the kumi heads, her kumi members, and her relatives. She stood up against them by refusing to straighten out the
succession of the house, sign the register, and appear before the Shimogata magistrate. The headman reports that when her
lineage members put pressure on her in the registration incident, "it was useless, she would not consent and started even
saying slanderous and irresponsible things." Taking another tack, they pleaded understanding for their own plight: as the
group responsible and liable for its members' behavior they would face difficulties because of this incident. But she responded
to this plea as she had to the headman's threat, retorting that this was precisely the point: "I have been waiting for this
opportunity with the population register in order to cause trouble for the goningumi (goningumi nangi itashisoro[*]tame ni
gochomen[*]inkei no setsu to aimachi makari arisoro[*] )." Ken had planned this. What in Ken's past had brought her to this bold
confrontation?

Goningumi, or official neighborhood groups, were not supposed to overlap with family networks: the first article of laws of
some villages in the area, like Makibuse also under bakufu jurisdiction, specified that these kumi ought not to include only
kinsmen or close friends, kith and kin (shinrui naka yoki mono ).[15] This rule obviously was aimed at preventing collusion of
interests and increasing local control. In the neigh-

[14] Ozaki, "Kenjo ikken," 49.

[15] Ichikawa Yuichiro[*] (Saku chiho[*] , 13-63) studied forty-one goningumicho[*]zensho , or village laws, of Saku district
covering the whole Tokugawa period. Included in these are four versions of the Makibuse village laws, from 1687, 1784, 1801,
and 1857. He discovered great similarities, if not complete identity, among many of them across jurisdictions (domains and
bakufu territories of various kinds) in three chronologically distinct periods separated by the 1660s and the second decade of
the eighteenth century. He appended to his study the full text of the standard version of each period: the laws of the bakufu
village of Shimo-Sakurai, nine kilometers east of Makibuse, of 1640 (21 articles), 1662 (56 articles), and 1766 (also 56 articles).
(The first two of these laws are translated in appendixes z and 3.) Among the Makibuse laws, the one of interest to us is the
first one, dated 1687, which according to Ichikawa is identical, down to the number of articles, to the 1662 version from Shimo-
Sakurai. When citing articles of the village law of Makibuse, it is to this text (appendix 3) that I refer. The article about the
proper composition of goningumi is the first one in the 1640 version of Shimo-Sakurai village law (appendix 2).

― 19 ―

boring village of Kodaira, kin-exclusive kumi had led to trouble;[16] in Makibuse also family networks and kumi overlapped:
the six households of Ken's lineage all belonged to the same kumi. In this region, shogunal law clearly had failed to break the
power of the local self-governing bodies of the lineages.

One of Makibuse's Poor Peasants: 1719-1744

Ken was a native of Makibuse, the daughter of a peasant named Rihei, whom we shall refer to as Rihei II since her grandfather
bore the same name (hence we will call the grandfather Rihei I). Born in 1719, she was forty-four at the time of the registration
incident. Her mother was from Yawata village, a station on the Nakasendo[*] two and a half kilometers east of Makibuse (the
last station in the Kita-Saku plateau before the highway winds its way westward into the mountains to the next station,
Mochizuki, where Ken's husband had come from); Makibuse's northern border also touched the Nakasendo[*] . In 1694, at the
age of sixteen, slightly earlier, perhaps, than was usual for girls, Ken's mother had married Rihei II.[17] She bore him a son,
Shinzo[*] , in 1700; a daughter, Ine, twelve years later; and finally Ken, her last child, seven years after that, in 1719.

Ken's father died the year she was born. By then Shinzo[*] was nineteen, old enough to work the fields, and Ine, seven, could
help care for her baby sister. At age fifteen, rather young, Ine married out to Yawata, where her mother (and paternal
grandmother)[18] were from, perhaps to reduce the number of mouths to be fed in the family.[19] Ken's first

[16] See the discussion of this issue in chapter 3.

[17] Ichikawa Yuichiro[*] (Saku chiho[*] , 113) found that the average marriage age in the area was twenty-three to twenty-four
for men and sixteen to seventeen for women. These are a couple of years lower than Hayami's findings for another district in
Nagano or Smith's findings in his study of Nakahara (Hayami Akira and Nobuko Uchida, "Size of Household in a Japanese
County throughout the Tokugawa Era," in Household and Family m Past Time , ed. Peter Laslett [Cambridge: Cambridge
University Press, 1972], 502; Thomas C. Smith, Nakahara: Family Farming and Population in a Japanese Village, 1717-1830
[Stanford: Stanford University Press, 1977], 94-96).

[18] See the sixth entry in the 1671 population register of Makibuse, in NAK-KS2 (1): 771.

[19] Smith (Nakahara , 93-95) found that daughters of wealthier families married considerably younger than those of poor
families. He suggests that this may have to do with the fact that the poor spent time as servants to supplement family income.
In other circumstances, however, early marriage of daughters may have been a solution to the problem of too many mouths to
feed.

― 20 ―

Plate 1.
Makibuse Village. Facing west toward Mochizuki, some three hundred
meters south of the Nakasendo[*] . Photograph by author.

marriage, in 1737 at age eighteen, brought her to Mimayose, two kilometers beyond Yawata on the Nakasendo[*] , leaving her
mother and unmarried brother in Makibuse.

Shinzo's[*] bachelor status at age thirty-seven needs some comment, although a specific psychological or other personal
explanation is not available. Perhaps he had been married and was divorced; we do not know. The thought that he might have
been an unattractive prospect seems culturally incongruent. Males must have had the strategic advantage in a society like that
of Tokugawa Japan. It is possible that Shinzo[*] could not afford a wife, that the reason of his bachelorhood was sociostructural
rather than personal. In 1671, at least, the only bachelors of marriageable age in the village were indentured ser-

― 21 ―

vants.[20] A socioeconomic explanation would thus point to the near-poverty situation of the family, although even "landless,"
that is, tenant, peasants did marry, as did Ken.[21]

Outmarriage was quite common in Kita-Saku district from fairly early in the Tokugawa period, contrary to Smith's assumption
of a "village rule of endogamy except for high-placed families who had to go outside the village to find marriage partners of
comparable family rank."[22] The 1671 population register of Makibuse recorded thirty-two peasant households (including
fourteen titled peasants), all but two headed by couples (see table 1). In addition, there were seven married sons, brothers, and
nephews of titled peasants co-residing with their main families in an extended-family pattern, which makes a total of thirty-
seven couples. Only seven of the thirty-seven wives were from Makibuse; all the others had inmarried from villages in the
same district or, in one case, from another district. Yawata tops the list of villages as sources for wives, with five. As for
outmarriage, fourteen daughters or sisters of current household heads married out. Only two Makibuse women married out to
the Kita-Saku plain; the rest settled in the mountain area. On the other hand, ten wives came from the plain. All twenty-three
bond servants (genin and fudai ) came from the mountain area, from within a radius of eight kilometers.

Thus, the number of women marrying within the village (seven) was

[20] According to the population register of 1671, none of the bond servants (genin and fudai) were married. There were
nineteen indentured servants (genin) in Makibuse—twelve males ranging in age from thirteen to thirty-four and seven
females between fourteen and twenty-three—and four lifetime servants (fudai) between the ages of nine and twenty-two (see
NAK-KS2 [1]: 771-80).

[21] Smith (Nakahara , 92) found that the probability of male celibacy in lower-class peasant families (those with holdings of
less than four koku) was from two to ten times as high as that among wealthier peasants (67 percent for the cohort aged thirty
to thirty-four and 41 percent for those aged thirty-five to forty-nine). Pierre Bourdieu discusses the structural necessity of
bachelorhood of nonsuccessor males for reproducing the social system in the late nineteenth century m the Béarn in southern
France in "Célibat et condition paysanne," Études rurales , nos. 5-6 (1962): 32-135.
[22] Thomas C. Smith, The Agrarian Origins of Modern Japan (Stanford: Stanford University Press, 1959), 61, 62-63 n. 1. My
occasional critique of this excellent study should be understood in light of the vast amount of Japanese scholarship produced
in the last thirty years. In Nakahara (133-34, 143) Smith reports only that "nearly all daughters were married our of the family"
but gives no figures for out-village marriage. See also the following note.

― 22 ―

Table 1.
Population of Makibuse, 1671-1764

Population Households

Male (genin ) Female (genin Total Horses Honbyakusho[*] Kakae Mizunomi Total [Kokudaka
) ]

1671 73 (13) 60 (10) 133 — 14 18 0 32 [175]

1700 93 (18) 80 (10) 173 — 16 24 0 40

1717 121 (12) 104 (10) 225 18 13 30 4 47 [205]

1725 129 (14) 99 (11) 228 16 12 33 3 48

1737 136 (15) 112 (9) 248 16 12 38 2 52

1750 130 (11) 112 (9) 242 12 11 35 2 48 [205]

1764 129 (11) 115 (9) 244 10 11 43 3 57

SOURCES : Adapted from Ozaki, "Kenjo oboegaki," table 9 (p. 97); the population for 1671 is from NAK-KS2 (1): 771-81, and
the kokudaka is from NAK-KS2 (2), appendix p. 53.

NOTE : Land development seems to have reached its capacity in the late seventeenth century. The kokudaka of 205 was
first recorded in 1704 (NAK-KS2 [1]: 613); between 1704 and 1834 another 16 koku were added.

― 23 ―

only half of that marrying out, and this in a community where thirty-seven wives were needed for households to reproduce. If
one adds all marital liaisons in and out of the village, the total is fifty-one, out of which only seven, or 14 percent, were
endogamous to the village.[23] This contrasts sharply with 50 percent and 75 percent for other villages in the area
(Komiyayama in 1669 and Wada in 1713, respectively), as Ichikawa Yuichiro[*] relates.[24] He also concludes that the average
marriage age in the district was twenty-three to twenty-four for men and sixteen to seventeen for women, and he notes that
almost all of the indentured servants (genin) and lifetime servants (fudai) were unmarried.[25] In 1671 none of Makibuse's
nineteen genin or four fudai were married.[26]

Let us return to the narrative. In 1742, five years after her first marriage, Ken's name reappears on Makibuse's family register,
which means that she had left Mimayose and her husband, although it is unknown whether this was a divorce. The household
she returned to, however,

[23] Finding a wife appropriate to one's economic status, which might necessitate liaisons outside the village, does not seem to
have been a motive either. After the 1670s the overwhelming majority of the peasants in Makibuse were very small
landholders. The seven marriages within Makibuse that were recorded on the 1671 population register were as follows
(honbyakusho[*] means "tired peasant"; kakae refers to a branch house in a lineage):

1. Headman (29 koku) and daughter of honbyakusho[*] (5 koku; 21 koku before partition)

2. Honbyakusho's[*] son (6) and kakae (4) daughter

3. Kakae (2) and kakae (5) daughter

4. Kakae (2) and kakae (5) daughter

5. Honbyakusho[*] (5; 23 before partition) and kakae (5) daughter

6. Kakae (5) and honbyakusho[*] (7) daughter

7. Honbyakusho[*] (7) and kakae (10) daughter

[24] Ichikawa Yuichiro[*] , Saku chiho[*] , 83. On the other hand, Makibuse's geographic distribution of spousal provenance is
very similar to that of Kodaira (a village studied in more detail in chapter 3), some four kilometers to the east. The latter's
population register of 1694 lists fifty-four couples, in only six of which both partners had been born in Kodaira; in the other
forty-eight, one spouse, usually the wife, came from within a radius of twelve kilometers, thirty-six from within a radius of
four kilometers (see Komonjo kenkyukai[*] dai ni han, "Genroku-ki no shumon[*] aratamecho[*] o miru," Mochizuki no
chomin[*]no rekishi 12 [1988]: 64-65).

[25] Ichikawa Yuichiro[*] , Saku chiho[*] , 110, 113.

[26] See n. 20. For comparative data on lifetime servants and indentured servants in the area, see ibid., 108, 120 ff.

― 24 ―

consisted now only of herself and her mother. Shinzo's[*] name had disappeared from the register. The possible reasons for
such disappearance are limited to a few: death, outmarriage to another locale (temporary employment elsewhere would be
duly recorded and not cause removal of the name), petition by relatives and kumi members for reasons of disinheritance
(kando[*] or kyuri[*] ), or abscondence to an unknown destination. The authorities kept a close tab on the whereabouts of all
subjects.

The previous year's register offers no clue about Shinzo's[*] fate. Ken's return, however, may have been prompted by the need
to take care of her mother, then sixty-four years old (or seventy, according to the entry of that year), and the minuscule plot of
land (0.19 koku, the equivalent of 2 ares or 0.04 acres, of medium-quality dry fields) for which her mother was responsible in
the village. Or was this a pretext to get out of a marriage that was not working? Ken was left completely alone when her
mother died in 1744, two years after her brother's disappearance. She was then twenty-five.

The death of her mother brought a change in Ken's "civic" status, as we know from the next year's entry. Until then the
household had been registered as a kakae , or branch family, of her uncle Gendayu[*] , her deceased father's younger brother.
Kakae literally means "embrace" or "hold in one's arms" and refers to the client-patron relationship of dependency between
branch and main houses; the term will be rendered hereafter as "branch house," "fully established branch house," or simply
"client." Now she was legally incorporated into her uncle's household and entered under his name (she was thus chonai[*] , or
"on [his] register," that is, co-resident), although it is unclear whether she actually moved into his house, since she had one of
her own.

In order to grasp the significance of this change, one needs to understand the highly structured village social and political
hierarchy. First, there was the great divide between peasants and nonpeasants, perhaps as great as the divide between rulers
and ruled in the society at large. The nonpeasant population increased throughout the Tokugawa period as craftsmen, doctors,
and others took up residence in the countryside in increasing numbers.[27] But these nonpeasants were marked in the
registers as belonging to an inferior class: all of them were registered as clients or bond servants (genin).[28]

[27] For Saku district, see ibid., I44-46.

[28] Ibid., 128.

― 25 ―

The first article in the village laws, including Makibuse's, reveals this hierarchy in the context of certifying that no Christians
were present in the village: "Art. 1. Re: Christians. Following the investigations of the past, each and every one, down to the last
person, has been thoroughly examined: not only (moshioyobazu[*] ) house owners [but also] men, women, children, servants
and semi-independent branch houses (kadoya ), renters, fully established branch houses (kakae), down to (sono ta ... itaru
made ) Buddhist monks, Shinto priests, mountain priests (yamabushi ), ascetics, mendicant monks with flutes or bells, out-
castes, common beggars and registered beggars (hinin , "nonhumans"), etc...." (see appendix 3). This comprehensive list is
clearly hierarchized; none of these undesirables seem to have resided in Makibuse.

Nonpeasants were second-class citizens within the village, yet some peasants also fell into this category. At the bottom were
mizunomi-byakusho[*] , literally, "water-drinking peasants," without land of their own, pure tenants whose number increased
over time. Next were non-titled landholders, who, although they were autonomous proprietors, were incorporated into a
political dependency relationship as real or fictitious fully established branch houses (kakae) to main families, lineage heads,
or patrons (kakaeoya ). Thus were constituted lineages and sublineages, crisscrossing or overlapping the kumi (as in
Makibuse). All landowning peasants, even immigrants, belonged to a "lineage." Prior to Ken's mother's death, her household
was a kakae under her uncle Gendayu[*] .

The titled peasants (honbyakusho[*] ) constituted the village council, and all lineage heads were titled peasants. In the eyes of
overlords, they were the official tribute deliverers. The two categories of landowning peasants—fully established branch
houses and lineage heads—often had a number of dependents: lifetime (fudai) and indentured (genin) servants, semi-
established branch families living in a separate dwelling on the premises (kadoya), renters, and dependent co-resident
members, like Ken after her mother's death.
This multilayered class, status, and social hierarchy, which allowed for some controlled mobility (specified later for Makibuse),
was thus from the bottom up: nonpeasants, lifetime servants, indentured servants, co-residents, pure tenants, semi-established
branch houses, fully established branch houses, titled peasants (who allowed for further power combinations with the
suprahousehold authority positions of lineage head, kumi head, and headman).

― 26 ―

In 1721 Makibuse counted thirteen titled peasants (the same number as in 1647 but three fewer than in 1700 [see table 4]) and
thirty-six others, among whom were four landless peasants who first appear on the rosters around 1710 (see table 1). In 1721,
moreover, of the forty-five landholding peasants sixteen had holdings of less than 1 koku, eighteen had between x and 4 koku,
and eleven had 5 or more koku. Ken's household was among the poorest in Makibuse. With its 0.19 koku, it was dangerously
close to landless status. Ken was not a titled peasant, and her shift in status from fully established branch house (kakae) to co-
resident (chonai[*] ) meant the loss of an already very circumscribed control over her own life.

Failed Marriages, Misfired Forgeries:1745-1756

A year after her mother's death, in the winter of 1745, Ken married someone from Kannonji-shinden, six kilometers northwest
of Makibuse, into the mountains. This second marriage, however, was not one that would provide her house with a male head
or successor because she had married out . Apparently, in this case, unlike in 1762-63, the village authorities put no obstacles
in the way of Ken's leaving. Could the only difference have been that this time she seems to have had no plans to transfer her
land to a temple? Ken was supposed to take up residence with her husband in Kannonji-shinden. Yet, oddly enough, she did
not seize this opportunity to leave the village, because she either returned very soon to Makibuse or remained there all along.

The reason for this short-lived union was recorded as having to do with Ken's husband's postponing the building of their new
house because of some problems with its proper geomantic orientation. One is led to believe that there was procrastination
and deliberate reluctance on Ken's part because her patron uncle and relatives showed impatience, making her sign a promise
in 1746/3 "to move without fail (soinaku[*] ) to Kannonji-shinden by next spring." This time her lineage wanted her out badly,
but she did not want to go—the opposite of the situation in 1762-63. The population register of two years later (1748), however,
indicates no change: Ken had not moved to Kannonji-shinden, and she remained listed as co-resident under her uncle
Gendayu[*] . Ken's return did not signal a firm determination to remain in Makibuse, though, because she went on a three-year
stint as a servant

― 27 ―

in two places quite different from Kannonji-shinden: Hirabara and Komoro. This absence, however, did not sever her ties with
Makibuse, where she remained registered.

Kannonji-shinden was a very small village of Komoro domain, that never amounted to much as it expanded quickly to its full
capacity of a meager sixty-four koku.[29] Ken's places of employment were not in the mountains but in the wealthier Kita-Saku
plain. She worked first for one year in Hirabara, some fifteen kilometers northeast of Makibuse, then for two years in the
castle town of Komoro, the center of the small fudai domain some thirteen kilometers almost due north (gradually reduced by
1700 to 15,000 koku from an original 50,000), of which Makibuse was a part for some time (1590-1647). Both Hirabara (a
sizable village of 1,474 koku in 1647, 2,018 in 1834, with a population of 602 in 1760) and the town of Komoro (with 357
households and a population of 3,620 in 1746) were situated on the Hokkokukaido[*] , the highway that splits off to the north
from the Nakasendo[*] in Oiwake at the foot of Mount Asama (see map 2).

It seems that, after her misfortunes, Ken sought to escape from her kin and had no intention of continuing the household by
adopting an heir or remarrying.[30] Nor did she want to leave the village for good by marrying out to a rather remote place in
the mountains like Kannonji-shinden. After her service, she reappears on the village register in 1752 as a resident, and then
we lose track of her until her third marriage, in 1756.

There is a short document dated 1755/12 that was signed by the priest of a temple from Koshigoe, some sixteen kilometers
northeast of Makibuse, beyond the mountains in the neighboring district of Chiisa-

[29] Originally, Kannonji-shinden was part of Innai village, which in 1624 had a kokudaka of 526 koku, to which 88 koku of
new fields (shinden ) had been added by 1703. It remained unchanged until the end of the Tokugawa period (at least until
1834). By 1703 Kannonji-shinden had split off as a separate village with 64 koku, which also remained unchanged. Its
population was registered as seventy-nine in 1760. For these figures, see NAK-KS2 (1): 7, 27, 120; and Nagano-ken, ed., Nagano-
kenshi: Kinsei shiryo-hen[*] , Vol. 2, Toshin[*]chiho[*] , pt. 2 (Nagano-shi: Nagano-kenshi kankokai[*] , 1979), app. 53 (hereafter
NAK-KS2 [2]).

[30] There may have been some affinal connection explaining Ken's choice of Hirabara, although this is sheer speculation.
According to the 1671 population register, Heijiro's[*] wife came from there and Heijiro[*] (C[4] of fig. 1) was a brother of Rihei I
(C ), Ken's grandfather.

― 28 ―

gata, south of Ueda city. It is a certificate of leave addressed to Makibuse village officials, and it mentions Shinzo's[*] name.
Ozaki Yukiya discovered several irregularities in the composition of this certificate, especially when compared with a similar
one dated a mere eight months later (1756/8):

Paper (issatsu no koto )

 Re : this Shinzo[*] . He is a person who was born in the village of Naka-Maruko. There is no question that [his family]
has been registered with this temple for generations. Now he is moving to your village as a successor [to a house
there]. Therefore, as the original guarantor of his religious affiliation, [I submit] the above document.

Horeki[*] 5/12 [1755/12], Koshigoe village

Zenbo-in[*] temple (seal)
[To :] Makibuse village officials.

Certificate of Leave for Rokuemon

Certificate of leave Re : Yohachi, 39 years old

The above mentioned person was born in this village and is certified as such. He is moving to your honorable village
as successor to Rihei. Please add him to your population register and put him under your jurisdiction. For future
reference, this sending document is as stated above.

Horeki[*] 6/8 [1756/8],

The [Nakasendo[*] ] Station of Mochizuki,
Omori[*] Hisazaemon, headman (seal)
[To :] Makibuse headman, Tsuchiya Chuemon[*][31]

The second document is addressed to Makibuse's headman (not to the village officials, as the first document is) by the
headman of origin, Mochizuki (and not by the person's family temple, as in the first document), where the person in question,
Yohachi (or Rokuemon), was registered. It records Rokuemon's age and the house (Rihei's) to which he will succeed in
Makibuse, both of which are missing in the former document, and requests that his name be added to the population register
in that capacity (also missing in the former document). The second document is more formal and complete.

Next to be noted is the name of Shinzo[*] , the same name as Ken's brother, who had disappeared thirteen years earlier,
leaving the house

[31] Both documents are cited only in Ozaki, "Kenjo oboegaki," 67, 68. The title of the second document was added on a
separate sheet by the headman for filing purposes.

― 29 ―

without a male head. Ken's two divorces and her seeming reluctance to adopt another husband had put the continuity of that
house in jeopardy. If someone by that name were now to be recorded in Ken's household, it would appear that the original
Shinzo[*] had returned. Everything points to a forgery in the making. Ultimately the attempt must have been abandoned,
because no Shinzo[*] appears on subsequent population registers in Makibuse, and the forged document might inadvertently
have been filed among the headman's papers. As noted, only one year later, in 1756/8, Ken married Yohachi/Rokuemon from
Mochizuki (her third husband).[32] The second document is the one Rokuemon brought with him when he moved to Makibuse
(the Rihei mentioned is none other than Ken's father).

It should be pointed out that this kind of document, a notification of a marriage union, does not mention the wife. Is it too
much to interpret this as indicating the absence of a legal status for women as contributing to a man's identity as his wife? The
two legal entities that are united here are a male and a house. Presumably, if the person being sent were a woman, she would
be spoken of as joining a male, as acquiring a new identity as someone's wife. Is the phrase "successor to a house," as Bourdieu
suggests, writing of Berber heirs in a similar situation, "an official euphemism allowing people to name the unnameable, that
is a man who could only be defined, in the house that receives him, as the husband of his wife"?[33]

Rokuemon's arrival brought another change in Ken's "civic" status: she was promoted from co-resident to head of a fully
established branch house (kakae). Moreover, her patron/lineage head was now Juzaemon[*] (E4 of fig. 1), not her uncle
Gendayu[*] , her legal guardian for twelve years, since her mother's death.

Shifting Power Within a Lindeage: 1670-1756

At the time, Ken's lineage was subdivided into two sublineages, headed by two patrons: Gendayu[*] and Juzaemon[*] . It had
not always been that

[32] Anne Walthall, upon reading this, made the interesting remark that while Shinzo[*] was presumably too destitute (0.19
koku) to take a wife, Ken succeeded in adopting a husband, which was supposed to be more difficult than marrying a woman.
Ken's obviously strong personality may perhaps partly explain this.

[33] Pierre Bourdieu, The Logic of Practice (Stanford: Stanford University Press, 1991), 179.

― 30 ―
Fig. 1.
Genzaemon's Lineage, 1629-1756
a village headman (plus lineage head and titled peasant)
b lineage or sublineage head
c household in dependent relationship with C line lineage
d household in dependent relationship with D line sublineage
e possession of one horse

Source : Adapted from Ozaki, "Kenjo ikken," table 4 (p. 57).

Notes : The total yield of the lineage property held by Genzaemon (A)
in 1629 is 20 koku; that held by 7 households in 1756 is 22 koku.
Note that Ken switched sublineages: she is not dependent on D2 but on E4 .

way. Originally Ken's paternal ancestors had headed the lineage. Since around 1670 there had been four households, headed
then by Rihei I, Ken's grandfather (C1 of fig. 1) Around 1700 there were five households (D1 —D5 ) headed by two patrons, Ken's
father, Rihei II (D1 ), being patron to two units, and Rihei's cousin (D5 ), to one. By the 1750s, however, power within the
lineage had shifted further away from Rihei's line, for now Rihei II's cousin's son Juzaemon[*] (E4 ) was patron to four
households (E1 , F, E3 , E5 ), while the other "patron," Gendayu[*] (D2 ), had Ken "registered under him," but not as a fully
established branch house, so that he was, properly speaking, no longer patron of a segment of the lineage.

― 31 ―

As mentioned above, the patron-client (or main house-branch house) lineage structure was a strictly internal village affair, not
regulated by shogunal law (although disputes concerning it, unresolvable in the village, could come before intendants). That it
was private or informal from the lord's point of view does not mean that lineage practice was not formalized and regulated
locally. The distinction between private and public does not pertain to the presence or absence of political regulation as such;
rather, it defines the modality of jurisdictional recognition between two social units one of which encompasses the other. Thus
the famous revenge of the forty-seven ronin[*] of Ako[*] , masterless samurai who avenged their lord, was for the bakufu not a
private but a public act and therefore under its jurisdiction. The bakufu left lineages alone, recognizing as legal entities only
neighborhood kumi as public institutions at the subvillage level. Nevertheless, as we shall see repeatedly, lineages were crucial
for the communal organization of villages.

Every peasant household in the Kita-Saku area had to be accounted for intramurally in a patron-client relationship; none
could function otherwise. Even new settlers from outside had first to secure a sponsoring patron in addition to the village
council's approval. Since political power was limited to titled peasants, who held access shares to irrigation, common grass or
mountain land, and so on, and lineage heads were always titled peasants, a client's access to power and its benefits was totally
dependent on his or her relationship to a patron and the latter's position and standing among the titled peasants. A client who
had serious complaints about his or her patron or lineage head could ask the village council to switch the household to a
different patron.[34] Such change, together with the reasons for it, was duly recorded in the village.

Ken was able to elevate her status from co-resident back to branch house, as it had been when her brother and mother were
still alive, by securing a male head and eventual successor to the house through marriage. This increased her autonomy vis-à-
vis her lineage patron. Ken's switch to Juzaemon[*] as her new patron was a double blow for

[34] For examples of newcomers seeking a patron, clients wanting m switch lineages (in the neighboring villages of Gorobe-
shinden and Yawata}, and other cases illustrating the position of clients, see Oishi[*] Shinzaburo[*] , Kinsei sonraku no kozo[*]to
ie seido (Ochanomizu shobo[*] , 1968), 125-37.

― 32 ―
Gendayu[*] , for now he lost control over Ken's household and his status as patron, for he had no other clients. The whole
lineage, except for himself, because he was a titled peasant, was now controlled by Juzaemon[*] , who had five branch houses
as his clients.

The reason for this change was recorded as "disharmony (fuwa )" between Ken and her uncle: "Horeki[*] 6 [1756]. Ken is taken
out as coresident from under Gendayu[*] . There being disharmony with her uncle [Gendayu[*] ], Ken is also removed as client
and will be client of her cousin [in reality her linealogical third cousin] Juzaemon." Ken's failed marriages and her three-year
absence from the village must have been related as cause or effect to this "disharmony," but relations must have been very
strained for her to want and to be granted severance from her uncle. We know the real reason from a petition she submitted
to the headman in the summer of 1757, a year after her marriage to Rokuemon.

An Open Secret Publicized After Seventeen Years: 1757

On 1757/7/22 Ken made the rounds of all her kumi/lineage members, the three kumi heads and the headman, in other words,
all the authority figures in Makibuse. She told them that she had a request to make. This was obviously a serious matter,
because that day the headman started keeping notes of meetings, recording also the absence of any officials and the reason for
it, as well as occasional quotations of who said what.[35] He obviously was creating a paper trail to cover himself, for Ken's
request concerned the violent death of her brother, Shinzo[*] . The headman recorded that he consulted with two villagers that
same evening. Two days later he got in touch with the Shimogata magistrate.

Thus prepared, before Ken had even made her request officially known, and most likely having agreed among themselves on a
strategy, the village officials took the first steps four days later, on 7/28. On that

[35]

― 33 ―

day the gathered officials (the headman and the four kumi heads) called all Ken's relatives for a hearing. When Rokuemon,
Ken's husband, was asked whether he agreed with Ken, he replied that "she had not involved him at all (watakushi e wa issai
kamawase mosazu[*] soro[*] )." While his answer may sound implausible, it is not impossible that he knew nothing about this
matter. He had moved to Makibuse to marry Ken only eleven months earlier. His answer was important, however, because it
was certainly irregular for a wife to engage in public action that took on the appearance of a lawsuit without her husband's
consent, since he held legal authority over the household.

The interrogators must have known that this was a wedge they could possibly drive into Ken's intentions. In Rokuemon's
written affidavit taken three weeks later, it is indeed clearly stated that "I did not know anything about it" and that "I am not of
the same mind [as Ken] on this (ichien doshin[*]tsukamatsurazu )."

His wife's pursuit of a lawsuit, with or without his knowledge or assent, must have humiliated Rokuemon, for it showed his
inability to keep order in his own household as was his duty. Moreover, lawsuits were to be avoided at all cost. Those who
resorted to them were written about in the village law as being on a par with other troublemakers and thus as people who
should be reported: "Art. 16. People who are not engaged in cultivation, trade, or any other occupation, or withdraw from
consultation with the village, or like quarrels and lawsuits, or do all kinds of bad things should not be hidden [but reported]"
(see appendix 3 ).

Rokuemon left Ken in 1759/2, a year and a half after Ken's lawsuit. He may have returned to his native Mochizuki, which was
very close. There is also a written pledge from around this time to stay away from sake; his problems must have led him to
drink. All his in-laws and the household heads of the Juzaemon[*] group tried in vain to bring him back. Shortly thereafter,
Ken added her voice and persuaded him to return.[36]

Obviously, life in Makibuse had become unbearable for Rokuemon, yet Ken prevailed for a while in keeping him home. Finally,
he could not take it any longer, and two years later, in 1761, he disappeared for good. Ken, however, must have had problems
too, for around the same time she started petitioning to leave the village. Indeed, as we shall see,

[36] The history of Rokuemon's earlier disappearances is discussed only in Ozaki, "Kenjo oboegaki," 72.

― 34 ―

her first request coincides with the moment she understood that her suit had failed.

As mentioned above, Ken directed her accusations against her relatives, among others, especially against her uncle and
former guardian, Gendayu[*] . This suit, which had been on her mind for some time, sheds light on Ken's second and third
marriages. Her second one came to a quick end because her move to another village under a different jurisdiction (Komoro
domain) would have made a suit nearly impossible. This marriage would have removed the only voice critical of what had
happened to Shinzo[*] .

Perhaps Ken sensed that this arranged marriage was an exile in disguise, which may explain her resistance under the flimsy
pretext of inauspicious geomantic matters. Yet her reinstatement as Gendayu's[*] coresident robbed her of the voice she
needed if she intended to file a suit, especially a suit against him. Once registered under him, Ken was legally as incompetent
as a child, and "hierarchy suits," in which children might sue fathers or retainers might sue their superiors, were outlawed.[37]

Ken's third marriage, to Rokuemon, however, solved the problem of legal voice, because it gave her back her status as a
member of an autonomous household. It is quite possible that through this marriage she intended not so much to put the
minds of her relatives to rest by securing the succession to her house (especially since later she would categorically refuse to
take the necessary steps for establishing an heir) as to dissolve her total dependency on Gendayu[*] , block any interference by
him, and gain the necessary authority to act publicly.[38] Rokuemon, perceived by others and perhaps himself as having been
used by a scheming woman, lost his bearings, fell apart, ran away, and finally fled for good.

Ken was present at the hearing of 2/28, but she was not asked about

[37] Dan Fenno Henderson, Conciliation and Japanese Law, Tokugawa and Modern , 2 vols. (Seattle: University of Washington
Press, 1965), 1:118: "Whether they were family or feudal, the rule was that suits brought by inferiors were not accepted."

[38] I would like to record Professor Yokota Fuyuhiko's cautious reservation as he communicated it to me in Kyoto in the
summer of 1994 regarding my interpretation of the legal dimension of Ken's third marriage. In his opinion, universal
registration in Tokugawa Japan created, in principle , the notion that everybody had the right to petition or sue, a "right" that
was, however, thwarted in many practical ways. In this view, strictly speaking, Ken would not have needed that marriage to
give her legal voice.

― 35 ―

her "request." Instead, she was told that an oral request would not do: if she had a petition to make, she should put it in
writing. The next day, a report of the investigation was sent to Shimogata, probably under the assumption that things had been
taken care of. By insisting on a written petition, the officials must have aimed at more than a postponement of the case,
calculating that without her husband's consent and assistance, and with the whole village against her, Ken would be unable to
draft a formal document.

Ken, however, was a resourceful woman. A week later, on 8/5, in the evening, she delivered a written petition to the village
authorities via the proper channel, that is, via Juzaemon[*] , her patron and kumi head. Early the next morning, Juzaemon[*]
rushed the document to the headman, who convened the village officials that same evening. Ten days later the document was
forwarded to Shimogata.

The first half reads as follows:

This is a request to hold the seventeenth memorial service for my elder brother Shinzo[*] . I have mountains and
mountains of grudges against Mr. Chuemon[*] [the headman]. He called my relatives and he called the whole village
together, deliberated with them, gave detailed instructions, and [thus] the innocent peasant of my house [my
brother] was killed. If he had been guilty [of some crime], my mother or myself would have been told the details by
my relatives. [But] because he was blameless they kept silent to us. He was killed [because the headman] had
wondered what [other misbehavior Shinzo[*] ] might engage in next, given the things he had been doing. Spite and
vexation about this murder have piled up mountain high, which I kept in my heart until this year. I want to pray for
the repose of his soul with a memorial offering so that your descendants may not suffer divine retribution. Anyhow,
following my judgment (nanibun ni mo kono kata no zonji yori ni ), I request that a memorial service be held.

On 1741/11/ 6 [a week after the murder], my relative Shinzaemon [E1 of fig. 1] called people together and I went
[Ken must have returned to Makibuse from Mimayose] and heard details about this. Shinzaemon said that an order
had come from the headman to have a memorial service held, since Shinzo[*] had been killed. I returned home and
told my mother the details [of what I had heard]. My sister also came from Yawata station [where she had married].
Anyhow, we thought of filing a suit for punishment with the shogunal authorities (gokogisama[*] ). But I was young
[twenty-three] and mother lost her mind because this had been too much for her, so that I abandoned the plan until
this year.

I have mountains of resentment against Mr Gendayu[*] [my uncle; D2 ] in this matter. [At one point, Shinzo[*] ] lost
the horse of my sister's husband Jingoemon from Yawata and came to borrow money [to pay back his

― 36 ―

brother-in-law], but my mother refused because they both [her son and son-in-law] were her children [she said], so
he went to Gendayu[*] , who lent him the money [on the condition] that [Shinzo[*] ] henceforward stop dealing with
horse people. All the relatives signed a note to that effect.

In addition, on 1741/16 his uncle had entrusted his own horse [to Shinzo[*] ], who had [mishandled] it, so that it
could nor be sold. For this reason [he] did not [dare to] come home. If he were to come home, my uncle, Mr.
Gendayu[*] , threatened to kill him. This being the case, Shinzo[*] pondered setting out for another province, but the
thought of leaving his old mother behind alone held him back. So he hung around in the area and stayed at a dry
field [owned by the family] at the foot of Ichiyama. There, everybody [got together] and a big crowd drove him out,
and so on 1741/10/31, in the evening, it was inevitable that he was killed (uchikorosaresoro[*]wa bitcho[*]gozaro[*] )
beyond the great bridge of Momozawa.

Shinzaemon [E1 ] delivered the first blow with a club, for which I heaped mountains of hatred on him. But now,
after seventeen years, I shall forgive him. [They] were told [to kill] this innocent man by the headman, but they
should [simply] have banished him. Dan'emon, who delivered blows, after seventeen years I can forgive. Uheiji [F]
is without blame, but Bunkyuro[*] [his brother], who died, he also delivered blows and is guilty. I do not know about
Chokyuro[*] [E3 ], Naoemon [peasant representative; 1.b of table 5], Iheie [kumi head; 5.F of table 5], Kichizaemon
[kumi head; 13.v of table 5] and the people of the village. But those who participated in this murder must know
about them.[39]
Thus, the reason for the disappearance of Shinzo's[*] name from the village population register in the spring of 1742 was that
he had been clubbed to death by his relatives and fellow villagers a few months earlier. Ken heard about it, although perhaps
not in great detail, at the time of the memorial service ordered by the headman, very likely to protect himself and all
accomplices against Shinzo's[*] avenging spirit. Together with her sister and mother, Ken had pondered the possibility of
pressing charges with the shogunal authorities, a plan she abandoned, but only for seventeen years, as it turned out. Over
time, Ken probably picked up bits and pieces of the story, which was a public secret. Realizing Gendayu's[*] role in her
brother's murder must have made life under him unbearable, a situation so perfectly euphemized by calling it fuwa ,
disharmony.

Ken's most serious accusations, however, were directed at the head-

[39] Ozaki, "Kenjo ikken," 50-51. The second half of this document is discussed below.

― 37 ―

man for allegedly ordering the murder of a man who, although admittedly a troublemaker, certainly did not deserve the fate
that befell him. Ken deduced Shinzo's[*] "innocence" from the silence of those who should have informed her and her mother
of the reason for his death. Shinzo's[*] troubles had centered around horse trading and gambling, both activities that were
disapproved. Article 51 of the village law specifically warns against horse thieves and go-betweens in horse trading (see
appendix 3). This angered most of all his uncle Gendayu[*] , who seems to have used Shinzo[*] as a middleman to sell his own
horse, which Shinzo[*] somehow must have spoiled.[40] Understandably angry, Gendayu[*] threatened to kill Shinzo[*] , which
implicates him heavily in this "death almost foretold," as Gabriel Garcia Márquez might put it. According to Ken, there had
been malign intention to kill her brother: an order by the headman and a threat by Gendayu[*] . She blames her relatives for
having followed a death order rather than simply expelling Shinzo[*] , and she names people who might know more about
other participants in the murder. Although Ken stated her accusation within the context of a professed desire to hold a
memorial service and a willingness to pardon some of those involved in the crime (about which more will be said later), she
was seeking at least some justice—from the accused—for a crime committed seventeen years earlier.

If the village officials had ordered the murder as a punishment, they were in double or even triple jeopardy: for contracting a
murder, for committing it, and for going beyond the bounds of their penal jurisdiction by issuing a death verdict, if, in fact,
that is what they did. This last point is important, for as we know, in principle, "private justice" was forbidden by shogunal law.

The historian Mizumoto Kunihiko lists seventeenth-century laws (which did not change during the next century) at various
jurisdictional

[40] In 1737 there were 16 horses in Makibuse, which then counted 12 titled peasants out of a total of 52 households, with a
total population of 248. In 1750 the figures were slightly different: 12 horses, 11 titled peasants out of a total of 48 households,
and a population of 242 (see table 1); thus there was almost one horse per titled peasant. The price of a horse is discussed as
follows by Tanaka Kyugu[*] (1662-1729) in his Minkan seiyo[*] of 1721: "Peasants need horses, but their price has gone up in
recent years. In the past, one could get a horse for one or two ryo[*] . Now, however, there are no horses for that price ... now
one cannot even get them for ten ryo[*] " (Nihon keizai sosho[*] , ed. Takimoto Sei'ichi, vol. 1 [Nihon keizai sosho[*] kankokai[*] ,
1914], 261). In article 13 of appendix 4, a price of "from 3 bu to one ryo[*] " is referred to in 1665.

― 38 ―

levels prohibiting punishments such as wounding or killing by villagers: shogunal laws issued for intendants to be applied in
their districts, bannerman laws for their fiefs, and bakufu-ordered village laws.[41] The heaviest penalty allowed at the village
level was banishment, although, as we shall see in chapter 4, scholars have recently discovered a small number of exceptions.
In addition, Mizumoto cites instances of enforcement of such laws: in Okayama domain a headman was beheaded for having
"secretly" (i.e., "privately" on his own initiative, without reporting it to the higher authorities), and with the help of others,
killed a thief (1676); village officials in a shogunal fief were fined in 1778 for having "secretly" banished a thief to another
province. Makibuse's village law explicitly prohibited the killing of evildoers (article 4) or the plotting by peasants of evil
things (article 17; see appendix 3).

The officials in Makibuse saw their options for getting rid of this case disappearing fast once an official accusation of murder
had been made. If behind the attempt to have Ken married out to Kannonji-shinden lay the intention to get rid of the one
dissident voice against them, they had failed. Doing away with the evidence by eliminating the litigant was obviously much too
extreme and risky. Besides, the recommended punishment for such a crime in shogunal courts was the death penalty, followed
by gibbetting of the severed head.[42] Persuading Ken to withdraw her suit did not work either. Hence, they had to transmit the
suit to the Shimogata office because murder was a matter beyond village jurisdiction and the investigation and punishment
had to be left to the higher authorities.

What the village officials resorted to, as already mentioned, was seeking secret advice from the Shimogata magistrate on how
to handle this

[41] Mizumoto Kunihiko, "Kogi[*] no saiban to shudan[*] no okite," Saiban to kihan , Nihon no shakaishi, 5 (Iwanami shoten,
1987), 285-86 (hereafter NNS 5); the information on village punishment in the following paragraphs comes from p. 286.

[42] Kujikata osadamegaki (1742), art. 71, par. shi, tsuika of 1744 (Tokugawa kinreiko[*]koshu[*] , ed. Ishii Ryosuke[*] , 4 vols.
[Sobunsha[*] , 1959-61], 3:419 [hereafter TKKk]). For an English translation of book z of the Kujikata osadamegaki , see John C.
Hall, "Japanese Feudal Laws HI: Tokugawa Legislation, Part IV, The Edict in 100 Sections," Transactions of the Asiatic Society of
Japan 41, pt. 5 (1913): 683-804. For a German translation of the whole work, see Otto Rudorff, "Tokugawa-Gesetz-Sammlung,"
Mittheilungen der Deutschen Gesellschaft für Natur- und Völkerkunde Ostasiens in Tokio , supplement to vol. 5 (April 1889).
 ― 39 ―

case before officially transmitting the suit to him. That much we know from Makibuse's headman's diary, but he did not record
the nature of this advice. We can only follow the developments surmising what it might have been.

The shogunal legislation that would be applied in an eventual trial was as follows. The law provided for severe punishments,
graded according to the nature of the murder, the perpetrators, their motives, their degree of participation, and the status of
the victim. For example, accidentally killing a nephew was punished by banishment; but if the murder had been premeditated
for gain, it was punished by death (shizai ). Contracting a murder was punished by death (geshunin ); executing a contract
murder, by banishment. Among accomplices, the one who struck the first blow received the death penalty (geshunin), those
who assisted were banished, and those who did not physically participate in the crime were punished by medium deportation
(10 miles).[43]

There existed some half-dozen different death penalties, of which shizai and geshunin were the lowest two. Death penalties
were always beheadings, to which various additional degrees of nastiness could be added, such as exposure of the criminal,
gibbetting, and so on. Shizai always entailed the availability of the corpse for sword practice (tame-shimono ) by samurai and
confiscation of the criminal's property (kessho ). Geshunin, as a rule reserved only for commoners, was based on the
retribution principle calling for the taking of a life if a murder has been committed; the corpse and property were left intact.
[44]

Investigating an Allegation

On 8/I6, a week after Ken's document was forwarded to Shimogata, Ken's relatives were called together by the village officials
to ascertain their intentions. As was to be expected, they disagreed with Ken's statement, which again left her alone to face the
whole village and its authorities. She had no allies.

The next day Ken was summoned by the headman and informed of her relatives' intention not to back her up, to which Ken
apparently retorted, "If my request is not settled after submitting it to the [Shi-

[43] Ibid.

[44] Hiramatsu Yoshiro[*] , Edo no tsumi to batsu (Heibonsha, 1988), 94.

― 40 ―

mogata] public office (Goyakushosama negai age ), I shall go all the way to the Edo office (Edohyo[*] )."[45] The headman's ploy
to keep the lid on this case foundered on Ken's forceful counterthreat to bring the case to court in Edo. Always upping the ante
in her confrontations with authority, Ken invariably found ways to respond to official threats with threats of her own. Thus,
the Shimogata office was officially informed of Ken's suit.

On 8/21 life came to a halt in Makibuse. Early that morning all those involved in Ken's suit, practically the whole village (the
headman, the kumi leaders, and all the members of the kumi), were summoned to Shimogata for interrogation. Yet the
question addressed to Ken was not about her allegations. It concerned Shinzo's[*] relationship to his household and Ken's
knowledge thereof: did Ken know that her brother had been disinherited? She testified that she had never been informed
about it but added that if Shinzo[*] had indeed been disinherited, "what I have said will have difficulty standing up (watakushi
mosu[*]bun aitachinikui tatematsuri zonji soro[*] )." In other words, her indictment might be null and void.

Ken had married and left home four years before her brother's disappearance. Therefore, it is possible that she did not know
that Shinzo[*] had indeed been disinherited only a few months before he was killed. Shinzaemon, however, produced then and
there the disinheritance deed. Seeing that the killers would escape prosecution, Ken spat out, "Now I hate you even more." We
shall return later to this question of disinheritance (kyuri[*] ). Let us first follow the investigation to its conclusion.

A second question pressed upon Ken was how she had drawn up the lawsuit. Her ability to do this, after all, had thwarted the
village officials' plan to ever let it come this far. Ken admitted that she had paid someone to draft the document, but she was
fierce in her refusal to reveal the name of the person who had assisted her. "I won't tell you," she fired back, "even if I lose my
head for it."

A war was being waged here over knowledge of documents. The disinheritance deed nullified Ken's suit, and she was
ultimately accused of lying when she maintained her ignorance of its existence. The officials, on the other hand, tried to
prevent her from producing a docu-

[45] Ozaki, "Kenjo oboegaki," 76.

― 41 ―

ment, the suit, in order to avoid an investigation, and they insisted (in vain) on knowing who had composed it. Ken's ignorance
was the weapon used by the officials against Ken (she should have known about the disinheritance), that she was quick to use
against them when she refused to reveal the name of the scribe.

Of course, we do not know who helped Ken. It is unlikely that it was someone from the village. If she had to find help
elsewhere, however, she would not have had to search far. Her family temple in Iribuse, to which she intended to commend
her property, was only one kilometer away. She could have gone to Yawata, where her sister was living and where her mother
and grandmother had come from. By this time in the Tokugawa period there were literate people in most villages, some even
with scholarly pretensions. For example, in Katakura village, two and a half kilometers beyond the hills west of Makibuse, one
of the village officials had a collection of books and manuscripts, even a copy of a work by Yamagata Shunan[*] , one of Ogyu[*]
Sorai's disciples.[46]

There were thus ways to circumvent the village elite's monopoly on writing and on writing in proper form (which meant
power in a regime that governed through documents). Ken's "petition," nevertheless, was not in proper form, for it gave too
much vent to her resentment and hatred and was devoid of the obligatory deferential formulae, besides being extremely
difficult to understand. It looks like a draft for a suit and allows a rare glimpse at a commoner's feelings, and rarer still, at a
woman's confrontation with authority.

At this juncture, a senior retainer (karo[*] ) from Shimogata joined the case and examined the evidence. Ken gave up her
lawsuit, owing to the legal point made of Shinzo's[*] disinheritance, and then requested to leave the village. She must have
realized that life there would be hell, and those with the power to decide whether to release her or to consign her to the village
for life must have realized it as well. She was given no final answer; instead, she was "entrusted" to her husband Rokuemon, a
euphemism for house confinement. This closed the first round of the investigation. Now that Ken's allegations had been
thoroughly discredited, the matter of Shinzo's[*] violent death could be taken up—rather safely, one should add.

[46] Nagano-ken, ed., Nagano-kenshi: Tsushi[*]5 (Kinsei 2) (Nagano-shi: Nagano-kenshi kankokai[*] , 1988), 251 (hereafter NAK-T
5).

― 42 ―

A whole month went by before, on 9/21, Gendayu[*] testified and admired to the murder as follows:

Shinzo[*] was doing evil things [gambling] roaming about the mountain where he lived. Therefore, the lineage and
the whole village banished him. They drove him out beyond the bridge at Momozawa, at which point Shinzo[*]
turned around, and because he raised his hand against me, I had to hit him with the club I held in my hand. In the
commotion that followed, it was hard to tell who was doing what (dare to mosu[*]koto mo naku bo[*]nite
uchimoshisoro[*] ). Blows fell, resulting in his death.[47]

Gendayu's[*] legal ground for defense was built into his testimony, perhaps on the Shimogata officials' private advice. Shinzo[*]
had started the fight, and his death was therefore not premeditated; it was a rebuttal of Ken's attempt to impute intention to
the accused. Moreover, he had been thrown out of the village for good reason (gambling), and it was his resistance to this
punishment that had led to the tragedy. On Ken's own admission, the headman had wondered "what would be next, given that
he does all these things."

That Shinzo[*] "raised his hand first" was most likely crucial to Gendayu's[*] defense—insofar as a defense was still needed.
This had two legal consequences. First, it turned Shinzo's[*] violent death into something other than murder: an unfortunate
accident for which the victim was to blame (he had resisted expulsion from the village). And second, the alibi introduced an
element that today would be labeled "justified and legal self-defense."

One article of the shogunal penal code read: "When, due to the unreasonable behavior of one party, one has no choice but to
cut him or her down, if it is certain that the relatives and village headman of the victim acknowledge that he or she was
ordinarily unlawful in behavior, offer no excuses for the person in question, and beg that the killer be pardoned, the
punishment is to be medium deportation [i.e., a lessening of the sentence]."[48] Moreover, although in principle arrests of
criminals by official or private parties had to take place without the infliction of harm, if it happened that the criminals got
killed under certain circumstances, the killers would not be held responsible.[49]

[47] Ozaki, "Kenjo ikken," 52.

[48] Kujikata osadamegaki , art. 72 (TKKk 4:1).

[49] Hiramatsu Yoshiro[*] , Kinsei keiji soshoho[*]no kenkyu[*] (Sobunsha[*] , 1960), 683.

― 43 ―

Gendayu[*] could not have known of this particular article of the penal code, because the code had been written that same
year and was secret, meant only as a set of guidelines for judicial officers of the shogunate.[50] However, the code was also
based on previous practice, and an awareness of mitigating circumstances must have existed, at least among bakufu officials.

Perhaps advised by the officials at Shimogata, Gendayu[*] may have aimed his testimony toward legitimate self-defense against
an opponent of dubious reputation for whom nobody would speak up. There certainly was enough time to structure
Gendayu's[*] testimony that way, since a month had elapsed between the first and second hearings.

A Dubious Case of Disinheritance

At this point the question of disinheritance comes in. Kyuri[*] was a distinctive Tokugawa form of disinheritance, different
from an older type (kando[*] ) although often confused with it even in Tokugawa times.[51] It drastically changed one's "civic"
status and is best understood in conjunction with other laws.

Japanese disinheritance practice has changed over time. Its main forms were gizetsu , kando[*] , and kyuri[*] . In antiquity,
gizetsu ("cutting the obligation") meant expulsion from home and loss of all inheritance rights. During the Kamakura period
(1185-1333) it was gradually replaced by kando[*] ("right measure after examination"), applied for unfilial behavior to escape
vicarious responsibility for crimes one is afraid one's children might commit, especially when their whereabouts are not
known, or simply when parents decide to exercise this right. The descendant had to leave the house and was freed from all
filial obligations and rights, including protection, but was not barred from official posts; and the decision could be revoked by
the parents. In the Tokugawa period, official permission was required for kando[*] , but now

[50] Dan Fenno Henderson, "Introduction to the Kujikata Osadamegaki (1742)," in Ho[*]to keibatsu no rekishiteki kosatsu[*] , ed.
Hiramatsu Yoshiro hakushi tsuito ronbunshu[*] henshu[*] iinkai (Nagoya: Nagoya Daigaku shuppankai, 1987), 504-8.

[51] For the information that follows, see Kokushi daijiten , ed. Kokushi daijiten henshu[*] iinkai, 15 vols. (Yoshikawa
kobunkan, 1979-94), 4:253 ("kyuri[*] ") and 3:884 ("kando[*] ") (hereafter KDJ). See also F. Jouon[*] des Longrais, L'Est et l'Ouest:
Institutions du Japon et de l'Occident comparées (six études de sociologie juridique ) (Tokyo: Maison Franco-japonaise, 1958),
387-90.

― 44 ―

there was also kyuri[*] ("separation for a long duration" or "old separation"). Kyuri[*] had the same prophylactic function
against the consequences of possible crimes by absconded offspring. Official ratification was by the parents, kumi, village
officials, and the intendant who ultimately permitted the fugitive to be taken from the population rosters, which automatically
dissolved marriage and blocked any possibility of holding office.

Laws and institutions often have ambiguous functions. They steer practice along norms and paths to produce good and just
behavior, which, however, may only be good and just in some respects but not in others. Hence, they also produce a need for
additional laws and institutions to correct these "dysfunctions." Kyuri[*] can be understood in such a way, as a response to the
undesirable effects of two other laws. One was the stipulation, inherent in the Tokugawa feudal configuration of power, that
crimes be judged in the jurisdiction of the criminal's registration. The simplest of these were domains or fiefs for peasants and
samurai. Matters were more complicated for uncommon commoners like priests, registered beggars, outcastes, and the
registered blind or for uncommon cases that involved two jurisdictions.[52] The other legal practice with a direct bearing on
kyuri[*] was the custom of vicarious guilt or communal responsibility for misbehavior (enza ): the relatives, the kumi, or
sometimes the village could be punished for a crime committed by any of its members. Hence they could suffer for a crime
committed by someone who had absconded and over whom they had thus lost control.

Legally separating an absconder from the group responsible for him or her through official disinheritance prevented such
injustice. Kyuri[*] was thus both a punishment in absentia for someone who had illegally left the village and a protection
against possible prosecution of his or her relatives and kumi members for crimes he or she might commit.[53] It should be
noted that in the second half of the Tokugawa period com-

[52] See Henderson, conciliation , vol. 1, chap. 4, esp. pp. 86-92.

[53] Oishi[*] Hisakata (1725-94) discusses kyuri[*] and kando[*] in these terms. A former village group headman from Kyushu[*]
who traveled widely in Omi[*] , Shinano, and the Kanto[*] , he was hired m 1783 as coordinator of rural administration in the
Takasaki domain (Gunma prefecture, east of Kita-Saku), where he composed for his lord a very detailed vade mecum for rural
administrators, the Jikata hanreiroku (see Oishi[*] Hisakata, Jikata hanreiroku , ed. Oishi[*] Shinzaburo[*] , 2 vols. [Kondo[*]
shuppansha, 1969], 2:123-54).

― 45 ―

munal responsibility was activated only for the most extreme crimes: killing one's lord or one's parent. Hence,
"disinheritance" was rather dormant as protection for the legally accountable group.[54] It served mostly as a means for the
person legally responsible for an absconder to evade social criticism within the village. Also, kyuri[*] could not be enacted
against a status superior: one could not "disinherit" one's absconding parent. Such disinheritance was a publicly certified act.
It had to be approved by the local headman, and valid reasons had to be stated. Since someone thus "disinherited" was erased
from the village registers, this measure needed to be sanctioned by the higher authorities. It changed the official status of the
person in question. A disinherited person was thus legally a nonperson and, because he or she was not attached to a legal
group, perhaps the village equivalent of a ronin[*] , or masterless samurai. Such a person was thus cut off from the village in
all possible ways, having neither obligations to the village nor privileges, including the privilege of protection. He was an
outlaw.

A remedy to the possible ill effects of another law, kyuri[*] itself was insufficiently corrected by further laws against its own
possible negative effects—except for the important general provision, central but implicit to Ken's argument, that life could not
be taken. Article 4 of the Makibuse village law made this clear: "People who have killed someone, or strange people who hang
out in shrines, forests, and mountains: the villagers, together with people from neighboring villages, should set out, arrest
them, tie them up, and hand them over. If it is difficult to catch them there on the spot, then you have to pursue them however
far and catch them where they settle. But no matter what kind of person they are, you cannot kill them" (see appendix 3).
Kyuri's[*] only other softening aspect was that it was revokable if the person affected by it mended his or her ways.[55] In
Shinzo's[*] case we see how kyuri[*] could be manipulated to prevent a violent death from being categorized as murder.

The next questions to be addressed are the reasons for Shinzo's[*] punishment of disinheritance and banishment. Why the
disinheritance, which was an action that ultimately only Shinzo's[*] mother could have sanctioned? And why did the villagers
and lineage members mete out

[54] KDJ 4:253.

[55] For an example of documents reversing prior verdicts of disinheritance and banishment, see Dan Fenno Henderson,
Village "Contracts" in Tokugawa Japan (Seattle: University of Washington Press, 1975), 190-91, 194.
 ― 46 ―

what turned out to be, in Ken's eyes, deadly punishment for Shinzo[*] —without trial, verdict, or, as reported by Ken, stated
reasons? Ultimately, Ken was arguing degrees of appropriate punishment and maintained that they had no right to kill
Shinzo[*] .

Two documents exist requesting the disinheritance of Shinzo[*] : one, signed by his mother and relatives, is addressed to the
village officials and dated 1741/7; the other, bearing the seals of the village officials, is the one the village officials forwarded to
the Shimogata office and is dated 1741/7/21, or barely three months before the murder. Since these documents were so crucial
in the murder case, it is of course not unthinkable that they were fabricated post factum for that purpose—after the murder
and even after Ken presented her suit.

If the disinheritance deed was a forgery, one might wonder how Shinzo's[*] mother's seal was secured. The modern
presumption that seals were always kept at home, however, does not apply universally to Tokugawa Japan. Ishii Ryosuke[*]
writes that it was not uncommon for a village headman to keep all the seals of all the households; some village laws even
warned against the possible abuses of entrusting one's seal to others, even relatives.[56]

Shinzo's[*] mother's request reads as follows:

Written Petition Submitted in Fearful Deference

My son Shinzo[*] , who made a living as a horse trader, was fond of heavy drinking and misbehaved in various
places. Moreover, he did not engage in agriculture. In particular, last fall he left fallow the small dry field we have,
leaving me almost starving [the following spring]. Again and again the lineage, united in its views, requested him [to
change his ways], but it was of no use whatsoever and the problems remained.
This year on 4/13, he disappeared with Gendayu's[*] horse without leaving a trace. I asked the lineage to look for
him everywhere, but he has not been found to this day, and this matter has [still] not been settled. It is hard to gauge
what other evil things he might be up to next. I therefore request the shogunal authorities to take him, starting this
year, from the population register and disinherit him.

1741/7 The village of Makibuse

Shinzo's[*] mother (seal)[57]

This petition served as the basis for the official document forwarded to Shimogata, which is in three parts: Shinzo's[*] mother's
(rewritten) peri-

[56] Ishii Ryosuke[*] , Inban no rekishi (Akashi shoten, 1919), 182, 184.

[57] This document to be found only in Ozaki, "Kenjo oboegaki," 77-78.

― 47 ―

tion, an endorsement by her four kumi members/relatives testifying to the truth of her statement, and an addendum by the
village officials including further explanations. The first two parts read as follows:

Written Petition

My son Shinzo[*] has not worked in recent years and not supported me, which has made life very hard for me.
Moreover, on 4/13 he disappeared, which is why I asked the lineage to look for him everywhere, but to this day he
has not been found, and even if we look further, I am alone with nobody to support me. Therefore, I petition to
remove [my son] from the population register.

The mother of Makibuse's Shinzo[*]

1741/7/21

The abovementioned Shinzo[*] , as stated in his mother's petition, made an unsubstantial living (fujittai yue yowatari
), which caused great hardship to his mother. Then, on 4/13, he disappeared with his uncle Gendayu's[*] horse. We
immediately started a search everywhere without results. On 5/21 we sought an official warrant [for an official
search], and we ordered the neighborhood kumi to look for him. We have searched until today but not found him,
and nobody has reported him, so that nobody from either the lineage or the kumi has been able to lay a hand on
him.
The above is without fallacy. Without any discrepancy do the relatives and kumi endorse the mother's petition.
Deferentially looking upward to the shogun (gokogisama[*] ), we humbly request the removal [of Shinzo's[*] name]
from the population register.

1741/7/21 The village of Makibuse

petitioner Gendayu[*] (seal)
petitioner Chokyuro[*] (seal)
same Shin'emon (seal)
same Masaemon (seal)
[To :] The village headman
The elders [i.e., kumi leaders][58]
According to these documents, Shinzo[*] had ceased to work the fields and instead had gotten involved in horse trading, with
the result that he was unable to support his mother. He drank too much and caused trouble. Moreover, he disappeared with
his uncle's horse. (Ken also refers to this matter in her petition, although rather obliquely). Private searches failed to locate
him. After one month a complaint was filed; the kumi ordered an official search, with no results.

[58] This document is available only in Ozaki, "Kenjo ikken," 53.

― 48 ―

Fleeing became legal absconding only after a certified effort had been made to apprehend the fugitive. A certain amount of
time (officially, six periods of thirty days) was fixed during which active searches were repeatedly ordered by the village or
supravillage authorities.[59] Hence the mention of the failed search in the disinheritance document (although fewer than six
months seem to have elapsed between Shinzo's[*] absconding and the disinheritance): disinheritance, as mentioned earlier,
was a legally certified "punishment" for absconding, while absconding was the legally certified status of a fugitive.

The official petition, signed by Shinzo's[*] mother and endorsed by her relatives, was forwarded to the Shimogata office by the
village officials, who appended a note with more details to complete the work of character assassination. They mention his
asocial conduct with reverberations in the community: drunkenness, general misbehavior, noncompliance with the "will of
the lineage," mentioned by Shinzo's[*] mother. Moreover, in the context of an investigation of theft conducted six years earlier
(in 1735) by another shogunal office (one in Hiraga, some fifteen kilometers east, in the Kita-Saku plain), Shinzo[*] had been
questioned because he was a horse trader (bakuro[*] ). Although Shinzo[*] seems not to have been involved in the 1735 theft
(otherwise it would certainly have been used more fully against him), and although this matter was six years old, one surmises
that it was made part of the record to give more weight to the argument for disinheritance by establishing, through innuendo,
a long record of misbehavior and bad companions. Following this investigation, the lineage had made him promise to
abandon horse trading and return to tilling the fields, a promise he had not kept. This, presumably, was "the will of the
lineage."

The previous fall, the officials also related, his absence of over twenty days had made him miss the wheat sowing time, which
had endangered his mother's food supply. Moreover, on 4/13 he had led his uncle's horse away to sell it and had disappeared
wearing his uncle's clothes. This last detail testifies, perhaps, to Shinzo's[*] poverty: members of branch houses often relied on
their patrons for material gifts or loans of daily necessity, so these clothes may or may not have been "given" to him by his
uncle. His relatives looked for him quietly for one month, but then the kumi was ordered to undertake a formal search, which
it did for an-

[59] Mizumoto, "Kogi[*] no saiban," 307; see also Oishi[*] Hisakata, Jikata hanreiroku , 2:113, 114, 119.

― 49 ―

other month, without results. Fearing what Shinzo[*] might be up to next, his mother had requested the removal of his name
from the register.

The case against Shinzo[*] is made again and again in these documents, at ever higher levels of authority: by his mother, his
lineage, the village officials. This reiteration no doubt had a reinforcing effect, solidifying the case against him. Clearly, things
were presented so as to put the blame for the removal of Shinzo's[*] name from the registers and the concomitant
disinheritance on Shinzo[*] himself. But one senses also the pressure the lineage put on Shinzo's[*] mother to take this drastic
step. At this point, however, she could not have foreseen the tragic outcome only three months away. And perhaps neither did
anyone else in the village, although there was the alleged threat by Gendayu[*] . Moreover, since kyuri[*] could be revoked,
perhaps Shinzo's[*] mother hoped for such a reversal when she consented to the disinheritance—if the document is genuine.

The Decline of a Prominent Peasant Household

In her disinheritance request, Shinzo's[*] mother mentioned that the family's landholdings were minuscule. Indeed, they may
have been too small to support her and her son, so that Shinzo[*] had sought additional income in horse trading and perhaps
also in gambling. How small was the property?

The family holdings Shinzo[*] inherited at the death of his father in 1719, when he was nineteen, amounted to no more than
1.9 koku, or 1.6 tan , the equivalent of less than 0.4 acres. (His father, Rihei 11, had not managed the property too well: he had
lost one third of it, since he had started out with 2.9 koku.) One koku of rice was the amount needed to feed an adult for one
year, and one able-bodied adult could normally cultivate three tan.[60] Shinzo[*] was obviously a very poor peas-

[60] Nakane Chie, Kinship and Economic Organization in Rural Japan , London School of Economics Monographs on Social
Anthropology, 32 (London: Athlone Press, 1967), 59. See also Oishi[*] Shinzaburo[*] , Kyoho[*] kaikaku no keizai seisaku: Dai
ichibu, Kyoho[*] kaikaku no noson[*] seisaku (Zohoban[*] : Ochanomizu shobo[*] , 1978), 20. A similar estimate is given in NAK-T
5:143, where it is stated that a couple could cultivate seven to eight tan (the equivalent of about ten koku). Thomas C. Smith
quotes a less productive ratio: four or five adults for one cho[*] of arable (one cho[*] equals ten tan), eight or nine for two, and
twelve for three (Agrarian Origins , 6). A holding's kokudaka is only an indirect reflection of its size, for it refers to the
computed putative yield of all products converted into rice equivalents, on the basis of which the tribute was calculated. This
tribute, in turn, was paid neither in rice nor m any other natural products. After 1726, in all bakufu land in Shinano, the
tribute amount, calculated as a percentage of the kokudaka, was converted into a cash amount at the going conversion rate
(NAK-T 5:86).

― 50 ―
Table 2.
Disposal of Shinzo's[*] Holdings in Koku, 1727-1740

First-grade Homestead

Holdings Paddies Dry Fields Dry Fields Total

Holdings

At start of 1.03 0.79 0.14 1.97a

1727

Disposed in

1727 0.43 — —

1729 — 0.14 —

1735 — 0.6 —

1736 0.26 — —

1737 0.17 — —

1739 0.15 — —

1740 0.02 — —

At end of 0 0.05 0.14 0.19b

1740

A 1.97 koku = 0.414 ACRES

B 0.19 koku = 0.042 ACRES

SOURCE : Adapted from Ozaki, "Kenjo ikken," table 9 (p. 63).

NOTE : The fields were sold to two distant cousins, E1 and E2 ,

belonging to the other sublineage, headed by D5 . The 0.14 koku of dry
fields constitutes newly developed fields; the remaining 0.05 koku is
also made up of newly developed dry fields. All paddles, plus the 0.60
koku of dry fields, are main fields or honden (fields registered on the
original seventeenth-century land surveys).

ant with insufficient land to keep him fully occupied, but in this he was not alone in the village. With 1.9 koku, his was among
the twenty-seven households, out of a total of forty-nine, with holdings under two koku.

In 1727, Shinzo[*] started to dispose of the little land he had (see table 2), until by 1740 all the honden ("original paddies": first-
grade paddies and first-, second-, and third-grade dry fields) and three-fourths of the newly developed dry fields were sold to
two distant cousins (E1 and E2 of fig. 1, who, it should be noted, in the 1720s and 1730s more than doubled their property, to 8.9
koku and 5.7 koku, respectively). Shinzo[*]

― 51 ―
kept only a tiny lower-grade dry field in the mountains and the superior-grade dry fields of residence land, a total of 0.19 koku,
or 0.04 acres. In other words, he wound up selling 90 percent of the land he had inherited, making him no better off than a
landless peasant, as mentioned earlier. In her petition of 1757, Ken stated that her brother Shinzo[*] "went to live in the
mountain on his plot of land." This plot of land, at the northern outskirts of the village, beyond the Nakasendo[*] , was the only
one he had left besides the residence land. He was killed a year after he sold his last paddy.

In table z one can see that in 1727 Shinzo[*] must have suddenly been in great need of cash. This was when he sold almost half
of his prime paddy to his cousin. Two years later he disposed of all the newly developed (low-grade) dry fields, except for the
tiny plot in the mountains, which he kept until the end. The next spurt of sales occurred between 1735 and 1740: first the main
(honden) dry fields all at one time in 1735, then the remaining precious paddy piece by piece.

It is not clear when Shinzo[*] began horse trading, but it is unlikely that he did so before he started to dispose of the land, in
1727. Until then he must have been working the fields he had inherited seven years earlier. In 1727 he fell into a spiraling
circle of horse trading, gambling, drunkenness, and land sales. At least this is the picture one gets from the disinheritance
document the village officials submitted to the Shimogata office. Yet these hardships seem to have come in spurts: the first one
in 1727-29 (loss: 0.57 koku), the second and seemingly the greatest in 1735-37 (loss: 1.03), the last in 1739-40 (loss: 0.17). Ozaki
surmizes that Shinzo's[*] "reprehensible" behavior was most likely not the cause but the result of hardships that may have had
something to do with the marriages of his two sisters. His sister Ine got married in this first period, his sister Ken in the second
period.[61]

These data Ozaki culled from records unrelated to the case. However, the 1763 incident (with the population register)
triggered an investigation, possibly ordered by the Shimogata office. The report reveals a wider background to Ken's lineage
and introduces us to some social and political dimensions of the history of Makibuse.

According to this report, Ken's ancestors had been peasants in Maki-

[61] Smith (Nakahara , 97) writes that for peasants "dowry was not normally a significant form of property." His source is late
Tokugawa. Yet, the "chest of personal belongings" the bride usually took may have seriously taxed a poor peasant's resources.

― 52 ―

buse for several generations, starting in the early 1600s with Genzaemon [A of fig. 1], headman at the time of the land survey
in 1629. His holdings were listed at 20 koku, among the four largest land holdings, each between 20 and 24 koku. For four
generations, down to Ken and Shinzo's[*] father, Rihei II, the family had been very prominent in the village and no doubt
jealous of its prestige, which may help explain the resentment provoked by Shinzo's[*] behavior. What happened to the family
between Genzaemon's time and Shinzo's[*] ? The story mirrors the fate of many families in the area. One has only to
remember that down to 1662 only three out of twelve families had holdings of under ten koku, while after 1721 that number
was forty-six out of forty-nine (and thirty-four of the forty-six had three koku or less [see table 4]).

In the second generation, in the late 1640s, the successor (B) lost the village headship because of an unspecified village conflict,
but his holdings were not affected negatively; in fact, they grew even larger (according to the resurvey of 1669-70, he had
twenty-three koku).[62] The third-generation successor (C1 , or Rihei I) also held the village headship for one year, in 1689. In
the transition from the second to the third generation, in 1684, the property was split among four sons. The successor to the
lineage headship (C1 , then the oldest son) received about one-third of the land, and the other three (C2 , C3 , and C4 ) each
received one-fifth. However, partitioning had already started more than a decade earlier. According to the population register,
in 1671 the family had been split into four separate households: Rihei I (C1 , then the second oldest son), Shinzaemon (B, the
father), Heishiro[*] (C4 , the youngest son), and Choemon[*] (C5 , the oldest son, who disappears from the registers thereafter).
The latter three, including the father , were all clients of Rihei I, their patron and also the only titled peasant in the lineage.[63]

[62] Ozaki has discovered no further details concerning the village conflict that caused the headship to be lost to another
lineage. The research by Mizumoto Kunihiko and Saito[*] Yoshiyuki, introduced in chapter 2, suggests that most of the village
disturbances of the 1640s were directed against the arbitrary use the village headman had made of the new power the lords
had vested in them (Mizumoto Kunihiko, Kinsei no mura shakai to kokka [Tokyo[*] Daigaku shuppankai, 1987], esp. chaps. 1
and 3 [hereafter Mura shakai ]; idem, "Murakata sodo[*] ," Chusei[*] no minshu[*] undo[*] ,Chuseishi[*] koza[*] , 7 [Gakuseisha,
1985], 289-307; Saito[*] Yoshiyuki, "Kinsei shoki no nomin[*] toso[*] to muraukesei: nengukanjo[*] sodo[*] o sozai to shire,"
Rekishi hyoron[*] ,no. 475 [1989]: 42-60).

[63] Ozaki ("Kenjo ikken," 57) charts the 1671 generation as one comprising three brothers (Rihei, 37; Shinzaemon, 64; and
Heijiro[*] , 34) and a fourth person (Choemon[*] , 39), whose relationship m the other three is unspecified (respectively, C , B, C ,
and C in fig. 1). Three of these four were relatively close in age, but the fourth, Shinzaemon, was twenty-five years older than
the oldest of the trio, which makes it possible that he was their father—assuming also that Choemon[*] is a sibling. According
to Ozaki's table, this Shinzaemon was the son of another Shinzaemon. My reasoning is based on the assumption that there is
only one Shinzaemon, son of Genzaemon (A) and father of Rihei I and his brother(s).

― 53 ―

It should be noted that the father moved out (a pattern of retirement sometimes called inkyo bunke , literally, "retirement
branch house"), and the headship was passed on to a house that branched off but was legally the head of the lineage and
inherited also the status of titled peasant. Primogeniture was by no means the rule in this region or, indeed, in most of rural
Japan in the seventeenth century.

More and Smaller Peasants

In 1734 a bakufu official sketched a picture of the optimal ratio of land to people in an ideal village. Such a model community
would have zoo koku (half paddies and half dry fields) and a population of 120 divided into 24 households, 60 males and 60
females, as well as 12 craftsmen or merchants and 6 horses.[64] In terms of arable, Makibuse met this standard: in 1750 it had
205 koku (see table 1). Its population, however, was double what it should have been: 48 households with a total of 242
members and 21 horses.

This model was put forward at the time of the first of three large-scale reforms, the Kyoho[*] Reforms of the 1730s, which
attempted to arrest developments that had weakened the bakufu's economic base. The earlier edicts prohibiting the sale of
land (1643) and the division of land into holdings of less than to koku (in the case of headmen, 20 koku) (1673) had failed to
strengthen the rulers' grip on agricultural surplus. A significant increase in the number of large landholdings syphoned off
part of that surplus as land rent paid to landlords by tenant farmers without land, or with too little land to support their
households. The history of these two prohibitions—"thou shalt not accumulate" and "thou shalt not divide"—and their
implementation is a complex one.[65] Put simply, it is clear that the laws were circumvented in various ways.

[64] Nakane, Kinship , 42.

[65] See Oishi[*] , Kyoho[*] kaikaku .

― 54 ―

The limitation on land partitioning was dealt with locally by officially (tatemae ) limiting the number of titled peasants and
recording branch houses as working part of the main family's land (buntsuke ). This practice can also be found in Makibuse,
for instance in Genzaemon's lineage. The 1677 cadaster shows the land registered under Rihei I's name and indicates the
amount tilled separately (buntsuke) by each household of the lineage.

Oishi[*] Hisataka, the eighteenth-century local administrator of Takasaki domain in a neighboring province, explained this
"partition" as marking the division of inheritance but not separately established branch houses (bunke ).[66] If branch houses
were established, their heads would be titled peasants and would pay tribute and corvée directly, he wrote. In Makibuse,
however, those households marked as buntsuke in the 1677 land cadaster had already been registered six years earlier in the
1671 population register as branch houses (kakae), that is to say, as separate branch houses under a patron. It is not clear
whether these houses paid their tribute directly, as Oishi[*] Hisataka suggested, or through their patron. What is certain,
however, is that they were branch houses without being tided peasants. Over time, the number of titled peasants in Makibuse
decreased somewhat, while the number of households providing tribute and corvée increased dramatically. Thus the bakufu's
definition of a titled peasant (as seen through Oishi's[*] eyes) was at variance with the village's. In the village of Makibuse, the
title became a privilege and a mark of political, not economic, power. Like shares or stock that could be sold, titles were
transmissible and negotiable. In order to purchase one, an individual needed "economic capital," but as figure z indicates, one
could hold on to it as symbolic (and political) capital even if one's holding was less than 2 koku. Note that household L of figure
z had only 0.4 koku, whereas the village headman had 33 koku, and they were both titled peasants.

As in the case of neighboring Kodaira, which is analyzed in detail in chapter 3, partitioning leveled economic differences
between titled peasants or patrons and (nontitled) branch houses. Such partitioning of the lineage's holdings by establishing
branch houses was a common practice in Makibuse. Other examples are the Iemon and Kichisaburo[*] lineages (see table 3
and fig. 2).

[66] Oishi[*] Hisakata, Jikata hanreiroku , 2:110. On Oishi[*] , see above, n. 53.

― 55 ―

Table 3.
Holdings of Iemon's Lineage in Koku, 1719-1736

Lineage Households

Lineage Patron and Co- A B C D E G K

residents

1719 A + d/e/f/g/h (5 sons) 13 4 5

1723 H + d/e/f/g (4 sons) + k/m (2 14 4 4

fudai)

1729 H + f/g (2 sons)+k (1 fudai) 7 4 4 3 3

1731 H + f(1 son)+ k (1 fudai) 4 4 3 3 3 3

1736 F 3 4 3 3 3 3 0

SOURCE : Adapted from Ozaki, "Kenjo ikken," table 3 (P. 55).

NOTE : Over a span of thirteen years, between 1723 and 1736, the lineage, headed
by A, added four households, D-K; by 1719, B and C already existed as autonomous
households. The position of economic prominence of the main house over the
branch houses disappeared as its size dropped from 14 koku to 3 and its holdings
were equalized with those of all the other units of the lineage. Notice that H, the
successor to A, was the youngest of five sons (d-h) and that the holding was divided
equally among them.

Of the two fudai of 1723, the female (m ) had disappeared from the registers by
1729; the remaining male fudai (k ) was eventually established as branch house K
without land (mizunomibyakusho[*] ) in 1736. It was quite usual for fudai to marry
and have families of their own after being set up as a separate branch. Thus, the
practice of branching off of fudai must have led to a population increase.

The same leveling trend was manifested throughout the village (see table 4). Between 1647 and 1662 the distribution of
holdings among titled peasants remained stable: four units had holdings of 20-24 koku (one of them being the A and B
generations of Genzaemon's lineage), five units had 10-14 koku, two had 5-9 koku, and one had 3 koku. The one household
with less than 1 koku in I647 had disappeared from the registers three years later, leaving a total of twelve titled peasant
households by 1662.

By 1677, however, the situation had changed drastically, although official registration maintained a fiction of continuity: four
households were shown to have more than 20 koku, four to have 10-14 koku, seven to have between 4 and 9 koku, and one
with 1 koku, bringing the total to sixteen, an (official) increase of four. Yet, in reality, seven out of this total of sixteen holdings
(each officially registered under one house) were divided among two to five households, usually evenly, resulting in a massive
(unofficial) increase in the number of households owning 1-10 koku of land. The distribution was as follows: one house of 29
koku (that of Chuemon[*] , the village headman); one of 11 koku and one

― 56 ―

Fig. 2.
Kichisaburo's[*] Lineage, 1700-1758
a titled peasant

Source : Adapted from Ozaki, "Kenjo ikken," table 7 (p. 61).

Notes : Nonce the kokudaka of 0.4 of title peasant L in 1758.

It will be remembered that the headman, also a titled peasant,
had over 30 koku. Smith (Agrarian Origins , 39 n. d) states that
to "ignore the minimum [below which partition was not practical]
was to invite ruin." Yet this is precisely what seems to have been
happening here and else where in the village. Hayami Akira wrote
that there were no cases of holdings under 5 koku being divided
("The Myth of Primogeniture and Impartible Inheritance m
Tokugawa Japan," Journal of Family History 8, no. 1 [1983]: 23).

of 10 koku; thirteen in the 5-9 koku range; and fifteen in the 1-4 range, bringing the actual total to thirty-one households, an
increase of nineteen over the 1662 total.
In subsequent decades the partitioning of lineage lands continued. By 1721 the total number of households had risen to forty-
nine (thirteen titled peasants, thirty-two branch houses, and four landless peasants), with two holdings above 30 koku; one of
x4 and one of 10 koku; seven in the 5-9 koku range; eighteen in the 1-4 koku range; and sixteen with less than 1 koku. Four
peasants were registered as without land of their own.

― 57 ―

By 1700, when Ken's grandfather Rihei I divided his property (increased by 1 koku to a total of 9), some branches of
Genzaemon's lineage were struggling to survive on very diminished holdings. Ken's father, Rihei II (D1 ), received only 2.9
koku, and her uncle Gendayu[*] (D2 ), 6.3 koku. Her father's cousin (D3 ), with only 1.2 koku, soon disappeared from the land
register. The other branch of the lineage was doing better: another cousin of her father's (D5 , who held 6.3 koku) formed a
sublineage with yet another cousin (D4 , holding 5 koku) and the latter's two successors (E2 and E3 ).

In 1703 Ken's father, Rihei II, transferred half of his already small holdings to his brother Gendayu[*] , retaining only 1.5 koku
for himself. He must have been experiencing financial difficulties, an indication that the decline of this line did not start with
Ken's brother, Shinzo[*] . Peasants in this situation had few options. Obviously, they lacked the means to develop substantial
new fields. Moreover, the terrain may not have allowed further development. Over more than four generations, between 1627
and 1756, the lineage's holdings increased by a mere 2 koku.

One solution was part-tenancy, but it is not clear whether this was an option in Makibuse. There were only a few holdings
large enough to use labor beyond the nuclear family (two holders of more than 30 koku, one of 14 koku, and one of 10 koku
[see table 4]), labor that may have been provided by hereditary or indentured servants who do not seem to disappear, even
though the population grew considerably and at a much faster rate than the village kokudaka (see table 1). On the other hand,
branch house or client status no doubt included some duties for the lineage head or patron that may have been remunerated;
indeed, some clients may have been quasi-tenants.

Another solution for surplus members of the household was to seek regular employment in towns or cities or short-term
employment during the slack winter season.[67] This is what Rihei II did. In 1712 he left his wife and daughter Ine (Ken was not
yet born) and took twelve-year-old

[67] Seasonal employment m Edo during the winter months was customary in some areas in Nagano, even for villages far
north, such as Yomase in Takai district. Documents show Yomase villagers petitioning to go to Edo from the ninth to the third
month in 1701-2. In 1757 in Otsukoto (Suwa district), which had 130 households, the number of those who officially petitioned
to spend the winter months in Edo was 33 (NAK-T 5:366). In Kodaira (see n. 24), exactly half of the peasants (9 out of 20 titled
peasants and 18 out of 36 kakae) were themselves listed as being in service outside the village (all titled peasants but one, and
no kakae, went to Edo); this, in addition to II of their children (see Komonjo kenkyukai[*] , "Genroku-ki," 67).

― 58 ―

Table 4.
Family Holdings in Koku, 1647-1721

1677

Koku 1647 1650 1662 Officially In 1721

reality

33 Z (1)

32

31 Y (1)

30

29 Z Z

28

27 X

26
25 W

[3] [1]

24 Z Z Z

23 x x x C1

22

21 W W W

20 A B B

[4] [4] [4] [1]

19

18

17

16

15

[0] [0] [0] [0]

14 1 1 1 V 1 (1)

13 1 1 T

12 3 2 1V S

11 S 1 1

10 S S B 1 (1)

[5] [5] [5] [4] [2] [2]

9 1 1x 2 (1)

8 1 1 1 1M 1 x C1 1 (1)

7 1 1vv 1 (1)
6 1 1 1 1 1m 2 (2)

5 xww 1

[2] [2] [2] [5] [13] [7]

(Table continued on next page)

― 59 ―

Table 4
(continued)

1677

Koku 1647 1650 1662 Officially In reality 172I

4 2 2www 1

ttt

s s s C4

3 1 1 1 7
(2)

2 m 3

1 1 1x 7
(1)

0+ 1 16
(1)

[2] [1] [1] [3] [15] [34]

0 4

Total 13 12 12 16 16+15=31 49
(13)

SOURCE : Adapted from Ozaki "Kenjo ikken," table 5 (p. 58).

NOTE : Letters A-C refer to households of the Genzaemon lineage (fig. 1),
M-Z to households of other lineages; small letters refer to households that
are branch houses of capital-letter lineages that were officially listed in
1677 under the patron families. Numbers indicate households whose
relationship over time could not be traced between 1647 and 1662. Totals
of 1647 through "1677 (Officially)" all refer to titled peasants; those of
"1677 (In reality)" comprise the same 16 titled peasants and their 15
branch houses after partition; the totals for 1721 are for the number of
households and, in parentheses, the number of titled peasants.
The number of households almost quadrupled between 1647 and 1721
(from 13 to 49), although the aggregate yield remained more or less the
same: approximately 166 koku in 1647 and 186 koku in 1721. (Numbers
have been rounded to the nearest even number.) Cf. table 1.

The increase in small and minuscule holdings was dramatic. Landless

peasants appeared in 1721 (4), when another 23 households with less
than z koku were very close to becoming landless. In 1647 only 4 out of 13
households held less than to koku; in 1721 that number was 45, and only
4 held more than 10 koku.

Shinzo[*] with him to Edo (some 180 km, or twenty-four way stations on the Nakasendo[*] , away), where they worked together
for two years. After his father returned home, Shinzo[*] stayed in Edo another three years; then he worked a year in the tiny
village of Gan (at the southern end of Makibuse's valley, some 7 km from home).

These efforts proved somewhat successful, for in 1716 Rihei II was able to "buy back" 0.5 koku from his brother, increasing his
holdings to 1.9 koku. This is the amount of land Shinzo[*] inherited three years later. However, when Rihei II sold part of his
land to his younger brother Gendayu[*] in 1703, he also relinquished, or very likely sold, to

― 60 ―

him his inherited lineage headship and his status as titled peasant.[68] Also, when Rihei II left for Edo, he had to step down
from the village office of kumi leader, an office he would be unable to recuperate upon his return because he lacked the
prerequisite status of titled peasant. Rihei II, whose great-grandfather and grandfather had been village headmen and whose
father, Rihei I, had held the office for one year in 1689, had become an impoverished, untitled peasant.

A brief excursus is called for here regarding the link between village office and peasant status in Makibuse. As table 5
indicates, in the first half of the eighteenth century the number of titled peasants in the village decreased by one-third, from
sixteen to eleven. These five households became clients of other titled peasants (household numbers 2, 4, 10, 14, and 15). On
the other hand, some clients became titled peasants, trading status with their patrons. This is what happened between Rihei II
(D1 ) and his younger brother (D2 ), and they obviously were not the only ones: see the three switches in the Kichisaburo[*]
lineage recorded in 1717, 1735, and 1764 (line II of table 5); see also figure 2. This table and figure also show younger brothers
superseding their older brothers (e.g., in lines 7, 8, II, and 12 of table 5), although one must here distinguish between shifting
the status to a different household headed by a younger brother (D1 and D2 in line 7) and keeping the status within the same
household with the younger brother succeeding his older brother as the head of the household (K and M in line 11). It is also
clear that these shifts do not occur in the more well-to-do lineages (lines 1, 3, and 5), which also are able to hold on to village
offices more durably. It should be noted that after 1735 the office of elder was combined with that of kumi leader (which
meant the lineage head, since kumi here were constituted along lineage lines).

Lineage Tensions

Let us recapitulate Ken's story in light of this information. Some time between his marriage in 1694 and the birth of Shinzo[*]
in 1700, Rihei II set up a separate household, leaving his younger brother Gendayu[*] with

[68] In Kasuga, a village southwest of Makibuse, across the hills, the village law stipulated the price a client had to pay to his
patron in order to be set free and become a titled peasant as 20 ryo[*] (Nagano-ken, ed., Nagano-kenshi: Tsushi[*] 4 [Kinsei 1]
[Nagano-shi: Nagano-kenshi kankokai[*] , 1987], 466 [hereafter NAK-T 4]). For a study of land pawning and redemption, see
Shirakawabe Tatsuo, "Kinsei shichichi ukemodoshi kanko[*] m hyakusho[*] takashoji," Rekishigaku kenkyu[*] , no. 552 (1986):
17-32.

― 61 ―

Table 5.
Titled Peasants and Village Offices, 1700-1764
SOURCE : Adapted from Ozaki, "Kenjo ikken," table 8 (p. 62).

NOTE : Column one is a guide to lineages diagrammed in other tables

(for example, Table 4.Y = table 4, lineage Y). Letters representing
household heads that appear in other tables are capitalized. All other
peasants are represented by lowercase letters in alphabetical order.

= village headman; ( ) = otonabyakusho[*] , which according to NAK-T

4:57 means hyakusho[*]sodai[*] , or peasant representative, usually a
small peasant, but not in Makibuse or m Irebuse before 1760 as
discussed m chapter 2; ____ = village elder, after 1735 kumi head; * =
earlier client status with preceding titled peasant;
= younger brother of preceding peasant;
= becomes client of next titled peasant (note the decreasing number of
titled peasants). Subscript figures indicate koku.

their mother. In 1700 Rihei was thirty-seven and Gendayu[*] was eighteen. The division of the property greatly favored
Gendayu[*] , perhaps because he took care of the mother: Rihei received 2.9 koku, while Gendayu[*] received 6.3, or more than
twice Rihei's portion. As we have seen, the succession rules in the village were not hard and fast, least of all that of
primogeniture. Nevertheless, because his brother got the lion's share, Rihei had to find ways to prevent his economic situation
from declining further. It was perhaps as a means to reassert symboli-

― 62 ―

cally the importance of his line over his brother's that Rihei adopted his father's name, thereby resurrecting his forebear and
affirming his claim to the latter's power and position inherited as titled peasant, patron, and kumi head. Yet economic reality
prevailed: Rihei wound up selling his titles and office to Gendayu[*] , which further increased the gap between the two
families, this time in terms of political power. This was part of the legacy he left Shinzo[*] when he died at age fifty-six.

Shinzo's[*] stay in Edo during his adolescent years must have changed his outlook on the world. He returned to a life as a poor
peasant without hope of improving his lot by working the fields. To the great dismay of his lineage members, who counted
village headmen among their ancestors, he sought his fortune in horse trading. Worse, he courted the company of gamblers
and suspected thieves.

From Shinzo's[*] perspective, however, it must have appeared that part of the reason for his misery lay with his uncle
Gendayu[*] , who had wound up with most of the family's property and with the family's titles. It should be noted that none of
the land Shinzo[*] sold in his seven separate sales went to Gendayu[*] : it all went to the competing and rising sublineage (see
table 2). The constant criticism of his ways by the lineage and their attempts to make him change his style of life must have
further alienated Shinzo[*] .

But the advice his relatives gave him carried more import than that of an ordinary uncle or aunt: as members of his kumi,
they were responsible for his behavior in the eyes of the community and the authorities. Their opinions, therefore, constituted
official admonitions. The head of his kumi was the representative of the village headman, behind whom stood the authority of
the bannerman and ultimately the shogun. This being the case, it was difficult for Shinzo's[*] mother to defend her son, and
her agreement to request the disinheritance, if it was her doing, undoubtedly signified her public disapproval of Shinzo's[*]
course of action. She had to align herself with the village authorities. Yet, even thus isolated, Shinzo[*] could not bring himself
to leave the village, for which he paid with his life.
As Ken pointed out in her lawsuit, the legal grounds for killing Shinzo[*] were open to question. The official effect of
disinheritance was to relinquish communal responsibility, but it was revocable. Although expulsion from the village could
have been a justifiable punishment for habitual gambling, it should not have meant irreversible exile. Yet it was pushed in that
direction, ordered by the headman, given full lin-

― 63 ―

Table 6.
Chronology of Ken's Life

1694 Rihei II [D1 of fig. 1] (30), lineage head, titled peasant, son of village head,
member of village council, marries young woman from Yawata (16)

1700 Rihei (36) and Gendayu[*] [D2 ] (20) set up separate households; Shinzo[*]
born

1712 Ine born; Rihei and Shinzo[*] (12) leave for Edo; Rihei stays two years;
Shinzo[*] stays five years and then works in Gan village one year

1719 Rihei dies; Ken born

1727 Ine (15) marries and moves to Yawata

1737 Ken (18) marries and moves to Mimayose

1741 Shinzo[*] (42), Gendayu's[*] client

4/13 Shinzo[*] disappears with Gendayu's[*] Horse
7/21 Shinzo[*] disinherited by his mother (63)
10/31 Shinzo[*] dies

1742 Ken returns and becomes Gendayu's[*] client (kakae)

1745 Ken's mother dies; Ken under Gendayu's[*] tutelage (chonai[*] ) Winter:
Ken (26) marries man from Kannonji-shinden

1746 Ken returns to Makibuse; lives alone

1749 Ken in service at Hirabara

1750-52 Ken in service in Komoro, then returns to Makibuse and lives alone

1756 Ken (37) marries Rokuemon (39), from Mochizuki; leaves Gendayu's[*]
tutelage

1757 7/22 Ken's oral petition for a memorial service for her brother; headman
starts keeping diary (until 11/18)
8/5 Ken's written petition
8/16 Affidavit by Rokuemon

1759 2/24 Rokuemon disappears

3/25 Written promise by Rokuemon to abstain from sake
4/15 Rokuemon returns
1761 9/- Rokuemon (44), disappears for good

1762 3/- Ken (40), requests to leave Makibuse

3/6 Succession question for Ken
3/7 Ken requests to leave Rokuemon on population register

1763 2/6 Ken requests to take Rokuemon off the population register
2/8 Ken refuses to sign the population register

1795 1/8 Ken dies

NOTE : Ages are given in parentheses.

― 64 ―

eage support. That Shinzo[*] resisted expulsion is understandable, but his resistance, if we believe Gendayu's[*] version,
resulted in his death.

Ken bitterly resented the outcome. Yet, when she challenged its legality, the disinheritance document saved the accused by
giving Shinzo's[*] violent death justification legitimized by authority (the Shimogata magistrate).

Protection Against an Avenging Spirit

In her efforts to bring this affair into the open and question its legality, Ken obviously could not count on any cooperation from
within the village. She said she paid someone to draft the suit, although that may not have been true. Payment, however, made
the relationship between herself and the scribe a business one, exculpating the scribe from any charge of sympathy for her
case. But bringing it into the open must also have been a catharsis for Ken, as the second half of her petition/suit shows:

As for his memorial service, following my judgment (nanibun ni mo kono kata no zonji yori ni ), it should be held so
that none of your descendants might suffer retribution. The service itself should consist of daily recitations at
Zenkoji[*] in perpetuity (kono yo aran kagiri ) and, also at Zenkoji[*] , a forty-nine-day ritual; daily recitations in
perpetuity at Koyasan[*] and a lantern; at the family temple of Baikei-in, recitation of one thousand sutras in
perpetuity.[69] If things turn out that way, I shall sponsor all this. I shall go to Zenkoji[*] [to arrange] for the service. I
shall bring money from you and I shall contribute my own part (sesshagi mo kono setsu dodo[*]itasu beku soro[*] ).
Koyasan[*] is a place where women are not allowed, hence my husband Rokuemon will represent me. He will also,
in a similar fashion, collect money, including travel expenses.
Although I am talking like this, I do not want to request even one sen. It would not be proper to do [as I say] if each
and every one is not convinced [that it is the right thing to do]. You should all get together and follow your judgment
(zonji yori ni nasaru beku ). I also am following my own. If you are convinced [of the rightness of this case], then
Chuemon[*] [the village headman], Gendayu[*] , Shinzaemon [E1 of fig. 1], Dan'emon,

[69] The forty-nine-day "ritual (tsukare )" probably refers to the period after death during which a number of purifying
exorcisms are performed. The lantern may. refer to a standard stone lantern or to one used in memorial services. Zenkoji[*]
was famous as a Pure Land to which all souls were said to go immediately following death. On Zenkoji[*] , see Donald F.
McCallum, Zenkoji[*]and Its Icon: A Study in Medieval Japanese Religious Art (Princeton: Princeton University Press, 1994).

― 65 ―

Uheiji [F] should write a deed [of apology for the murder]. If that is what they think is right, then I shall write a
document [to forgive them]. I also want a deed [of apology] from Chokyuro[*] [E3 ] and all the other officials, and
then I in turn shall write a document [to forgive them]. I [shall] feel no [further] resentment toward my relatives. I
have no resentment toward Juzaemon[*] [E4 ], Heihachi [E5 ], and, among the officials, toward Tadasuke [kumi head;
9.1 of table 5; the other two kumi heads were mentioned as guilty in the first half of the petition]. As for the people
of the. village who did not show up here today, I do not need their monetary contribution for the service. I have
nothing else to say. As stated above.

1757/8 Ken (seal)

[To: ] The village headman
The officials[70]

Thus Ken had devised a number of rather extraordinary memorial events—perhaps commensurate in her eyes to the crime to
be expiated—in order that the descendants of her brother's murderers might not suffer divine retribution. She hints at
supernatural retribution and vents her resentment but is also determined and resourceful to overcome a difficulty she, as a
woman, might encounter in executing her plan. Ken would take charge of it all and collect monetary contributions from
everyone, but she wanted these only if people were convinced that the murder had been a mistake and were ready to
apologize for it in writing. She appealed to everyone's conscience. She wanted a verdict from a tribunal less than she wanted
an apology from everyone. That, however, she would not get.
For all those involved in Shinzo's[*] murder, the deed of disinheritance had wrapped up the case. On 9/21, accused of false
testimony concerning her lack of knowledge of the disinheritance, Ken was "entrusted" to her kumi; that is, she was restricted
in her comings and goings and put under surveillance. A month later, on 10/26, she was summoned before the Shimogata
authorities again and told that her suit was an "unspeakably impudent petition (iwarezaru daitan naru negai )" that on
principle ought to be investigated. Instead, she was told, in the condescending voice of paternalistic authority, "you are
forgiven because you are a woman [but] will be ordered house confinement because of your offense."

When the village officials received this order, they used it to restore

[70] This second half of Ken's petition is to be found only in Ozaki, "Kenjo oboegaki," 93-94.

― 66 ―

"peace" in the village. Through the priest of her family temple, they promised Ken that they would not enforce the house arrest
order if she reconciled herself with her relatives, that is, if she stopped harassing them—a second put-down, turning the victim
into the cause of the trouble. This new proposal made Ken furious, and at first she resisted it forcefully. Finally, perhaps
recognizing that it would be futile to further resist her relatives and the whole village, backed as they were by bakufu
authority, Ken stated her intention to comply with their demands.

The failure to obtain an apology or win her lawsuit had not broken Ken's spirit. The trouble arising from her refusal to affix
her seal to the population register in 1763 proves this. That year, it so happened, was the twenty-third anniversary (according
to Japanese reckoning) of Shinzo's[*] death, something that must not have escaped Ken. (Individual dead are memorialized
with more than the annual rites at special anniversaries—the seventh, eleventh, thirteenth, and twenty-third.) This time Ken
did not accuse her relatives or the village of wrongdoing, but pleaded hardship in having to work the fields, pay tribute, and
perform corvée. She asked for understanding for her plight and proposed a solution, namely, that the fields she was supposed
to work be removed from the tax register and commended to her temple and that she leave the village. In fact, Ken's solution
was another expression of her total rejection of her relatives and her village, no doubt prompted by her unwillingness to
accept their denial of any wrongdoing in her brother's murder.

In this she resisted, of course, the solution proposed by the village authorities, namely, to provide for an heir. This threw them
into confusion once again, but they could not let a woman who had caused them so much trouble have her own way. The two
sides were at an impasse again. The village wanted a successor to Ken's house and therefore did not want to let her go, so that
the land would remain in the village. She did not want to stay, nor did she want her land to go to anyone in the village. There
seems to have been more behind village officials' insistence that she establish an heir (part of a new agreement Ken signed, in
which she also promised to continue as a peasant and keep the peace) than, as Ozaki surmises, a desire to comply with
overlord law not to let peasant houses die out. Ken was not even a titled peasant; her holding was tiny, and many families
could have used more land.

What came of this promise? Twenty years later, according to Maki-

― 67 ―

buse's population register of 1782, Ken was still living alone. She also appears on the tax register of 1792 as a tribute payer. She
still had not complied with the wishes of the authorities. Ken's tombstone, next to her mother's and Shinzo's[*] , perhaps
erected by her lineage in fear of their avenging spirits, records the date of her death as 1795/1/8. She lived alone in Makibuse
until she was seventy-six. One can only imagine what her life must have been like during those last thirty years. The best way
to describe it, leaving room for the imagination, is the way her life under Gendayu[*] was described: fuwa , without harmony.

Conclusion

Through the story of Ken we have made the acquaintance, in unusually vivid detail, of a peasant woman and some salient
events in her life. Ken's life and that of her family were tied to a past of increased economic hardship, which many others
shared. It also unfolded in a wide geographic setting beyond the village, reaching all the way to Edo. There her father and
brother hoped to accumulate savings to alleviate the family's plight, and there Ken thought she might eventually find justice.
Ken was a stubborn and strong woman who followed her own conscience (see the several instances of kono kata no zonji yori
ni in her petition). Her sense of justice was reflected in a trail of documents recording a series of confrontations with authority
at various levels: her husband, her relatives as a lineage and as a neighborhood kumi, the village officials, the headman, and
shogunal representatives.

Ken's sense of justice clashed with the authorities' concern for order. With a firm grasp of the system of governance, geared
toward the maintenance of order, Ken succeeded more than once in strategically subverting it to pursue what she judged to be
just. She manipulated two of her marriages this way. Although illiterate, she produced an important document, in format a
petition for a memorial service, in reality a demand for justice in the form of an apology from the village authorities, in effect
a lawsuit against them for murder. She knew how to gain a public voice to lodge this suit and how to embarrass the headman
and her kumi. She threatened to wield the symbolic weapons provided by religion and by the courts to find retribution and
justice from two faraway centers that she must have imagined as detached from the nasty world she experienced: Zenkoji[*] , a
nonsectarian Buddhist pilgrim center

― 68 ―
Plate 2.
Ken's Tomb. Left to right : the tombstones of Ken, her mother, and
Shinzo[*] . Photograph by author.

and also a transit point for the souls of the recently deceased; and Edo, a political center where powers and authorities
converged and to some degree lost their local character.

Ultimately, this match between a woman seeking justice and a system seeking "harmony" was uneven. Authority was not to
bend to law, and law was not to bow to reason, according to a Tokugawa saying. In practice, therefore, authorities did not issue
public apologies, because only subjects could be in the wrong. Ken, who thought justice might be a concern of the shogunal
authorities, found that they functioned with the same logic as her village superiors. Once she publicly forced the issue of an
apology, the two levels of authority, intra- and extravillage, formed a common front. Then Ken's resistance turned passive.
Within the village that became her prison, she kept control over her property and her life until the end of her days by not
adopting an heir or a husband as her superiors would have had her do.

Power in a village, as we see it used by both sides in Ken's story, is social power that is steeped in public, official power. One's
relatives and lineage provide an immediate social framework that has at the same

― 69 ―

time a public character, because it overlaps with the official neighborhood system and its internal hierarchy of main and
branch houses, which takes on an official dimension by the distinction between titled and nontitled peasant households. In
addition, the manipulation of power at the village level has wider dimensions because of the strategic possibilities offered by
the intersection of intra- and extravillage authorities. Here we can mention the attempts by the village authorities to prevent
the petition/suit from leaving the village; the threats by Ken either not to appear before the extramural authorities (in the
registration incident) or to bring her suit to them (in the murder case); the collusion between the intramural and extramural
authorities with regard to the investigation of the murder; and the pressure on the part of the village authorities to extort
promises by granting suspension of a punishment (house confinement) decided beyond the village.

Because Ken was a woman, her interactions with village authorities and the expectations she confronted must have been
those of patriarchal power. Yet, in apparent control of her three marriages, firm and outspoken in her confrontations with her
kumi members and village officials, and countering the headman's threats with some of her own, Ken flaunted those
expectations. When the officials finally dug in their heels, denying Ken her request to leave the village and transfer her land to
a temple, they put her in her place as a woman , which was to secure males (a husband or an heir) on the land to produce
tribute. Ken, however, was equally firm in sticking to her earlier warning: "There is no need to come and speak about this to
me again."

Ken must have been scolded often by her relatives and superiors. One clear instance was when the Shimogata authorities told
Ken that she was not given the full punishment she deserved for her "unspeakably impudent" suit because she was a woman.
Here condescendence yields, by way of scolding, the double profit of a feigned compassion and understanding and a real
punishment. Scolding is perhaps the first means to activate explicitly when threatened, the (usually unformulated) power of
doxa or common sense, which as it is used here is that of gender-asymmetrical propriety.

In all moves made by parties in the power field encompassing both village and supravillage authorities, written
documentation played a crucial role. Conflicts and disputes revolved around, and were ultimately resolved by, the creation of
legally recognized documents; for example, in the case of the headman's diary, the record of steps taken by him

― 70 ―

could eventually be used in court. The following instances of the crucial role of documentation come to mind: the ability to
write petitions in proper form; the manipulation of population records; the importance of written and certified disinheritance
documents that were forwarded beyond the village; a petition for a memorial service that could not be suppressed because it
contained a charge of murder; the many occasions where written promises were exacted (to not drink sake, to establish an
heir, to move in with a new husband in another village, to live in harmony, to work the fields, etc.). Documents were
indispensable for regulating conduct within the village. Village life was highly legalistic and formalized, a far cry from a
Gemeinschaft governed by informal understandings, where the spoken word is binding.

Moreover, intervillage linkages of various kinds played an important role in defining networks that could be exploited for a
number of purposes. Not only was Makibuse exogamous to a very high degree but some of its inhabitants spent long periods
of time working outside the village. Its position on a major highway made Edo appear available and within reach to the
villagers and must have increased the flow of information that could be put to use in various ways. Villages were less isolated
than often imagined.

The question of village autonomy has been debated back and forth by Japanese historians, most of whom choose either to
defend it or to deny that it had any substance. The Tokugawa political landscape, being neither homogeneous (which would
make the village insubstantial) nor divided into two heteronomous spheres (which would disregard the grave imbalance of
power between villages and lords), is perhaps best visualized as two fields of power that created at their intersection a space
for manipulation from either side.

― 71 ―

2
Class Politics
Privileges associated with past status are acknowledged, but times have changed.
Bannerman to village headman, 1650s

Class and status are introduced in two separate chapters in this book. The distinction between the two, however, hard to
justify even analytically, should certainly not be taken as an objective separation, as the epitaph on status in this chapter on
class suggests. It is merely a question of emphasis, of foregrounding one against the other. The simultaneous economic and
political decline of Ken's line in Makibuse illustrates how inextricably related the two could be in Tokugawa village life. Yet
status obviously is not a mere reflection of class: some of Ken's neighbors, although impoverished, were able to hold on to the
status of titled peasant and the political power that came with the title. That micro narrative was firmly anchored locally,
venturing only occasionally beyond the confines of the village. In this chapter the analysis moves inward toward the village
from the wider class and power configurations that were shaping society mainly in the first half of the Tokugawa period.

A cardinal feature of Tokugawa society was the concentration of the new warrior rulers in urban centers, separated from the
land and the people they had conquered in the last decades of the sixteenth century. The rulers' self-enforced spatial
separation from the fruits of their conquest, which resulted in domination from a distance, made particularly acute the need
for reliable, quantifiable data for the full exploitation of their new sources of revenue. Thus, in the late sixteenth and early
seventeenth centuries the land was measured and the people on it

Epigraph: Saito[*] Yoshiyuki, "Kinsei shoki no nomin[*] toso[*] ," 53.

― 72 ―

counted. This created two new public and universal (because divisible and exchangeable) values: land aggregates and
populations. As value and data, the former had existed before, although in a different form, that is, less comprehensive and
not measured in the equivalent of bushels of rice. Population data, however, were altogether new. In Tokugawa Japan
everyone was to be accounted for.

The smallest unit for the collation, circulation, and exploitation of these values was the village as it emerged from the turmoil
of a century of warfare. Further military conquests having become impossible, the victors turned their full attention to tribute,
exacted now by decree rather than by the sword. The new rulers were not particularly interested in remote village affairs per
se and, indeed, did not interfere beyond taking the minimal steps necessary to extract tribute and corvée.

This broad picture conveys the impression of an incorporation or cooptation of villages into the Tokugawa power structure
with a minimal impact on them except for the new tribute system. In the past, agricultural surplus had flowed out of the
villages in a number of different directions. Now it went only to a single class of beneficiaries. Such an interpretation is
reminiscent of the conventional view of Tokugawa ideology, which assumes the appropriation of a ready-made doctrine, neo-
Confucianism, that was just "waiting in the wings" to be mobilized for legitimating purposes by the new regime. Until recently,
the Tokugawa village was also presumed to have been "already there," available for outside political exploitation. Scholarly
historiographical exploitation in turn consisted in writing about this entity within a framework that, like that of the rulers,
assumed stability.

More recently, however, historians have amended this view of the socalled Tokugawa village as a "natural community"
typified by "village-level" tribute exaction, "consensual" decision making, and "autonomous" self-governance; they have
historicized these qualifiers. This village, or more precisely, perhaps, because it is more image than anything else, the social
reality that (mis)informed the above image, took six to twelve Tokugawa decades, depending on the region, to develop, and
once established it continued to change, in unforeseen ways. In other words, the "Tokugawa village" (mura ) should not be
conceived of as a stable entity, but as an ever-changing Tokugawa village practice, a practice driven by external and internal
forces that were closely interrelated. A projection backward (and forward) has often informed the discourse, often structural,
about the Tokugawa village: the outcome

― 73 ―

of a long development during the Tokugawa period was assumed to have been there from the beginning (and to have
remained throughout).

According to the perspective adopted here, one does not confront an entity only to record its history. Instead, one traces its
genesis, which never ends, because, strictly speaking, time never produces lasting entities; they are produced only by
administrative records, and the historians who read them, who often wind up in abstractions that relate only vaguely to
practice.[1] Concretely this means that a "village" in the early seventeenth century, economically self-sufficient and consisting
predominantly of, and ruled by, a few extended families, had little in common besides its name with the form it took two
hundred and fifty years later, when it consisted predominantly of numerous small, often impoverished families tied into a
wide regional administrative network and a national economy.

Of course, change has not been totally absent from historical discourse on Tokugawa village society. The changes that have
been identified, however, have been mostly economic, generated by a commercial economy that developed during the
eighteenth century. From this perspective, the village has been viewed as a stable social formation, and change as externally
generated, economic in nature, and historically late rather than internal, political, and ongoing.

Historians and other scholars who look into the past who profess an interest in change often wind up missing it in their
analyses. I have discussed this problem at some length with regard to Tokugawa ideology, arguing that many students fell
victim to both that ideology itself and their own presumption that there is a universal need for all societies to have an ideology
formulated as a doctrine.[2] What, then, one may ask, blurred historians' vision so often and for so long when they looked at
Tokugawa villages, especially of the first half of the period? Why has

[1] These reflections on writing a history versus tracing a genesis are inspired by Michel Foucault's "Nietzsche, Genealogy,
History," in Michel Foucault, Language, Counter-Memory, Practice: Selected Interviews and Essays , ed. and trans. Donald F.
Bouchard (Ithaca, N.Y.: Cornell University Press, 1977), 139-64; and by Claude Lévi-Strauss, The Savage Mind (Chicago:
University of Chicago Press, 1966), 256-62.

[2] See my Tokugawa Ideology , as well as my "Neo-Confucianism and the Formation of Early Tokugawa Ideology: Contours of a
Problem," in Confucianism and Tokugawa Culture , ed. Peter Nosco (Princeton: Princeton University Press, 1984), 27-61.

― 74 ―

the myth of stable, undiversified, harmonious Tokugawa rural communities persisted for so long?

Diversifying the Picture of Early Tokugawa Rural Society

Some of the reasons are the same as the general reasons why "popular history" did not gain academic legitimacy until the
1960s. In a sense, while Japanese historians discussed at great length "the Tokugawa village," they did not see (into) the
villages, because "real" history presumably was happening elsewhere.

In the West, Thomas C. Smith more than any other historian has been sensitive to the divisions and tensions (mainly, but not
exclusively, economic ones) in seventeenth-century Japan. However, in 1958 he wrote that not much was known about the
important political issue of allocating within the village "the tax bill among its individual holders."[3] In his Agrarian Origins of
Modem Japan , he elaborated further but still had only his historical and sociological imagination, which he used with great
skill, to rely on because of a lack of data.[4] Much research has taken place since the publication of Smith's classic study in
1959.

Mizumoto Kunihiko, whose lead I follow, has argued that postwar Japanese scholars remained stuck until the 1970s in the
historiographical consequences of a Marxist structuralist interpretation of the rural policy that Toyotomi Hideyoshi (1536-98)
established and the early Tokugawa bakufu continued.[5] This policy would have succeeded in dividing larger clan holdings
into independent smaller ones.

[3] Thomas C. Smith, "The Land Tax in the Tokugawa Period," in Studies in the Institutional History of Early Modern Japan , ed.
John Whitney Hall and Marius Jansen (Princeton: Princeton University Press, 1968), 284, reprinted from Journal of Asian
Studies 18 (1958): 1.

[4] Thomas C. Smith, "Political Conflict in the Village," in his Agrarian Origins , 181-83. Most of that chapter, however, focuses
on the last century of the Tokugawa period, especially on conflicts centered on miyaza , or shrine associations, which were
most prevalent in the Kinai region.

[5] Mizumoto Kunihiko, "Bakuhansei kozoron[*] kenkyu[*] no saikento[*] —Asao, Sasaki-shi no shigoto o sozai hi," Atarashii
rekishigaku no tame ni (Minka Kyoto shibu rekishibukai kikanshi) 131 (1973): 13-23. For a full historio-graphical survey,
consult Kodama Kota[*] , Nagahara Keiji, et al, eds., Kinsei , Nihon rekishi taikei, 3 (Yamakawa shuppan, 1988), 385-407
(hereafter NRT 3). Mizumoto presents his argument in the form of a critical assessment of Asao Naohiro and Sasaki
Junnosuke's work, but his critique reaches beyond them to Araki Moriaki, Miyagawa Mitsuru, and even the prewar
scholarship of Furushima Toshio: Asao Naohiro, Kinsei hoken[*]shakai no kiso kozo[*] (Ochanomizu shobo[*] , 1967); Sasaki
Junnosuke, "Kinsei noson[*] no seiritsu," Iwanami koza[*]Nihon rekishi: Kinsei (2) (Iwanami shoten, 1963), 165-221; Araki
Moriaki, Bakuhan taisei shakai no seiritsu to kozo[*] (Ochanomizu shobo[*] , 1959); Miyagawa Mitsuru, Taiko[*]kenchiron , 3
vols. (Ochanomizu shobo[*] , 1959-63); Furushima Toshio, Nihon hoken[*]nogyoshi[*] (Shikai shobo[*] , 1941). Many of these and
other scholars formulated their interpretations of Tokugawa village society in essentialist Leninist terms, periodizing societies
according to their dominant socioeconomic formation as determined by their particular mode of production. Lenin, in his
Development of Russian Capitalism , used the Russian term uklad (literally, "structure") to refer to such socioeconomic
formations. Many Japanese historians made a fetish of the term, using it as an explanatory device that obviated any need for
further analysis. On the uklad theory, see also Mizumoto Kunihiko, "Bakuhanseika no nomin[*] keizai," Nihon keizaishi o
manabu: 2. Kinsei (Yuhikaku[*] , 1982), 66-70. A handy summary of the debate can be found under "ukurado[*] " in KDJ 2:56.
According to this entry, the locus classicus is in Lenin's 1921 paper "On the Food Tax" and his earlier Development of Capitalism
in Russia . Thus, the specific Tokugawa uklad was from the beginning one of independent, small landholders, created by
Hideyoshi's national land survey, which presumably had the effect of a land reform. Asao and Sasaki later added a second
uklad—of the extended patriarchal family—attaching differential values to the two uklads by arguing that local magnates, a
negative historical force, were eventually overcome through struggles of small peasants aided by the bakufu (Mizumoto,
"Bakuhansei kozoron[*] ," 14-18).

― 75 ―

Asao Naohiro, for instance, emphasized the crucial role played by the local leadership of the so -type[*] villages, predominant
in the Kinai region, around the capital, in the fifteenth and sixteenth centuries. So[*] villages were autonomous corporate
villages or groups of villages (of one go[*] , a medieval administrative district, or one shoen[*] , or medieval estate) where part
of the land was commonly cultivated and full administrative, judiciary, and often penal powers (including the death penalty)
were held by an oligarchy of prominent peasants.[6] The overlords subcontracted many of these corporate villages via their
elite as official units responsible for the total tribute due from their territory. This for-

[6] An excellent brief summary of the history and operation of corporate villages can be found in KDJ, s.v. "soson" (8:565-65);
see also the entries "jigeuke" (6:721), "sosho[*] " (8:549), and "gosonsei[*] " (5:424-25). For a case study of the most well-known
so[*] village (because best documented), see Tonomura Hitomi, Community and Commerce in Late Medieval Japan: The
Corporate Villages of Tokuchin-ho (Stanford: Stanford University Press, 1992). On jigeuke, see also Fujiki Hisashi, "Ikoki[*]
sonrakuron," in Nihon chuseishi[*]kenkyu[*]no kiseki , ed. Nagahara Keiji et al. (Tokyo[*] Daigaku shuppan, 1988), 208-10.

― 76 ―

mula of collecting tribute, then called jigeuke (perhaps best rendered as "land receivership"), was quite similar to the
subsequent Tokugawa bakufu's policy of murauke ("village receivership"). The bakufu, Asao argued, was able to win this
"corporate elite" to its own side and use it as the backbone of its local structure of domination. Thus, it would have been
wealthy rather than small peasants who enjoyed bakufu support.

For Sasaki Junnosuke, the mura, or village, was a Tokugawa creation, headed by both the janus figure of the headman, or
shoya[*] (the link between village and supravillage authority), and village officials. Under a general rubric of establishing
small, independent landholders, the bakufu would have implemented a differentiated policy toward local leaders, displacing
some and supporting others. Sasaki's view was based on his interpretation of late Tokugawa ikki , or uprisings in which
villages acted as a whole. He then applied this image of unified communities of small landholders fighting overlords in his
discussion of the early period.

Mizumoto's critique is based upon a careful examination of early Tokugawa so[*] village disturbances in the Kinai region.
Small landholders did not confront rural magnates, nor did solitary villages confront their overlords. Rather, these peasant
struggles were against the arbitrary rule of their own village leaders, struggles unrelated to official land policies.[7]

The thesis of the bakufu's promotion of independent small landholders was based on an examination of land surveys and laws
around the turn of the seventeenth century. Dissenting scholars, however, pointed out that the titled peasants officially
established through these surveys were, regardless of size, yakuya , or corvée houses, selected by overlords to deliver village
quotas of tribute and corvée (yaku ),[8] a responsibility that also, because it was not shared by all villagers, functioned as a
mark of local distinction and privilege. These titled peasants invariably not only controlled land but also had homesteads
(yashiki ) and were often land magnates and local bosses, sometimes with a semiwarrior background, in other words, the
village economic elite.[9] Thus, the social picture of early Tokugawa rural Japan consisted of registered

[7] Mizumoto, "Bakuhansei kozoron[*] ," 18-21.

[8] Mizumoto, "Bakuhanseika no nomin[*] keizai," 65.

[9] Ibid., 63-64.

― 77 ―

peasants/households, of whom some were singled out as accountable for tribute and others were not burdened with that
responsibility. The former were "titled peasants" (not to be understood in the restrictive sense of title to land); the latter,
although without homestead (muya-shiki ), often had some direct claim or perhaps merely a relationship of usufruct to land
that was entered under their name.[10] The picture was further complicated by the fact that some of the titled units were large
extended families, while others were small households.

Mirco Struggles I: Headmen, Elders, and Small Peasants in Corporate Villages (So[*] )

Mizumoto's discovery of internal tensions within early Tokugawa villages stimulated a reevaluation of the character of the
kyodotai[*] , or community, which historians assumed had been coopted by the new rulers. It was generally accepted that this
incorporation had been achieved by the policy of extracting tribute on a village basis (murauke), the village being understood
unproblematically as a "natural" community, its perimeters and internal compostion reflected in the land and population
surveys from around the turn of the seventeenth century. Yet these so-called villages (mura) that served as units of tribute
exaction were not infrequently artificially established territorial units created by the surveys, and not preexisting functioning
communities (kyodotai[*] ) merely recorded by them.[11]

The village cadasters did not include land residents held outside the village but did list the names and fields of nonresident
landowners.[12] Some cadaster "villages," therefore, actually counted more outsiders than insiders. In Higashi-Tenkawa
village, Settsu province (Hyogo[*] prefecture), for instance, insiders (89) were outnumbered by outsiders (133), residents in ten
different villages.[13] At the time of the land surveys some communities were split up into several villages, some regrouped
with neighboring ones to form a new unit based on the medi-

[10] Naito[*] Jiro[*] has stressed this point throughout his Honbyakusho[*]taisei no kenkyu[*] (Ochanomizu shobo[*] , 1968); see,
e.g., 28, 66, 178-79, 258-61, and 293.

[11] Mizumoto drew new attention to this problematic in Mura shakai , pt. 2, chap. 2.

[12] Mizumoto, "Bakuhanseika no nomin[*] keizai," 76-77.

[13] Mizumoto, Mura shakai , 145.

― 78 ―

eval go[*] , others simply moved to make room for castles, and many villages were brand-new, created on newly developed
land (shinden mura ) or as way stations on the highways.[14] These administrative interventions often triggered various forms
of resistance, protest, and local maneuvering, if not to undo them, then at least to mitigate some of their negative
consequences.[15]

Mizumoto's analysis of early Tokugawa disturbances in corporate villages of the Kinai problematizes the notion of these
"natural communities."[16] "Early Tokugawa" has a very specific meaning here, referring to the decades between 1600 and the
1680s, a period further divided at the 1640s. Compared with the well-known peasant revolts of the mid and late Tokugawa,
these early Tokugawa disturbances were distinguished by two characteristics. First, there were relatively few ikki, or large-
scale uprisings, and they were almost all village affairs. Second, although similar to later intramural (or intra-mura) suits
against abuses by village authorities concerning the calculation and allocation of tribute and corvée or their exemption
thereof, the class identity of plaintiffs and defendants changed over time, roughly, in the 1640s and again in the 1680s. Before
the 1640s, suits and petitions were signed only by the village elite, peasants who were from the same economic class as the
headman but who directed their litigation at him. After the 1640s the suits targeted, besides the headman, that same elite,
which now had gained power in village governance; the litigants now came from a new class of well-established landholders.
Toward the end of the century the plaintiffs were predominantly peasants with very small holdings.[17]

[14] For examples of split villages, see ibid., 139; for examples of several villages collapsed into one, see ibid., 146, and
Mizubayashi Takeshi, Hokensei[*]no saihen to Nihonteki shakai no kakuritsu , Nihon tsushi[*] , 2 (Kinsei) (Yama-kawa
shuppansha, 1987), 136; for villages displaced for the building of Komoro castle in Kita-Saku district, Shinano, see NAK-T 4:188;
for shinden mura in the same area, see Oishi[*] Shinzaburo[*] , Kinsei sonraku no kozo[*] , chap. 3; for a recent work on such
villages in the Kanto[*] plain, see the first two articles under the heading "Shinden kaihatsu to minshu[*] ishiki" in Chihoshi[*]
kenkyu[*] kyogikai[*] , ed., "Kaihatsu" to chiiki minshu[*] : sono rekishizo[*]o motomete (Yuzankaku[*] , 1991), 137-94.

[15] Mizumoto, Mura shakai , 137-90.

[16] Mizumoto, "Murakata sodo[*] ," 289-307; idem, Mura shakai , chaps. I and 3.

[17] Mizumoto, "Murakata sodo[*] ," 289-94; idem, Mura shakai , 7-17.

― 79 ―

The predominant complaint in pre-1640s suits of corporate villages was about the arbitrary exercise of power by headmen,
who made important decisions without consulting the sobyakusho[*] ("all the peasants"). The inclusive nature of this formula is
misleading, for it referred, not to all villagers, but only to the village elite and often to the core of that elite, the toshiyori , or
elders. In pre-Tokugawa times that elite was not ruled by someone with special powers like those of a headman.[18] It is
important to keep in mind that in these early decades of the seventeenth century the "elders" were not village officials in the
eyes of the overlords, even though they continued to play traditional, key roles in community and intervillage disputes.[19]
Headmen were established by the superordinate extramural (or extra-mura) powers in the 1590s and were picked from the
ranks of the village elite. Thus, this elite came to direct its early suits against the member from its own circle who had been
elevated to hold the new office of headman.

In the late 1640s, however, the character of these village disturbances changed in two ways: (1) the former plaintiffs were now
made the targets of legal suits; and (2) these lawsuits became an integral part of efforts to shift the village basis of privilege
and political power from pedigreed houses of tided peasants to households that were holders of taka , or tribute land. This
new class of plaintiffs identified themselves no longer as so-byakusho[*] , "all peasants," but as ko-byakusho[*] , "small
peasants"[20] (a term I shall elucidate later; for now, let us stay with the defendants).

The complaints voiced by these "small peasants" concerned village heads and elders, who came to be officially incorporated
into the structure of village governance around that time. On the yearly tribute rosters overlords sent to villages, they now
appeared as a special category of addressees (besides the headman and "all the peasants") because they filled the new village
offices of kumigashira , or kumi heads (of the five-household neighborhood groups created by the bakufu).[21]

[18] Fujiki Hisashi disagrees with Mizumoto on this point and stresses a continuity of headmen from pre-Tokugawa to
Tokugawa times in his "Ikoki[*] sonrakuron," 208.

[19] Mizumoto, Mura shakai , 19; see also Miyajima Keiichi, "Kinsei nomin[*] shihai no seiritsu ni tsuite: 1. Chusei[*] zaichiho[*]
no 'hitei' to 'naizai,'" Chihoshi[*]kenkyu[*] , no. 171 (1981), 1-11.

[20] Mizumoto, "Murakata," 297; idem, Mura shakai , 7, 20-22.

[21] Ibid., Mura shakai , 20.

― 80 ―

The development of this institution clearly illustrates the transference of military organizational formulae to civil rule. In 1597
Hideyoshi introduced five- and ten-man units in his army to better check internal discipline. Tokugawa Ieyasu (1542-1616)
used the idea to bring order to Kyoto in 1603, and his successor introduced the system in all bakufu territories. The five-
household groups were adopted nationwide in the mid 1630s in order to establish multipurpose subvillage administrative
units, which created new village power holders besides the village headman.[22] This tightening of control within the village
via a paramilitary institution has usually been understood as part of the bakufu's strategy against hidden Christians and
troublesome ronin[*] (masterless samurai). In light of Mizumoto's argument, however, other pressures were probably at work
as well, because the creation of these political offices signaled a victory by the old elite over a former equal, the headman.

The overlords thus modified the intravillage power distribution in two significant ways. First, they undermined the traditional
elite's oligarchic rule in corporate villages in matters of intravillage tribute and corvée allocation by appointing a headman.
Then the overlords broke the headman's monopolistic power through the establishment of kumi heads. By specifying the
elders' duties and prerogatives as kumi heads, the bakufu restored them to the political prominence they had all along
believed to be rightfully theirs. Thus the power base in the village was broadened, even though the majority of the peasants
were still left out.

This analysis modifies the standard picture of the relationship between village and state in the Kinai in important ways.
Villages were not ready-made units that were simply subsumed by the new state structures. To a certain extent they were the
creations of the warrior rulers, and not only because the surveys officially certified village borders. Yet these villages were not
made once and for all either. They developed in ways unanticipated by their creators, because they constituted differential
power fields whose members reacted to the new superordinate powers and forced those powers in turn to react to them.

Moreover, during the latter half of the seventeenth century village disputes in the Kinai region were political maneuvers by
landholders for formal recognition as fully vested village members on a par with

[22] For a brief sketch of this history of the goningumi, see KDJ 5:936, and Kodansha Encyclopedia of Japan , 9 vols. (Tokyo:
Kodansha International, 1983), 3:45.

― 81 ―

the titled peasants. As mentioned above, cadaster entries did not differentiate between peasants who held titles to land and
others who did not, but between peasants with and without homesteads. The disputes of the second half of the century were
new to the extent that they included attempts (with various degrees of success) by peasants, increasingly the small
landowners, to shift the village political and status order from one based on privilege attached to some households (originally
"homeowners") to one based on taka holding. I shall postpone the question why the new class of peasants could not simply
argue that many of them now also had homesteads.

What were taka holders? Taka means "amount" and stands for kokudaka , "(rice) yield (measured) in koku," koku being
equivalent to 4.96 bushels or 180 liters. Kokudaka refers, however, to assessed rather than actual yield. It was the official crop
index, recorded in cadasters and yearly tax rosters and used to determine the yearly tribute (nengu ). This index was
calculated by assessing the grade of the arable (superior, medium, inferior; sometimes there were as many as thirteen grades)
[23] and multiplying the assessed yield per square measure (e.g., per acre) by the actual size of the field. Dry fields and even the
size of homesteads were also converted into koku equivalent. A taka holder was thus a peasant whose name was associated
with some field listed on the cadasters, or more often, since few new surveys were made after the first ones, with some portion
of the village tribute or corvée on the yearly tax registers. Thus, tenants without land of their own were not registered as taka
holders, but their landlords were. It is of some importance to note that while taka has a close relation to plot size, it was not
identical to it. The taka value—tribute or official value, which of course affected market value—of two plots equal in size
differed if one plot had a higher assessed value than the other.

Village burdens could be allocated in a number of ways, depending on which constituent unit of the village—adults,
households, "titled peasants," the taka value of holdings—was the basis of computation. The drive in the second half of the
seventeenth century was for taka holders to be treated as titled peasants. More precisely, it was an attempt by newly
established taka-holding households to join the ranks of the traditional "titled peasant" households, a struggle by a new class

[23] Mizumoto, Mura shakai , 287.

― 82 ―

and generation of peasants for fuller recognition by the old one. Implicit in this drive was the logic of the convertibility of
economic power into political clout. The old guard's political importance was based on economic wealth; now, new
landholders wanted the same for themselves.

The earliest recorded attempt, dating from 1649-51 in Kami-Kawarabayashi village, Settsu province, was not initiated by
economically small holders (the number of similar disputes increased rapidly thereafter, peaking in the 1680s, when disputes
clearly were led by members of the small peasantry).[24] This early request to distribute various yaku (corvée and other
burdens) per assessed yield (see table 7) was one among a number of complaints against abuses by the headman and the elder.
The socioeconomic status of the two men accused (Y and Z) and of the fourteen plaintiffs (A-N) within the village can be judged
by the size of their individual holdings (kokudaka) and their ranking among the fifty-five taka holders of the village (all except
L, an unclear case, owned houses).[25] Some of the peasants also held land in the neighboring village of Shimo-Kawarabayashi
(see table 8).[26]
The plaintiffs were predominantly middle-ranking taka holders, A and B being exceptions as they rank even higher than the
elder (Z) in terms of holdings and servants (the elder had no servants, while A had four), and the plaintiffs were all titled
peasants. One may wonder why this drive for expanded political participation was led by titled peasants, who already held a
monopoly on matters of village governance. Were they eager to share their power? Only twenty-one of the fifty-five taka
holders were registered as tided peasants (here called honyakunin ), and three more were listed as half-titled peasants
(hanyakunin ), and the request by about one-third of the titled peasants was to turn all taka holders into tided peasants. We do
not know whether this request was ultimately granted, but in the 1670s L was listed as a titled peasant.

[24] Ibid., 69-70.

[25] Ibid., 72.

[26] Peasants often held land in neighboring villages; thus, some of the involved parties's names also show up on Shimo-
Kawarabayashi's tax roster. I recomputed and added data on the basis of the 1655 cadaster of Shimo-Kawarabayashi, in
Nishinomiya-shi shi , vol. 4, Shiryo-hen[*] , pt. 1 (Nishinomiya: Nishinomiya shiyakusho, 1962), 496-97. Kami-Kawarabayashi's
cadaster can be found on 494-95, its population register of 1659 on 620-32. If we assume that some identical names on the two
rosters refer to the same persons, Y of table 7 would have held an additional 9.2 koku in Shimo-Kawarabayashi; C, 4.6; H, 7.5;
and K, 16.6.

― 83 ―

Table 7.
Plaintiffs and Defendants, Kami-Kawarabayashi, 1649-
1651

Taka Holding

Peasants Koku Rank Status

Defendants

Y 30.4 1 Headman

Z 15.0 10 Official elder

Plaintiffs

A 22.2 3 Titled peasant

B 15.8 8 Titled peasant

C 14.7 11 Titled peasant

D 14.5 12 Titled peasant

E 14.3 13 Titled peasant

F 14.2 14 —

G 11.2 17 Titled peasant

H 9.0 22 —

I 8.4 23 —
J 6.7 26 Genin of B

K 5.9 29 Titled peasant

L — — Retired head of B

M — — —

N — — —

SOURCE : Mizumoto, Mura shakai , table 1 (p. 72).

NOTE : The headman also has eleven co-resident and

five separately housed bond servants (genin). Taka
holdings are limited to land held within the village .

To answer the question why titled peasants might agitate to have their ranks opened to others, it is important to draw a
distinction that may seem odd at first. In Tokugawa Japan, being accountable for tribute was not the same as contributing to it.
Prior to these requests, the burden and privilege of public accountability for yaku (mainly tribute rice but also buyaku , corvée)
was limited to "corvée households" (rifled peasants). This title, as mentioned above, did not identify taka holders: although all
villagers with some taka shared or contributed to the tax burden, only the tided peasants were accountable for it in the eyes of
the overlords. This distinction simplified tribute extraction for the

― 84 ―

Table 8.
Taka Holders in the Neighboring Villages of Kami- and
Shimo-Kawarabayashi

Taka Holders (number of known plaintiffs )

Koku In Kami- In Shimo-Kawarabayashi

Kawarabayashi

30+ 1 0

25-30 1 1

20-25 4 (1) 3

15-20 4 (1) 5

10-15 7 (5) 5

5-10 14 (4) 8

1-5 13 11

-1 11 26
Total 55 59

SOURCE : Nishinomiya-shi , vol. 4, Shiryo-hen[*] , pt. 1, 496-97.

overlords because it limited responsibility for it to the prominent villagers—a simplification of social space and also of social
time, since it made no provisions for changes in village demographics.

When siblings or other dependents of titled peasants were established in separate quarters on a land parcel or when
household heads retired on a separate plot, inkyo bunke (Kami-Kawarabayashi counted thirteen such cases in 1649, L being
one of them), these branch houses did not automatically become titled peasants. The drive for taka-based access to the status
of titled peasant was the result of the proliferation of new landholders. They shouldered tribute and corvée burdens without
sharing the political privilege attached to the title. Given the large number of branch houses established by retiring titled
peasants in Kami-Kawarabayashi—thirteen—the attempt to give these new households full political status as titled peasants
may have been part of a strategy to increase the power of expanding lineages. It is important to remember that this is the first
occurrence of such protest found to date and that it is somewhat unusual in that the initiative came from titled peasants. The
number of such petitions, however, increased dramatically in the 1670s, when nontitled taka holders initiated them.

These petitions for entitlement, when granted, often signaled the end of the traditional miyaza in corporate villages, where a
limited number

― 85 ―

of "shrine families" had succeeded in hanging on to their religiopolitical monopoly. Originally, in the fifteenth and sixteenth
centuries, within the framework of corporate villages of a district or estate, the elders not only were secular leaders but also
monopolized ceremonial life through the Shinto shrine associations, or miyaza. In the late sixteenth and seventeenth
centuries, however, other peasants came to function as liturgical celebrants (kannushi ) within the individual villages that
emerged after the dissolution of the larger corporate units.[27] It should be noted, however, that miyaza continued to function
as exclusive political clubs in a great many villages in the Kinai.

The Kawarabayashi area offers an example of such a dissolution of miyaza-based power. Its Hino shrine originally venerated
the dominion lord Kawarabayashi as the ujigami , or tutelary deity. In the early 1600s he remained the patron god of only
three villages, including the two Kawarabayashi's, and in the second half of the seventeenth century his sacred jurisdiction
shrunk even further, comprising only Kami-Kawarabayashi and its branch village of new fields, Goroemonshinden[*] .
Ceremonial control was now in the hands of three families in the village. Toward the end of the century even they lost their
status because villagers took turns presiding over religious ceremonies.[28]

Initially, as already mentioned, these struggles were led by traditional titled peasants against the headmen and the elders.
They were followed in midcentury by demands from a different sector of the village population, namely, small taka holders
who were not titled peasants. Although their motive may have been political, their professed aim was survival. They argued
that their share of village expenses was disproportionately high since these were divided equally per household rather than
calibrated according to holding size (taka).[29] As can be seen in table 7, many of the taka holders in Kami-Kawarabayashi
were dangerously close to having no taka at all.

Around 1700, plaintiffs explicitly incorporated in their petitions the "landless" peasants (mudaka , "without taka, " or
mizunomibyakusho[*] , "water-drinking peasants"). These terms, like "titled peasant," should

[27] Mizubayashi, Hokensei[*] , 32-35.

[28] Ibid., 223-24. Mizumoto presents another case where the struggle against entrenched religiopolitical power was waged at
this level: Tateishi, in Yoshino district, Yamato province, in 1677-79 (Mura shakai , 74-74).

[29] Mizumoto, Mura shakai , 76-80.

― 86 ―

not always be taken literally, for many mizunomi peasants had some, if little, land of their own. In other words, they were not
necessarily pure tenants. These suits were now directed explicitly not only against headmen and elders but also against great
taka holders (otakamochi[*] ).[30] These plaintiffs perceived themselves as virtually landless and made common cause with
those who actually were.

Various class fractions of the peasantry realigned themselves thus in their attempts to participate fully in decisions that
touched their lives. Before the 1640s it was the traditional elite against headmen. In the second half of the century old titled
peasants challenged headmen and elders. Then, in the 1670s all taka holders pressured the village leadership for widening
decision-making participation. And finally, around 1700 the marginal peasants made common cause with the "landless" ones.
At this last juncture, however, the argument changed because landless peasants could not be incorporated on the basis of their
taka. They pleaded for corvée to again be computed per household rather than per taka.

This last point is hard to explain, as Mizumoto admits, because the request seems to work to the economic disadvantage of the
poorer peasants.[31] Landless homeowners would be tax-exempt in a system where corvée and tribute were based on land
holding but not in a system based on home ownership. This leaves one hypothesis, namely, that the poor peasants stood to
gain politically from being recategorized closer to the status of full-fledged village members. It is important to note that the
system they wanted, based on household, or ie , was different from that of the early Tokugawa period, because now it would
include all households, not just a privileged few.

This drive did not succeed. As we shall see in chapter 3, political membership did not open up. On the other hand, many
compromise solutions were reached in which burdens were computed in part on the basis of holding size, in part on the
number of households. At the level of consciousness, however, things did change, as Mizumoto argues in his perceptive
analysis of the term kobyakusho[*] , "small peasants."[32] We are now in a position to clarify the uses and meanings of this
term, for there are several.

[30] Ibid., 80-85.

[31] Ibid., 82.

[32] Ibid., 85-102.

― 87 ―

In the first four decades of the Tokugawa period petitioners signed collectively as "all the peasants," a phrase that in reality
referred only to the village elite, who had always spoken for "all the peasants." "Small peasants" was the identity taken by all
successive categories of later plaintiffs discussed above: first the heads of "corvée households," or titled peasants, then taka
holders, followed by marginal peasants, and finally marginal and landless peasants together. As time went by, the term came
to include, for the people , more and more sectors of the village population, all of them politically disenfranchised, and it
gradually came to include the economically less well off.

For the bakufu , however, it was a different matter. There was stability of meaning over the same period. In its documents and
legislation, the bakufu (and some domains) differentiated a specific large category of "small peasants" from other villagers.
The term excluded village officials, bond servants (nago and genin), handicraft workers, and outcastes and referred only to
taka holders, even very small ones. Small peasants were thus tribute producers , even if they themselves did not directly pay
the tribute, for example, branch houses for whose tribute patron families or lineage heads were responsible.

These discrepancies between rulers and ruled in the referents for the term "small peasants" reveal that initially the bakufu
defined it more broadly (all taka holders) than the commoners (titled peasants). Subsequently, the scope of both usages
converged when the plaintiffs included all taka holders beyond the titled peasants. And finally, the peasants expanded the
term's meaning beyond the bakufu's to include nonholders.

In the area of village practice other than suits, Mizumoto further documents the growing importance of taka holders in village
affairs. His analysis of the identity of those who affixed their seals to tax rosters and village budgets in a couple of villages in
the late seventeenth and early eighteenth centuries reveals a sudden expansion of those who endorsed these documents and
participated in their composition. In one case, 45 percent of all taka holders signed the budget in 1676 but eleven years later
100 percent signed. In another, all taka holders signed the tax roster in 1706, while only 85 percent had signed four years
earlier and only 54 percent had done so in 1680.[33]

[33] Ibid., 95-101.

― 88 ―

These various developments in consciousness and practice invite comments on interpretations of class in early Tokugawa
rural society. Herbert Bix has attempted to document a rising peasant class consciousness and an expanding oppositional
political practice against the ruling samurai starting in the eighteenth century, when according to his analysis class became
important.[34] One weakness of his argument is that for the seventeenth century Bix worked basically with the official, legal,
Tokugawa status definition of peasants and samurai and thus did not pay sufficient attention to economic and class divisions
within the villages. The present analysis, however, reveals political action based on the recognition of the unbalanced relation
between political and economic power, not between samurai and peasants, but within the peasantry. This configuration of
practice, perception, and the reality of inequality was framed by the social and political structure established by the ruling
samurai. Its setting was limited to the village, and the lines along which the confrontations took place were those that the
rulers initially drew within the field of power—the village itself—lines subsequently redrawn when the population expanded.

The bakufu, as stressed earlier, was interested in villages only insofar as they were units of submissive tribute producers.
Through legislation, it engaged in shaping the world of the village for that purpose. This "world-making" (to use Bourdieu's
term) through legislation produced the headman, responsible for the village tribute, and titled peasants, the tribute payers.[35]
Yet as the cadasters show, the rulers were also aware that various relations to the land existed at a level below that of the titled
peasants, who often leased, granted, or gave parcels of their land to dependents. It appears that the rulers decided not to deal
with these messy and shifting proprietary relations and instead chose the most cost-effective way to extract tribute: they
assigned responsibility for paying it to the headman and the titled peasants. At the same time, however, they wanted to make
sure that all agricultural (and hence, ultimately, tribute) producers, that is, all those who held some taka in one form or
another (all small peasants) were notified of their share of

[34] Herbert Bix, Peasant Protest in Japan, 1590-1884 (New Haven: Yale University Press, 1986), xvii, 104.

[35] Pierre Bourdieu, "The Force of Law: Toward a Sociology of the Juridical Field," Hastings Law Journal 38 (1987): 837-40, 846.

― 89 ―

the tribute burden. Thus, the bakufu also created a new class of "small peasants," because it named them as such in official
documents.
The bakufu thus reinscribed existing power relationships, a move welcomed by those whose interests were thus reinforced.
Responsibility for the delivery of tribute was used for internal political purposes, first by the headman, then by the headman
and the official elders, and finally by the titled peasants. The others, however, whose numbers increased dramatically through
the division of property (widely practiced during the seventeenth century), leveraged their status as taka holders to increase
their voice in village affairs. Mizumoto even surmises that only when some internal political unity was achieved, around the
turn of the eighteenth century, did villages start to agitate in unison in the larger peasant disturbances directed against
supravillage domination in the eighteenth century.[36] Be that as it may, this suggestion certainly reverses the standard picture
of harmonious village communities in the early Tokugawa period, which would have begun to dissolve much later, when the
peasantry fractured according to diverging class interests.[37]

Macro Power I: The Colonial Variety

The dynamic relationship between Tokugawa villages and lordly power can be properly grasped only if one understands the
specific nature of the political order established by the shogun and the daimyo. This new order was a regime of conquest like a
colonial regime. Once the shogun and daimyo had taken over the country and eliminated or neutralized challengers from
without (the court, religious establishments, commoner armies) as well as from within (rival daimyo or powerful vassals), they
regrouped their troops in the towns and cities like a colonial army and faced the task of extracting tribute from the villages
without direct use of coercive force. For this job they picked prominent natives (insiders) from among the local elite, backing
them with their lordly military authority, and granted them privileges, mainly in the form of tribute exemptions. This policy
created a new site of power in the villages and ample space for arbitrary local rule. The traditional elite

[36] Mizumoto, Mura shakai , 101.

[37] This is discussed in chapter 3, where the views of American scholars are presented.

― 90 ―

rebelled, wrangling a share in the power from those thus selected from among its ranks. The lords accommodated this new
power distribution because it eliminated a trouble spot at the site of tribute production.

Colonial powers profoundly affect the life of villages from a distance, if only through new ways of recording them or by their
contacts with inside authorities. Similarly, the massive shifts in the field of macro power beyond the village during the decades
around the turn of the seventeenth century affected village practice in the corporate villages of the Kinai—and other types
elsewhere as well, as we shall see. The authority of powerful, albeit absent, rulers now loomed behind the local (partly new)
leadership, thereby transforming its character, limiting its authority in some respects and enlarging it in others. Moreover,
absentee rulership meant the exercise of coercive force through decrees, regulations, directives—rule by law and documents
—which in itself transformed the nature of village governance.

In this context, let us return for a moment to the question of the purpose and social effect of the national land surveys, which
constituted the first drastic intervention in village life in the period under study. Japanese scholars are divided on this
question, although perhaps less than twenty years ago.[38] Mizubayashi Takeshi, in his brilliant synthesis

[38] American scholars have taken the following positions on this question: John Whitney Hall, in the context of a discussion of
the social effect of the land surveys, shifting the focus to village government, states that the "'landlord peasant' status was in
fact protected and recognized by the structure of village self-government which relied on the wealthy peasant for positions of
responsibility within the village" (Government and Local Power in Japan, 500 to 1700: A Study Based on Bizen Province
[Princeton: Princeton University Press, 1966], 321). Thomas C. Smith shares this opinion in "The Japanese Village of the
Seventeenth Century," in Hall and Jansen, Studies , 265-68, 280. Mary Elizabeth Berry mentions the two sides of the
controversy, allowing some room for the equalizing theory, but ultimately seems to decide that Hideyoshi's surveys were not
revolutionary with respect to landholding patterns (see her Hideyoshi [Cambridge: Harvard University Press, 1982], 118-21).
Stephen Vlastos, on the other hand, speaks of "a new order in rural Japan characterized by a profound leveling of social and
economic status" (Peasant Protests and Uprisings in Tokugawa Japan [Berkeley: University of California Press, 1986], 27). See
also Philip C. Brown, "The Mismeasure of Land: Land Surveying in the Tokugawa Period," Monumenta Nipponica 42, no. 2
(1987): 115-55. Recently Brown has given the question the most extensive discussion available in English in his Central
Authority and Local Autonomy in the Formation of Early Modem Japan: The Case of Kaga Domain (Stanford: Stanford University
Press, 1993), 16-19, 54-88, a book that came to my attention as this manuscript was going to the publisher.

― 91 ―

of Tokugawa history, offers cogent reasons why the cadasters did not establish a peasantry of small landholders and were
never intended for that purpose. According to Mizubayashi, the three kinds of village-based data compiled under Hideyoshi
had the combined effect of sanctioning existing land holding and land management patterns of small and large households (ie
).[39]

The most well known of the registers is the land cadaster, or kenchicho[*] , listing plots of land, of insiders and outsiders alike,
in the order that the plots lined up geographically in each village. The nayosecho[*] , or name register, listed these plots within
each village per holder household (ie). Although compiled in the village, this name register was an official document since it
was used to indicate the amount of tribute owed by the village.[40] A third register was the iekazucho[*] (literally, "register of
the number of households"). It listed by and large the same households that were recorded on the other two registers. Here,
however, the purpose was to identify those households that were responsible for corvée to the overlords. These documents not
only established a tribute relationship between overlord daimyo and their subjects but also linked the village population of
peasants as national subjects to the "state" at large under the shogun.
This relationship of national corvée for commoners was similar to that between the warriors and the state. For the warriors
the unit for performance of service was each ie, as registered in the name rosters of the vassal bands; for the peasants it was
the village as a collectivity of ie, recorded in the household register, another instance of a transfer of military organization
techniques to the nonmilitary sector of society. The amount of tribute, corvée, or service was determined according to the size
of the land listed in the cadaster, and thus, although it was channeled through the "titled" peasants (owners of homesteads), it
was also produced by the nontitled (muyashiki) peasants. Some of the latter,

[39] Mizubayashi, Hokensei[*] , 129-36. Brown (Central Authority , 75-76, 91) shares this opinion.

[40] Shimadani Yuriko makes the same argument, but her research indicates that some of these name registers were compiled
by intendance (see her "Kinseki 'honbyakusho[*] ' saikosatsu[*] no kokoromi," Jinmin no rekishigaku , no. 86 [1985]: 14). She
also argues that lordly corvée in the early Tokugawa period, specifically for the battle of Osaka, was not levied only on elite
yakuya , or service households, but was levied on all holders according to the size of their holding, taka, and that the land
cadasters did not have lists of yakunin , service people, attached to them (ibid., 13, 21). This, she suggests, was not exceptional.

― 92 ―

as outside owners of plots within the village, may have been titled peasants in their home villages.[41]

A comparison of entries in a name register with those in its corresponding cadaster reveals that the latter contains far more
names, because it includes, besides the names of household heads, those of a number of dependents who worked the land.[42]
This association of names with plots of land provides the basis for the argument that the cadasters reflect a policy by
Hideyoshi to break down large holdings managed by extended families into smaller units and that the name registers
compiled in the village were perhaps a means for the large holders, who listed only themselves there, to resist Hideyoshi's
policy.

Mizubayashi questions the validity of this hypothesis because the cadasters did not focus on land ownership ; if they had, their
entries would have been identical to those on the list of household names. Instead, since one of the purposes of the cadasters
was to make it possible to enlist manpower for service even at the national level, official cadasters recorded dependents
working field parcels of extended families. If the overlords' intention had been to promote these dependents to economic
independence, it is hard to imagine how this could have been achieved for people not even listed as owners in the household
list simply by entering them on the cadasters as independent. The cadasters listed the manpower that worked the fields but
could be put to other uses as well.

Dependents working identifiable plots were not necessarily fully independent owners. The two lists converge only much later,
toward the end of the seventeenth century, when the number of owners increased and that of dependents on the cadasters
decreased. The late-sixteenth-century cadasters simply reflect the composition of the various types of ie: the large, extended
peasant households (what modern historians call dogo[*] ), single peasant families, and even nonpeasant families and mixed-
economy units. The modern historiographical term dogo[*] refers to wealthy landed villagers with local political clout, which
in the sixteenth century could be, and often was, temporarily converted into military power. This class of landowners included
"samurai of the soil" (jizamurai ) and "men of the provinces" (kokujin ), who enjoyed a more

[41] This important point is made by Naito[*] , Honbyakusho[*]taisei , 294.

[42] For a comparison of the two registers of one village, see Mizubayashi, Hokensei[*] , 133.

― 93 ―

official status of authority, at least nominally.[43] Villages ruled by dogo[*] families were typified by pronounced internal
economic (and concomitant political) discrepancies compared with so villages, where such stratification was less severe.

Certain households were thus registered on the household registers, compiled nationwide in 1591 by order of Hideyoshi, as
responsible for a wide range of corvée.[44] Takagi Shosaku[*] has drawn attention to the national character of some of this
corvée that was earmarked as separate from that owed to the local overlord.[45] It included auxiliary noncombat corvée,
needed by campaigning armies for transporting materiel and provisions (also provided by the villages) and building roads and
dikes and was official corvée away from home ordered by the highest authority, the shogun (kuniyaku , or national corvée).[46]
With the development of a system of way stations on the network of highways,

[43] Ibid., 134. For definitions of these and similar medieval terms, see John Whitney Hall, "Terms and Concepts in Japanese
Medieval History: An Inquiry into the Problems of Translation," Journal of Japanese Studies 9, no. 1 (1983): 1-32, esp. 23-32. For
a case study of one such family, see Suzanne Gay, "The Kawashima: Warrior-Peasants of Medieval Japan," Harvard Journal of
Asiatic Studies 46, no. 1 (1986): 81-119.

[44] In the registers these households were variously referred to as honbyakusho[*] , basic or titled peasant; honke , main
house; buyakunin , corvée person; kogiyakunin[*] , official servant; or kujiya , official house. Modern Japanese scholars refer to
them as yakuya, or "service households" (Mizubayashi, Hokensei[*] , 141-42).

[45] The major findings of Takagi's research on the mobilization methods of late-sixteenth-century and early Tokugawa armies
can be found in the following articles: "'Kogi[*] ' kenryoku no kakuritsu," in Bakuhansei kokka no seiritsu , ed. Fukaya Katsumi
and Kato[*] Eiichi, Koza[*] Nihon kinseishi, 1 (Yuhikaku[*] , 1981), 151-210; "'Hideyoshi no heiwa' to bushi no henshitsu:
chuseiteki[*] jiritsusei no kaitai katei," Shiso[*] , no. 721 (1984): 1-19, translated as "'Hideyoshi's Peace' and the Transformation
of the Bushi Class: The Dissolution of the Autonomy of the Medieval Bushi ," Acta Asiatica , no. 49 (1985): 46-77; "Bakuhan taisei
to yaku," Ken'i to shihai , Nihon no shakaishi, 3 (Iwanami, 1987), 309-41 (hereafter NNS 3). See also his lengthy contribution to
NRT 3:160-223. For an answer to his critics, see his "Kinsei Nihon ni okeru mibun to yaku: Minegishi Kintaro-shi[*] no hihan ni
kotaeru," Rekishi hyoron[*] , no. 446 (1987): 90-108. Minegishi argues that Takagi overstresses the effect of Hideyoshi's status
laws at the expense of the economic underpinnings upon which they were based (see Minegishi Kintaro[*] , Kinsei mibunron
[Azekura shobo[*] , 1989], 183-218).

[46] KDJ distinguishes between kokuyaku (5:697- 98) and kuniyaku (4:845-46). The former seems to be limited to extraordinary
national or provincial levies of pre-Tokugawa times (1054-1546), ordered by the imperial or shogunal courts. The discussion of
the latter is limited to the Tokugawa period and to corvée by artisans. These volumes, published in 1984 and 1985, did not take
into account Takagi's work, perhaps because it was too late to incorporate it or, less likely, because the authors of these entries
disagree with it.

― 94 ―

portage duties, or sukego , were also assigned to surrounding villages by shogunal authority.

This mobilization of peasants by the highest national authority for noncombat military service affected local authority
relationships in a number of ways. The villages, as identified on the rosters and registers, became production and support
brigades for the "national," or shogunal, government in times of war. Thus, the whole country was, at least latently, a "garrison
state," as Takagi put it.[47] When the daimyo received orders from the shogun to mobilize, many "natives" were assigned to
serve in the "colonial" (samurai) army of daimyo, who were often transferred from elsewhere to rule them. This arrangement
incorporated not only villages as a whole but also individual peasant households directly into a national organization, giving
those selected for this service an enhanced local identity. Since there were virtually no national mobilizations during the
Tokugawa period, however, the greatest effect of this system was not its practical use but the creation of an additional prestige
value within the village, a form of symbolic capital, to borrow Pierre Bourdieu's famous simile, that was put to use to yield
local political profit.[48]

This institution affected dogo[*] magnates in two ways. On the one hand, it strengthened their hold over their dependents,
since national authority added a surplus dimension to some aspects of their local power. On the other hand, this local power
was now in certain respects shared with a higher authority, clearly stipulated, and hence limited. This limitation of dogo[*]
power was the result of a mobilization formula that was imposed on them via the kokudaka system of the cadasters,

[47] Takagi, "Hideyoshi's Peace," 46.

[48] Pierre Bourdieu writes about the relationship between economic and symbolic capital as follows in his In Other Words:
Essays towards a Reflexive Sociology (Stanford: Stanford University Press, 1990): "A capital (or power) becomes symbolic
capital, that is, capital endowed with a specifically symbolic efficacy, only when it is misrecognized in its arbitrary truth as
capital and recognized as legitimate and, on the other hand,... this act of (false) knowledge and recognition is an act of practical
knowledge which in no way implies that the object known and recognized be posited as object" (112, italics in the original; see
also his Logic of Practice , 118).

― 95 ―

and it differed from the one used by the warlord daimyo of the sixteenth century. When these daimyo mobilized, their orders
pertained only to warrior retainers, and not to noncombatants. Specifications for these daimyo's retainers stipulated only the
"kind and number of weapons and those wielding them";[49] support personnel were not listed. Each warrior provided his
own followers through his personal authority as a local magnate, and he shouldered their expenses as well. The novelty of the
Hideyoshi/Tokugawa system consisted in that now (1) the over-lord's mobilization order reached commoners directly, over the
heads of dogo[*] /retainers, the local bosses; (2) the number of porters and other followers was specified; and (3) expenses
were shouldered by the overlord. In other words, local magnates saw the autonomous nature of their authority undercut by a
strengthened center.

Battle array charts (jindatesho ) thus included a great number of noncombatants, mostly peasants. The original purpose of the
kokudaka system was precisely to determine the number and duties of these noncombatants, who constituted the bulk of the
armies. Takagi gives as an example the composition of the army of Sakai, daimyo of Maebashi (130,000 koku), toward the
beginning of the eighteenth century. Only about one-third of the total number of men and horses of the main force were
assigned combat duties: 2,000 men out of an army of 5,344. The vanguard consisted of a corps of "engineers" who built field
camps for the troops; it included 10 carpenters, 50 porters (ninsoku ), and 2 horses (see table 9).[50]

Since under the Tokugawa peace there were no national mobilizations after the Shimabara rebellion of 1637 (except in the
two campaigns against Choshu[*] domain in the 1860s), one might wonder about the significance of the inclusion on paper of
so many commoners in the armies of the daimyo and shogun. This mobilization formula, although not put into practice for
over two centuries, contributed to the peasants' identity.

First of all, some of the national duties for which peasants were responsible were converted into permanent tribute categories
in the second half of the seventeenth century. A portion of their tribute, cash (busen ) as well as rice (bumai ), was earmarked
as military tax.

[49] Takagi, "Hideyoshi's Peace," 54.

[50] These figures are Takagi's, taken from NRT 5:191-92; they are more precise than those he gives in "Hideyoshi's Peace," 47.

― 96 ―

Table 9.
Composition of Sakai's Army, Early 1700s
Combatants

Mounted samurai 347 (1)

Samurai on foot 87 (1)

Foot soldiers 1,129 (10)

Noncombatants

Rear vassals attending mounted 1,694 (0)

samurai

Lackeys, servants 306 (10)

Porters 1,179 (50)

Grooms 573 (10)

Artisans, doctors, cooks, etc. 29 (10)

Horses

Riding horses 347 (1)

Pack horses 567 (10)

SOURCE: NRT 3:191, table 1.

NOTE: Figures in parentheses refer to the

vanguard; a number of the rear vassals should
be included among the combatants.

Furthermore, the grandiose shogunal progresses to Ieyasu's shrine in Nikko[*] , a total of fifteen, including those before its full-
scale construction in 1636, also functioned as mobilization exercises, as did, on a smaller scale but more frequently, the
daimyo's yearly marches between Edo and their domains, as required by the system of alternate attendance (sankinkotai[*] ).
In the progress of 1776, for instance, the shogun, escorted by his three collateral houses (gosanke ) and twenty daimyo, was
accompanied by no less than 620,000 low-ranking soldiers (zohyo[*] ), 250,000 porters, and 305,000 horses. These processions
took one week and were so large that, according to an eyewitness, they stretched the whole length of the road, which had
twenty-one way stations, from Edo to Nikko[*] (145 km).[51] In the Choshu[*] wars of the 1860s, the peasants from the bakufu
and some daimyo domains were effectively mobilized according to the formulae fixed over two hundred years earlier, readily
accepting noncombat assignments as a "natural" part of their duties as subjects. By the same token, they resisted as
"unconstitutional" the

[51] Watanabe Hiroshi, "'Goiko[*] ' to shocho[*] : Tokugawa seiji taisei no ichisokumen," Shiso[*] , no. 740 (1986): 138.

― 97 ―

attempts to arm them for combat, insisting that if they were to bear arms they should be treated as samurai.[52]

Undoubtedly, the great division in Tokugawa Japan was between warriors and commoners. However, the peasants were
essential even for the military organization of the country, and their public selfidentity included noncombatant military
service at a trans-domain, if not "national," level. To that extent, in addition to being subjects of their daimyo, they were also
subjects, not of the emperor, but of the shogun, since he decided when mobilization was required.
Micro Struggles 2: Peasants Versus Magnate Headmen (Dogo[*] )

Mizumoto's regional study of predominantly the corporate type of village left unexplained the relevance of its findings for the
rest of Japan, especially the "less advanced" regions, where dogo[*] households dominated village life. A typical example would
be Toyota village in Izumi province (Osaka), where two dogo[*] households (5 percent of the total number of households), each
with an average of ten dependents, together controlled 30 percent of the arable.[53]

Saito[*] Yoshiyuki has looked into this question with regard to a number of villages of eastern Japan, in Niigata, Nagano, Akita,
and other prefectures, and presents the following argument linking protests regarding the fair distribution of tribute
throughout the village with the establishment of the village-based system of tribute levy (murauke).[54] In other words, he
indirectly traces the genesis of yet another basic Tokugawa institution. Saito[*] starts by pointing out that this murauke system
was not entirely a bakufu innovation. As mentioned earlier, a similar practice (jigeuke) was not uncommon on the old estates
and in corporate villages.[55] Nevertheless, even there its implementation by the newly established headman caused problems.
Saito discerns two issues related to headman governance: abuse of authority and excessive privileges , the latter an extension of
their informal power as local magnates; and different responses to these problems by the bakufu before and after the

[52] Kurushima Hiroshi, "Kinsei gun'yaku to hyakusho[*] ," Futan to zoyo[*] , Nihon no shakaishi, 4 (Iwanami, 1986), 273-317
(hereafter NNS 4).

[53] Mizubayashi, Hokensei[*] , 42.

[54] Saito[*] Yoshiyuki, "Kinsei shoki no nomin[*] toso[*] ," 42-60.

[55] Ibid., 43.

― 98 ―

early 1640s, first unambiguous support for the headmen , then an acknowledgment of the peasants's complaints .

The issue of abuse of authority is illustrated by a case from Echigo, Otorisu[*] village, where forty peasant households were
ruled by two headmen (kimoiri , as they were called there). The headman of the East Kumi (possibly consisting of several
neighborhood kumi), backed by all his peasants and some from the West Kumi as well, complained in 1662 to the intendant of
Nagaoka domain that his colleague from the West Kumi had secretly overtaxed the new fields (shinden) of the East Kumi for
the last eight years. The suit stressed that the accused headman had acted "without consulting the East Kumi's head and all the
peasants." The case was settled out of court, the official yearly tax rosters (nengumenjo[*] ) were made public, and the head of
the West Kumi was deposed. An apology from those West Kumi peasants who had not endorsed the suit from the beginning
closed the affair. Like corporate (so[*] ) villages, this was a village without great economic discrepancies between landholders.
The issue was abuse of authority by a beadman .

Other protests addressed the issue of the informal power of local magnates. The village of Shimo-Okamoto in Utsunomiya was
ruled by a peasant magnate (dogo[*] ) with 166 koku, while the other fifty-three peasants together held 840 koku. Over a span
of forty years (1637-78) at least six protests were lodged against the headman. The first was a complaint by all the peasants
about the headman's exemption from his portion of corvée (proportionate to his 166 koku), which they had to shoulder.
Headmen from neighboring villages mediated a settlement whereby the headman in question had to pay a fee in corvée rice
(bumai) on half of his holdings. Seven years later, in 1644, fourteen peasants successfully petitioned the domain to form a
separate village perhaps in order to escape shouldering the headman's exempt corvée burden, which now fell on only thirty-
nine peasants. Ten years later they demanded the removal of the headman. Finally, in 1678, after a few other protests, they
asked the intendant's approval for apportioning corvée not per household but according to each holding's assessed yield ("per
taka"). These suits did not concern flagrant abuse of power by a headman; rather, they aimed at changing the traditional
patriarchal authority the headman, as a local magnate, wielded over small landholders , who were very dependent on him.

The bakufu's response to these two types of protest shifted significantly after the Kan'ei period, which ended in 1643. Before
that turn-

― 99 ―

ing point the bakufu unambiguously protected its village-based tax agents by siding with the headmen, leaving their public
authority unaffected, and by withstanding attempts to limit their privileges.

Two examples from Shinano province in 1621 illustrate how the bakufu backed its appointed headmen. In one, the intendant's
representative, together with powerful figures from neighboring villages, saw to it that embezzled money, whose absence was
discovered after the village account books were opened, was returned, but no further action was taken against the headman.
In the other, the intendant refused two successive requests by the peasants to check the tax lists. Obviously, access to the
records of intramural tribute allocation was not backed by bakufu authority.

The bakufu's protection of magnate power, not against accusations of abuse, but against attempts at limiting it can be seen in
the bakufu's response to peasants who complained in Akita in 1617 that they were unable to shoulder the burden of more than
10 koku of the headman's total of 16 exempted koku. They wanted to cap the exemption at 10 koku. The intendant, however,
ruled that headmen had a right to as much as 20 corvée-exempt koku even in villages of only 100 to 150 koku. With only 16
koku of the headman to take care of, he added, the petitioners should count themselves lucky. The same standard was upheld
in 1620 in another bakufu case.

In the early 1640s the bakufu abandoned its unquestioning support of headman authority and privilege. By then
embezzlement was clearly illegal and punished. Some headmen even wound up in prison.[56] To prevent abuses, the
headman's monopolistic power was curtailed de jure. The first directives concerning intramural calculation and apportioning
of tribute were issued in 1642 and 1644 as part of the bakufu's legislative response to the famines of the Kan'ei period. No
doubt pressure had also been exerted by numerous petitions and suits—altogether a clear indication of the bakufu's initial
reluctance or inability to regulate village affairs. The frequent settlement of these disputes without recourse to bakufu
authority or through mediation of other headmen, as well as the protracted nature of some of them, further point to the same
conclusion. The regulations of 1642 and 1644 stipulated that tribute allocations must be made public, and that all (taka

[56] For examples dating from 1643 and 1648-50, see ibid., 50.

― 100 ―

holding) peasants must participate in decision making.[57] This, Saito[*] concludes, amounted to the institutionalization of the
famous murauke system, whereby the whole village, and not simply the headman, was in a way subcontracted as the legal
agency for allocating tribute.

Two points should be made here. This is first of all a good illustration of Tokugawa legal expansion as a response to ad hoc
problems. What James White argues with regard to peasant uprisings also applies to their petitions and protests. Peasants
never succeeded or even wanted to overthrow the system, yet their petitions and protests were not without effect. Peasant
agitation propelled legislative production, and insofar as the bakufu's legislative practice as such was never put into question
its authority to rule was thereby legitimized in an indirect but significant and real way.[58] Moreover, as this shift in bakufu
policy was the result of local struggles, it consolidated a new power distribution and political consciousness in the village and
perhaps produced a new "civic" identity among certain sectors of the peasantry. This, at least, was the potential of this policy,
even though it did not affect all community members; nor was it implemented in all communities at the same time. Much
later, other peasants would agitate to participate in decisions regarding tax allocation.[59]

Some of the special powers and privileges of the dogo[*] were being limited by other bakufu legal judgments. Thus, in the early
1650s in a village of a bannerman the question of corvée allocation to noncorvée households was raised; it was ruled that
"privileges associated with past status are acknowledged, but times have changed (jidai onaji de arazu ) and henceforward
corvée [and other burdens] will be allocated equally according to assessed value (taka: byodo[*]no takawari ni )."[60]

The drive to apportion corvée and duties per assessed value rather than per household cut into the traditional power of
dogo[*] peasants. That these magnates resisted this as best as they could is illustrated by

[57] The decisions are quoted in ibid., 51. For the text of these laws, see Tokugawa kinreiko[*] zenshu[*] , ed. Ishii Ryosuke[*] , 5
vols. (Sobunsha[*] , 1959-61), 5:155 (no. 2784, art. 11) (hereafter TKKz); and Kinsei hoseishiryo[*]sosho[*] , also edited by Ishii
Ryosuke[*] , 3 vols. (Sobunsha[*] , 1959), 2;156 (no. 280, arts. 9, 15).

[58] James White, "State Growth and Popular Protest," Journal of Japanese Studies 14, no. 1 (1988): 7 n. 27, 17, 20-21.

[59] Smith presents a dramatic example: in the town of Kurashiki some two thousand families broke the monopoly of thirteen
titled peasants in 1790 (Agrarian Origins , 183-87, esp. 186).

[60] Saito[*] Yoshiyuki, "Kinsei shoki no nomin[*] toso[*] ," 53.

― 101 ―

two cases analyzed earlier. In the 1650 suit in Kami-Kawarabayashi village (see table 7) the headman capitulated, but only
temporarily: shortly thereafter he reverted to the old system, a move that resulted in a new suit by the peasants. The series of
suits and petitions spread over forty years in the village of Shimo-Okamoto in Utsunomiya, the second case cited in this
section, started with the attempts to limit the headman's privilege in the form of exemption from corvée and wound up with a
petition to switch to the taka-based system in 1678.

Macro Power 2: Dogo[*] Power and the Dominant Class

The mid seventeenth century was not the first time dogo[*] struggled against extramural interference. Two generations earlier,
around the turn of the century, they had even organized armed resistance against land surveys in a number of places. In Higo
province (Kumamoto) they mobilized twenty thousand peasants in opposition to Sassa Narimasa, who was in charge of the
land survey. Similar rebellions took place in the north, where dogo[*] power was strongest. As Mizubayashi points out (and as I
shall discuss shortly), this contrasts with the economically more developed Kinai region, where since the time of Oda
Nobunaga (1534-82) this dogo[*] class had spontaneously begun to split into those who sought fortune primarily in military
campaigns and those who sought it on the land. In the "backward" areas, however, the rural magnates clearly felt threatened
by the warriors and their land surveys.[61]

This struggle they lost. The rural magnates could stop neither the warriors nor the surveys, but they survived quite well the
restructuring of political power that was taking place over their heads. It is important to clarify the changing structural
position of this dogo[*] class in order to understand how the dogo[*] made the transition into the Tokugawa period, when they
functioned as the overlords' local agents for over a century, until the bakufu dissolved their power in the first decades of the
eighteenth century in part as a response to abuses.

Under warlords of the sixteenth century like the Rokkaku, the traditional control the dogo[*] , as often semimilitary rural
magnates and

[61] Mizubayashi, Hokensei[*] , 145. On dogo[*] uprisings, see Donald Burton, "Peasant Struggle in Japan, 1590-1760," Journal of
Peasant History 5, no. 2 (1978): 144-47.
 ― 102 ―

fief holders, had over their retainer/servants (hikan ) and dependents remained firm. When they and their followers were
called upon for military or other tasks by these warlords, they were recruited in toto as small coherent units. Oda Nobunaga,
however, started to sever the dogo's[*] ties to their subordinates when he used his formidable authority to bypass these local
bosses by recruiting some of their dependents directly. Nobunaga presented projects such as the building of Azuchi castle in
1576, employing this labor force as kuniyaku, or "national service," and he indemnified the masters of these recruits with
some tax exemption.

A second way in which the ties between dogo[*] and their followers were dissolved was through transfers of warlords. When
Nobunaga assigned a new territory to a daimyo, he ordered him to take his dogo[*] along, as he did when he sent the Maeda to
Noto.[62] These local magnates thus lost their power base of land and followers/servants and, as full-time retainers, became
completely dependent upon their lords.

Many dogo[*] were jizamurai, or landed men of arms; in the Kinai region, however, many of them had considerably reduced
the portion of their holdings that they worked themselves. Thus loosening their ties to the land on their own initiative, many
had joined Nobunaga's armies without giving up their landed property, expecting greater gains from military campaigns, in
the form of increased holdings, than from working or managing the land they owned. Hideyoshi, as is well known,
institutionalized this initial voluntary and incomplete removal of armed men from the land through his Sword Hunt Edict
(1588) and his "Status Regulations" (1591).[63] Henceforward armed men would be clearly distinct from peasants. This had a
profound effect on the social position of these landed samurai or armed rural magnates, splitting this class into two since it
forced them to choose either to take the warrior road or to remain in their mini-dominions working the land. Under Nobunaga
such a choice had been neither total nor final; under Hideyoshi it was both.

It is important to stress that either way, as new, full-fledged retainers

[62] Mizubayashi, Hokensei[*] , 109. For a detailed discussion of Maeda's case, see Brown, Central Authority , chap. 5.

[63] For the text of these decrees, see Tsunoda Ryusaku, William Theodore de Bary, and Donald Keene, comps., Sources of
Japanese Tradition , 2 vols. (New York: Columbia University Press, 1964), 1: 319-22. The reason for quotation marks around
Status Regulations will be explained later.

― 103 ―

without land in the castle towns or as rural magnates without arms in the country, they did not become men without power in
the new order. This class was reconstituted at the lower echelons of the ruling hierarchy, where it performed a number of
essential functions with or without arms. The fact that some were now legally samurai and others were not should not
obfuscate the fact that the new rulers encompassed nonsamurai and the ruled included peasants with a samurai background.
Without a significant component of nonwarriors in the ruling cadres the system would have been unable to function, since
almost all warriors were now separated from their subjects. Peasants were thus not only tribute producers, separated socially
and geographically from the samurai; they were also essential to the operation of the system, in times of war, as we have seen,
and in times of peace.

One can distinguish three ways in which the dogo[*] were incorporated into the new power structure. Some became district
intendants (daikan ) or their assistants (tedai ), others became village group headmen (ojoya[*] ) in charge of a number of
villages, and still others turned into a kind of contractor for "public works," especially land reclamation projects. We shall first
take a look at the third group. Our example is from Kita-Saku district, a few kilometers east of Makibuse (see map 2).

As is well known, the seventeenth century witnessed the greatest boom in land reclamation in premodern Japan. Between 930
and 1450 Japan's arable increased by 10 percent. Another 70 percent was added by 1600. Equating the arable of 1450 with the
value 100, it was 90 in 930, 100 in 1450, and 173 in 1600; then it almost doubled, to 314, by 1720 (and it was only 322 in 1874),
as table 10 shows. Unlike the land reclamation of the eighteenth century, most of which was financed by merchant capital, the
much larger projects of the seventeenth century were initiated and funded by the daimyo and the shogun. Often dogo[*] and
their bands of dependents were employed in these projects. This was the case with four such shinden (new paddies) projects
in Kita-Saku.[64]

The most famous of the four shinden is Gorobe-shinden, the others being Mikage-, Shiosawa-, and Yaehara-shinden. They were
developed between 1630 and 1662. The irrigation networks for the four projects, totaling some 175 km, took twenty years to
build. The ducts that brought

[64] Oishi[*] , Kinsei sonraku no kozo[*] , chap. 3.

― 104 ―

Table 10.
Total Acreage in Japan, 930-1874

Acreage (cho[*] Percentage of 1450

) figure

930 862,000 91.1%

1450 946,000 100.0

1600 1,635,000 172.8

1720 2,970,000 313.9

1874 3,050,000 322.4

SOURCE: Mizubayashi, Hokensei[*] , table 17 (p. 301).

NOTE: 1 cho[*] =9,917 m[2] .

the water to Gorobe-shinden alone were about 20 km long (including some 2 km hewn in rock, nine tunnels totaling 1 km, and
a duct crossing a river) and passed through nine villages.[65] All four projects were undertaken by descendants of local vassal
and subvassal houses of Takeda Shingen (1521-72). The warrior ancestors of these developers had continually sided with
losing parties in the wars of the late sixteenth century. For instance, the Ichikawa house, to which Gorobe belonged, was first a
subvassal of the Uesugi, then of Takeda Shingen, and subsequently of Oda Nobunaga and the Hojo[*] when its holdings were
greatly reduced. The remaining Ichikawa property, mountain forest and some undeveloped grassland in the Kita-Saku plain,
was finally divided among followers of the house (between fifty and one hundred). Gorobe, reluctant to leave the area, refused
twice to enter Tokugawa Ieyasu's service; instead, he took the opportunity given by Ieyasu's decision to develop new arable. In
return, Ieyasu acknowledged Gorobe's jurisdiction over his numerous hereditary vassal-servants (fudai and genin). In 1642,
Gorobe received from the lord of Komoro domain 150 koku of the new fields as his "fief." Then the fifty or so fudai literally
came out of the woodwork, writing to Gorobe, "We are sons of fudai

[65] The calculations are based on data given by Saito[*] Yoichi[*] in his Gorobe-shinden to hisabetsu buraku (San'ichi shobo[*] ,
1987), 1, 2, 56-57. The most detailed history of a shinden mura in English is the work of an anthropologist, Jennifer Robertson's
Native and Newcomer: Making and Remaking a Japanese City (Berkeley: University of California Press, 1991), 76-103, 136-143.
The remaining information on Gorobe-shinden and the developers of the other shin-den that follows is again taken from
Oishi[*] , Kinsei sonraku no kozo[*] , chap. 3.

― 105 ―

who for generations have been in your family and fought together until your father's generation, but then we moved to the
mountains not so long ago because there was no land left in the family. Now, however, that you have received 150 koku, we
shall remain loyal in any battle that may lay ahead." They worked for Gorobe again, first as vassals, then for wages, and they
were finally set free in 1713. Meanwhile, the Ichikawa had left the area in 1670 and become absentee landlords.

The founding families of the other new villages have similar histories. Kashiwagi Koemon, founder of Mikage-shinden, having
also sided with losing lords, went to Suruga for a while to take an office under Tokugawa Ieyasu, but soon he was back in Kita-
Saku in charge of bringing new land under cultivation. He eventually received eighty-three koku in tax-exempt new fields,
enjoyed labor services from the new settlers, and wielded considerable power in the new village. He passed on to his
descendants the right (exercised as late as 1819!) to veto any election of village officials by the titled peasants and the right to
keep all the records, which he lent to the officials at his own discretion. For the first fifteen years there was not even a village
headman; when one was finally appointed, in 1665, he came from a branch house, and the office remained in that family for
half a century.

A second road taken by dogo[*] in the seventeenth century, more frequently by dogo[*] remaining on the land, led to the new
rural office of ojoya[*] (village group headman). They thus became charged mainly with collecting tribute from a number of
villages, their area of jurisdiction often coinciding with the pre-Tokugawa districts (go[*] ).[66] In this way, they supplemented
their economically based and informal patriarchal power as local magnates with a new official, administrative dimension. Just
as the sixteenth-century warlords used "the state" (which they built) to continue furthering their interests, the landed dogo[*]
added the same surplus political value to their traditional authority.

It should be remembered that the purpose of Hideyoshi's land surveys, conducted and registered at the village level, was not
tribute exaction but the mobilization of a national army. For taxation purposes, Hideyoshi relied on the old local
administrative units (often the go[*] , encompassing a number of villages), which often became the juris-

[66] For this paragraph and the following one, see Mizubayashi, Hokensei[*] , 136-37. Brown also stresses the indispensability
of nonsamurai at the district and intendant level in Kaga domain (Central Authority , 114-41, 193, 205).

― 106 ―

diction of a village group headman. Thus, one of the two dogo[*] from Toyota village, mentioned earlier, filled the post of
village group headman for thirteen villages.[67]
The village group headmen, as stipended agents for the city-bound warriors, were an integral part of the dominant class in not
only structural but also economic terms, for in both status and privilege they were considerably removed from the ruled in the
villages. Located administratively beyond the village and performing their corvée or service to the overlords as officeholders,
in their village of residence they were exempt from the obligations shouldered by regular villagers.

As developers of new paddies, the dogo[*] expanded the overlords' tribute base, and as village group headmen they channeled
this tribute upward. In addition, they functioned at the next higher administrative level also, as district intendants or, more
often, as their representatives or assistants. One usually thinks of intendants as samurai, but especially in the early decades of
the Tokugawa period a good number of them were dogo[*] . Why?

The indispensability of the disarmed fraction of the dogo[*] class to the new military rulers was not only a consequence of the
concentration of the samurai in urban centers. It was also the result, in some important cases, of massive daimyo transfers to
new territories. After the Hojo's[*] defeat in 1590, Tokugawa Ieyasu was ordered by Hideyoshi to leave the five provinces he
controlled and take charge of the eight provinces in the Kanto[*] plain, most of which constituted the former Hojo[*] domain.
This transfer affected the dogo[*] of Ieyasu's old and new domains in different ways.

In his old domain, the transfer put into practice, in one stroke, the separation of warriors from peasants, splitting the dogo[*]
class into those who took their arms and followed Ieyasu to the Kanto[*] and those who stayed behind. It was a move that also
strengthened the integration of his vassal band, which, cut off from its economic base, now became totally dependent on its
lord. In his new domain, however, the local dogo[*] class became indispensable for ruling the countryside, since Ieyasu
concentrated most of his army in and around Edo, granting many (but not all) rural administrative posts—hatamoto, or
bannerman, fiefs in the "Tokugawa houseland" and daikan or gundai intendant offices for

[67] For the system of village group headmen in the Ueda domain, see Bix, Peasant Protest , 62-64.

― 107 ―

the remaining portion of the Tokugawa domain—to only a small fraction of the thirty-five thousand warriors under his direct
command.[68]

The figures are quite striking. Forty percent of the Tokugawa domain was entrusted to some twenty-two hundred fief-holding
bannermen, each with his own retainers, but the overwhelming majority of the latter were stationed in Edo, not in the fiefs
themselves. For instance, the Sengoku bannerman, who in the 1790s held a fief of two thousand koku (in eight villages) in
Shinano and an additional seven hundred koku in two other provinces, had thirty-two retainers in Edo but only four in
Shinano. His Shinano office was headed by someone with a dogo[*] background who also relied on locals: two wariban (the
equivalent of village group headman) and the eight village headmen.[69] The remaining 60 percent of the Tokugawa domain
was ruled by a mere forty intendants. They were assisted by helpers, whose total number in 1839 was less than a thousand,
slightly over half of them based in Edo. Under Hideyoshi the intendants had been recruited from the ranks of vassals and
dogo[*] . In the early Tokugawa period they were former retainers from defeated daimyo (the Imagawa, the Takeda, the Hojo[*]
) or local dogo[*] who had served under them. This local entrenchment, perpetuated through heredity, led to numerous abuses
that were further facilitated by the way these intendants were paid, namely, by allowances they themselves took from the
tribute they collected. In the 1680s the majority of the intendants (thirty-five) were purged, and after 1725 intendants had to
forward the total amount of tribute to the center, which substituted fixed stipends for their allowances. In addition, their posts
were now increasingly being filled by staffers from

[68] Mizubayashi, Hokensei[*] , 143. The figure 35,000 comes from Conrad Totman, Politics in the Tokugawa Bakufu , 1600-1843
(Cambridge: Harvard University Press, 1967), 135. By the end of the eighteenth century only 44 percent, or 2,264, of the
bannermen held fiefs, with a total kokudaka of 2.6 million koku (see KDJ 9:381, s.v. "chigyoseido[*] "; for similar figures, see
also Totman, Politics , 135), spread over forty provinces, and governed from 3,677 local offices because many fiefs consisted of
separate territories, 43 percent of these fiefs having parcels in from two to six different provinces. The portion of the
Tokugawa domain that was not parceled out in fiefs (each less than 10,000 koku) to bannermen (3.2 million to 4.2 million koku
—the amount fluctuates over time) was Tokugawa "houseland" (tenryo[*] ), administered by intendants, called daikan (50,000
to 100,000 koku) or gundai (over 100,000 koku), whose number stabilized at around forty (KDJ 4:1040-44; see also Totman,
Politics , 66-85).

[69] Mizubayashi, Hokensei[*] , 268-69.

― 108 ―

the bakufu's central Finance Commission. These purges and reforms signaled the end of those dogo[*] who had survived as
intendants.

Many of the intendants who were not dogo[*] but samurai took up residence in Edo and, like the bannermen, maintained only
skeletal staffs in their country and Edo offices. The two intendant offices in Shinano—Nakanojo[*] (69,000 koku) and Nakano
(54,000 koku)—had staffs of only three and four men, respectively, while their Edo offices had nine and eight. (Shinano also
counted two branch offices, one in Mikage with a staff of three and one in the highway station of Oiwake with only one post.)
Their small staffs consisted of helpers called tedai and tetsuki . The latter, shogunal retainers, were introduced only in the
1790s; the former were local peasants.[70] The situation was slightly different in the domains, for there the offices were often
overstaffed. The small domain of Tanoguchi (12,000 koku) in Saku district, for instance, counted only twenty-five villages, yet
there were three intendants.[71]

Given the sparse use of samurai for rural administrative purposes, these intendants and fief-holding bannermen had to be
assisted by locals both in their small territorial offices and in the villages, where they were assisted by village group headmen
and village headmen. Moreover, although one might expect responsibility for the expenses, if not for the personnel then at
least for the upkeep of these offices, to have been assumed by the mighty bakufu, that was not the case. Routine maintenance
expenses were shouldered by villagers, who were also charged with the corvée duties of cleaning, changing the paper of
sliding doors, operating the hot bath, and so on. They also had to provide messengers to communicate between the rural and
Edo offices of these administrators.[72]

[70] This information on the daikan and gundai was taken from KDJ 4:1040-44; see also Mizubayashi, Hokensei[*] , 270-71, and
Totman, Politics , 66-85. In 1853 the bakufu employed 640 tedai (291 of these in Edo) and only 225 tetsuki (146 in Edo). Thus,
when in the 1790s the bakufu created the new post of tetsuki, to be filled by retainers, next to the identical post of tedai, filled
by commoners, not only were its numbers significantly lower than that of the tedai but fewer than 100 retainers were
deployed in the countryside (KDJ 9:885, s.v. "tedai").

[71] NAK-KS2 (1): 486. One of these intendants, appointed in 1851, kept a detailed list of all incoming and outgoing presents for
his first year in office: he himself gave gifts on 188 occasions, 4 times to headmen from five villages, and received gifts on 596
occasions! (Ichikawa Yuichiro[*] , Saku chiho[*] , 220-47).

[72] KDJ 8:711, s.v. "daikansho"; 7:940, s.v. "jinya."

― 109 ―

The system was extremely cost-effective and user-friendly for the samurai because the peasants, both elites and others, not
only paid for it but ran it as well. The analogy of firing squad victims who first have to pay for the bullets and dig their own
grave comes to mind. This analogy might seem far-fetched, bur in Tokugawa times villages that for one reason or another had
a member in prison had to pay the "boarding" costs.[73] Notwithstanding Confucian rhetoric about peasants' being "the basis of
the country," the samurai were unwilling to dirty their hands through contact with peasants, not even for the purpose of
levying tribute.

Samurai rural administrators, a rather rare breed in the first place, ranked low in the prestige and stipend hierarchy. Ogyu[*]
Sorai (1666-1728) wrote that "the intendants' sole preoccupation being to levy tribute, they appointed people just because they
could write and count and they themselves were of low status ... which was the reason why no samurai with a pedigree
became intendants, and this in turn was why no samurai, including the upper ranks of counselors, had any real sense of the
people."[74] The fundamental division between manual and mental labor stressed by Karl Marx and Max Weber[75] is basic to
the separation between warriors and peasants, although in Tokugawa Japan, as Ando[*] Shoeki[*] noted in the eighteenth
century, it was more a question of manual versus no labor.[76]

The nonsamuraized fraction of the pre-Tokugawa dogo[*] class was thus reconstituted as an essential lower fraction of the
dominant class and maintained local power well into the eighteenth century. Its demise, however, is not to be attributed only
to bakufu initiative against abuses that diminished the flow of tribute to the upper fractions of the domi-

[73] Those confined to an intendant's jail whose provenance was known (i.e., those who were not vagrants, mushuku ) had to
arrange for their own food, make payments for their expenses to the prison guard, and provide even their own lamp oil. The
situation was slightly different in Edo (see Hiramatsu, Kinsei keiji soshoho[*] , 936 ).

[74] Ogyu[*] Sorai, Sorai sensei tomonsho[*] (1727), in Ogyu[*]Sorai , ed. Bito[*] Masahide, Nihon no meicho, 16 (Chuokoronsha[*]
, 1983), 319, 320-21.

[75] Karl Marx and Frederick Engels, The German Ideology (New York: International Publishers, 1970), 51; Max Weber,
Economy and Society , ed. Guenter Roth and Claus Wittich, 2 vols. (Berkeley: University of California Press, 1978), 2:936.

[76] Maruyama Masao, Studies in the Intellectual History of Tokugawa Japan (Princeton: Princeton University Press, 1974), 252-
53.

― 110 ―

nant class, a story of rationalization and bureaucratization; there was also a political side to this story.

Protests and petitions had led to limitations on headman authority in the first decades of the Tokugawa period. Then these
dogo-headmen[*] were forced to share power with a class of new landholders, often former dependents. Moreover, since the
late seventeenth century their authority had been further checked by a representative from the other peasants, a
hyakushodai[*] (a new village office to be discussed below). The position of village group headman was abolished in the bakufu
domain in 1713 as a response to a violent peasant rebellion in the Murakami fief in Echigo. Similar rebellions against abuses
by village group headmen followed in daimyo domains (Kanazawa, 1712; Hiroshima, 1718; Aizu, 1720; Kurume, 1754;
Fukuyama, 1789), leading there also to the abolishment of the office. The bakufu and domains, however, could not long do
without some conduit between their regional offices and the villages, hence the on-again, off-again status of the office of the
village group headman (on again in the bakufu in 1734, on again in Fukuyama in 1791 and off again in 1823). In the bakufu
the office was now staffed on a rotation basis by headmen from the villages under its jurisdiction.[77]

Knowledge and Power: The Struggle for Information

Data on land, population, and tribute were essential for the exercise of domination, which for Tokugawa Japan is a more
accurate term than administration . Once established, power needed knowledge more than the sword to carry on. This
information was organized around the kokudaka system, a thorough quantification of the country's economic potential
established by Hideyoshi for military purposes of a national order. The daimyo who gathered these data for Hideyoshi
benefited too, because the data gave them a more precise knowledge of their own power base, which was also expanded, since
for the first time dry fields, commons, and even homesteads were rated and taxed. These data, while official because collected
by and for those in power, was not public, precisely because of its link to power. Access to this information was

[77] KDJ 2:613, s.v. "ojoya[*] ."

― 111 ―

obtained through struggle, that is to say, through the application of a kind of power that forced the authorities to yield it.

Village suits against headmen's abuses in tribute allocation were struggles for access to kokudaka data. The plaintiffs had in
mind a specific use for these data, different from the one the overlords had in mind when they created them: they wanted the
data as a leverage for justice, for a fair intravillage distribution of corvée and tribute. The logic behind the strategy to obtain it,
however, was similar. Through their land surveys, Hideyoshi and the Tokugawa shogun after him made official and created
for themselves knowledge that thus far had been the private (and less accurate) knowledge of the daimyo and thereby
increased their power over them. They were able to force the daimyo to create and render this knowledge in part because
they had converted their warlord superpower into public authority (kogi[*] ) through a number of officializing strategies.[78]
Hideyoshi and the shogun wielded a higher authority, which it was in the best interest of the daimyo not to resist. The
peasants achieved the same result in a somewhat similar way by wresting privately held knowledge from their village
superiors. For this the peasants needed a higher authority, which they gained when, after the 1640s, the bakufu supported the
peasants in their demand for open village ledgers.

The story of the struggle for making official knowledge public does not end there, however, but the principle was established
in the 1640s. I trace this struggle until the end of the eighteenth century in the following section. Here I shall first return
briefly to the drive started in the mid seventeenth century to base village corvée allocation on a taka rather than a household
basis. This was possible only because knowledge of each field's kokudaka was forced into the open, and resistance to it was
strongest in dogo-ruled[*] villages because there the plot discrepancies were the greatest. In such villages, if corvée were
allocated according to holding size rather than divided evenly among all households, the elite would lose proportionately
more than they would in places where economic discrepancies were less pronounced. Second, we shall examine the bakufu's
endorsement of publicizing tax rates in the 1640s. Although implementation of the decrees to that effect was uneven, that

[78] See my Tokugawa Ideology , chap. 2.

― 112 ―

decade constitutes an important turning point with regard to the issue of accountability.

Knowledge of the kokudaka provided peasants with a standard of value they could use in various ways, depending on local
circumstances, for quantifying things other than tribute. Sometimes complicated formulae that obviously were the outcome of
protracted wranglings and negotiations, to which small landholders must have been partners, were worked out. Thus, in
Kitasawa village (Kita-Saku) the following agreement using different measuring units for various items was reached in 1687,
through the mediation of the local temple, it should be pointed out.[79] The preamble clearly states that the process involved
"large and small peasants." The agreement covered both kinds of corvée, that owed to the bakufu and that due to the village,
and stipulated the following:

Tribute from the mountains: manpower to be assigned on the basis

of horse ownership—one man per horse;

Bakufu-ordered [extramural] corvée: labor and number of days to be assigned per taka; [bakufu-ordered] intramural corvée:
labor and number of days to be assigned per household for both large and small peasants;
Village fees: to be determined one-half per taka, one-half per household;
Sake portion of village fees: two-thirds per taka, one-third per household.

Such a formula of corvée allocation necessitated knowledge of each holding's kokudaka. What did the peasants know about
the share of the yearly village tribute they owed the overlord?

The amount of yearly tribute, determined by the overlords, was communicated to each village head by the intendant in a
document most commonly called the nenguwaritsukejo[*] (literally, "yearly tribute [nengu ] rate letter"). From 1591 until
around 1635, when they were standardized, in most bakufu territories most of these documents were very short, notes really.
[80] They clearly reflect the sole concern of the

[79] Saito[*] Yoshiyuki, "Kinsei shoki no nomin[*] toso[*] ," 56.

[80] For the discussion that follows, see Arai Eiji, "Nengu waritsukejo[*] no seiritsu," in Kinsei no jikata-machikata monjo , ed.
Nihon komonjo gakkai, Nihon komonjogaku ronshu[*] , 12 (Kinsei 2) (Yoshikawa kobunkan[*] , 1987), 228-55 (hereafter NKR
12).

― 113 ―

overlords—tribute—in that they simply list the village taka, the assessment rate (not always), the amount of rice tribute in
bushels (koku, sometimes in bales, hyo[*] , one koku being the equivalent of 2.857 hyo[*] ), and the date the tribute was due.
Adjustments for bad years were usually made by modifying these total figures.

This skeletal system worked because the headmen and the overlords had information that showed the further breakdown of
the total village figure into amounts per holding, but this information was not to be found in most of these documents. The few
exceptions list in detail and separately paddy fields, dry fields, and homesteads according to the grade of the arable; these
were usually issued the year after a survey was taken. Why is it, then, that around 1635 in most areas of the bakufu territory
and in the 1660s and 1670s in a few remaining pockets these forms became standardized, now always giving the detailed
information that was only very rarely available earlier?

A tighter central control over intendants was not the only reason. Political pressure was also applied from below, not,
obviously, from village headmen, who could thwart the system to their own advantage, but from the peasants. There were too
many complaints about abuses that occurred because peasants were kept in the dark concerning the fairness of their share of
village tribute.

As a first response to these political problems, starting around 1635 the intendants provided all the necessary detail in their
yearly tribute letters, thereby binding headmen to extract such and such an amount from such and such a plot. This, of course,
still did not make the headman accountable to the tribute payers, which led to numerous disputes. Indeed, the shogunal
directives of the 1640s ordering intendants to make sure that headmen made the yearly tax list available to all the "small
peasants" for their approval invariably mention that this was being required "in order to avoid quarrels in the villages."[81]

For the bakufu, as we have seen, "small peasants" included all taka holders below the titled peasants. These small landholders
were thus granted new public political importance within the village: they yearly ratified the tax list. It is possible that the few
regions that succeeded in postponing the implementation of these measures until the 1660s and 1670s (Sagami's Osumi[*] and
Musashi's Okutama) were areas where

[81] TKKz, 4:122 (no. 2105, art. 14).

― 114 ―

dogo[*] power was stronger. In other words, this development indicates that the "small peasants" were not fully incorporated
into the murauke system as far as trans-village tribute was concerned until the 1680s, that is, when villages as a whole, and not
simply headmen, became responsible for the delivery of tribute.

The kokudaka system was thus manipulated as a standard of fairness by the people below in a class struggle, not against
samurai, but against village leaders. This worked to the bakufu's benefit in two ways. Inasmuch as its village tribute system
was widely used by the tribute producers as an ultimate scale for communal fairness , the system itself was safe from being
subjected to other standards of justice; thus, intra-peasant practice legitimized overlord domination. Hints that appeals in
petitions to a broader concept of Justice (which were common in premodern Europe) transcended the standard itself can only
be found in vague references to notions of benevolent government or the need for peasants to reproduce themselves, to
survive. These notions were marshaled in peasant protests, but they did not question the system as such.[82] Moreover, the
bakufu, through its decision (under pressure) to widen access to the system for more and more peasants beyond the headman,
first the kumi heads as village officials and then all the peasants as certifiers of the tribute allocation, created a new village
practice that strengthened overlord legitimacy among a widening constituency of subjects.

The expanding political empowerment of villagers discussed thus far was limited to taka holders. Toward the end of the
seventeenth century, however, the number of non-taka holders increased. They also had a stake in access to information first
monopolized by headmen and then shared by the "small peasants," and this for two reasons. First of all, tenants were
interested in knowing the breakdown of tribute allocation because the rent they paid to landlords included the amount of
tribute due on the parcel they rented, tribute for which the landlord was responsible. Correct information about this tribute
allowed tenant peasants to judge the fairness of their rent. Moreover, as we have seen, these peasants shared intramural
corvée and fees, and they could use the taka system to argue for a fairer allocation of these duties.

In principle this information was made available by law, starting in

[82] Stephen Vlastos discusses the arguments of these appeals in his Peasant Protests , 15-17.

― 115 ―

the 1640s. A directive of 1642/8/10 stipulated that "for the calculation of the yearly tribute, etc. the intendant has to make
village headmen meet with the small peasants and have them affix their seal to the tax roster of that year in order to avoid
headmen's doing injustice to the small peasants."[83] In another directive, dated 1644/1/11, sent to the intendants of the Kinai
and the Kanto[*] , the bakufu reiterated that "every year the tax rates (osamekatawaritsuke ) have to be shown to all the
peasants without exception (sobyakusho[*]nokorazu ) who have to affix their seal to it so that the quarrels of the past can be
avoided."[84]

Now, empirically minded scholars invariably raise the question of practice: was not the fact that these laws were reissued
again and again proof that they were a dead letter? Indeed, they were reissued, for example, in 1713/4 in a long directive
overhauling intendant governance that admits to faltering practice in recent years, quoted in part here:

The tribute roster that the intendant forwards every year to the village officials (murakata ) has to be made known
in detail to the peasants, small and large alike. This is a law from the past. In recent years, however, we have heard
that nobody has seen them except the headman, and so on, and, moreover, that the headman, and so on, do not
make the village budget available to the smallest peasants because there are unnecessary expenditures or illegal
matters. Henceforward, as stipulated in past laws, not only the tribute roster, which goes without saying, but also
the specific items of the village budget have to be made clearly known to the peasants large and small, who will
affix their seal to the budget. And every year, the intendant will examine the budget document of each village and
make inquiries of all the peasants of the village concerning details, this to avoid illegalities on the part of the
headman, and so on, and unnecessary expenditures on the part of the large and small peasants.[85]
Documentary confirmation of actual compliance is, obviously, only sporadic, but as we saw earlier, seals of more and more
peasants do appear on these documents, while they were absent at the beginning of the period. A more interesting question
than that of the impact of legislation on practice is the reverse one, namely, what kind of practice prompted this legislation. It
is clear that those who through these

[83] TKKz 5:155 (no. 2784). For the discussion that follows, see Otsuka[*] Eiji, "Nenguwaritsuke to mura nyuyo[*] no kokai[*] ni
tsuite," Shinano 43, no. 9 (1991): 1-24.

[84] TKKz 4:122 (no. 2105).

[85] Ibid., 134.

― 116 ―

directives became, so to speak, politically more enfranchised were the ones who had exerted the pressure.

This continuing tug of war between those who wanted information made public, in principle officially backed by the bakufu,
and those who resisted such requests resulted in more concrete legislation in 1745, 1750, and 1767. These directives required
that a short version of the tax rosters, written by the intendant and listing the total amount of tribute and the precise amount
due in each category of taxes, be posted on the public notice board or at the headman's gate. It was to remain there for the
whole year and then be exchanged for the next year's one at the intendant's office.[86]

Budget Control

Tokugawa villages were not only tribute units; they were also communities of people who did things together other than pay
tribute, and communities have common expenses. Control over decisions concerning the size of the community budget, what
it is to be used for, and how the burden is to be shared within the community translates into political power. And the degree to
which such control lies within or outside the community largely determines the degree of community autonomy.

There are two opposing stereotypes concerning the autonomy of Tokugawa villages. One stresses their autonomy, because
they were self-governing communities (kyodotai[*] ). The other stereotype maintains that villages were subject to overlord
authority in so many important ways that self-governance was a sham. To maintain the distinction between these two
interpretations in the discussion that follows, I use communities in reference to these social units insofar as they were
autonomous and villages insofar as they were constituted by the overlords. Much of this controversy has been a matter of
definitions and the political views that inform them or lessons one wishes to draw from them.[87] A discussion of local budgets
may throw some light on this

[86] Otsuka[*] , "Nenguwaritsuke," 6-7. Otsuka's[*] research suggests that these digests were posted, but most probably only for
a few weeks.

[87] For a good survey of the historiographical polemics surrounding this issue since the early twentieth century, see Uesugi
Mitsuhiko, "Kinsei sonrakuron: kinsei sonraku to 'jiji,'" Nihonshi kenkyu[*]no shinshiten , ed. Nihon rekishi gakkai (Yoshikawa
kobunkan[*] , 1986), 175-93.

― 117 ―

question of autonomy, which is usually argued from a legal and juridical angle, as we shall see in chapter 4.

The degree to which Tokugawa villages, sanctioned by the land surveys, were already also communities with communal
economic interests and expenditures varied widely. Whatever the case may have been, they were certainly tribute units. And
since tribute production and payment loomed so large in the life of these villages, this undoubtedly furthered their
development into some sort of communities if they were not already communities to begin with.

In Tokugawa Japan, decisions on sharing community expenses were similar to those concerning tribute, because tribute was
also a communal affair. To get a sense of the relative importance of community and tribute expenses, let us consider briefly
the main categories. Tribute consisted of (1) tribute in kind, mostly rice (nengu), but also various other contributions
(komononari ) and (2) corvée (yaku). Corvée for overlords did not always take the form of labor: it could include cash (busen)
or rice payments (kuchimai , bumai); kuniyaku, "national" corvée or service the daimyo required for duties owed to the
shogun; buyaku, which daimyo or intendants requested on their own authority; and a special category, sukego, or portage
labor performed by groups of villages attached to way stations on the highways. In addition, there was corvée connected with
the collection and transportation of the rice tax. Many of these services included materiel needed to perform the labor. A
number of other village expenses were indirectly the result of overlord demands, including remuneration for village officials;
their travel expenses on official trips; expenses for village meetings held in relation to the payment of taxes; and the cost of
brushes, ink, paper, and so on, to draw up official documents. All of these levies ultimately had to do with the reproduction of
the samurai ruling class.

Community expenses proper, on the other hand, had to do with corvée and contributions of cash or materiel needed for the
reproduction of the peasants (the ruling class's reproducers), in all spheres of their communal life (economic, political,
spiritual): roads, bridges, and irrigation systems had to be maintained, festivals held, donations to temples financed, and,
occasionally, the expenses of suits (not lawyer costs but travel expenses) shouldered, interest on loans paid, and so on. Aside
from the important political question—how this burden was spread

― 118 ―

throughout the village—attention needs to be drawn to a number of other points.

First, the combined village and community expenses were high, and they increased considerably in the nineteenth century,
when many financially strapped overlords, through sheer neglect, shifted infrastructural maintenance costs to the villages. For
instance, for the village of Kichijoji[*] , outside of Edo, which had no particularly heavy way station portage duties or irrigation
systems to maintain, in 1845 these expenses amounted to 28 percent of its tribute (nengu). For Hara village (Saku district,
Shinano), which had way-station and dike maintenance duties, between 1734 and 1841 the average expenses amounted to 44
percent; in some years it was go percent or even a full 100 percent.[88] Second, often only a tiny fraction of these total expenses
were community expenses proper; in Hara village in 1734, expenses were virtually limited to festivals and religious occasions.
Third, since the bulk of their expenses were direct or indirect overlord levies, the villagers had little control over the size and
allocation of the expenses they had to share. Furthermore, overlords could and did increase these levies at will, while they
were more reluctant to alter the yearly rice tribute. These structural constraints put severe limits on village autonomy.

Thus far I have avoided talking of budgets because Tokugawa villages did not have what we today understand by that term.
Our notion of budget is forward-looking in that a budget is an itemized allocation of a known or projected amount of funds.
Tokugawa village budgets were different, more like retrospective expense accounts drawn up at the end of the year, as
itemized lists of expenses incurred during the year. An official copy of the budget was forwarded to the intendant at the
beginning of the next year; the original draft, often at variance with the official version, remained in the village. The expenses
were, of course, paid as they occurred; often, however, they were paid not by the villagers but by the headman or another
better-off member of the com-

[88] For Kichijoji[*] , see NRT 3:441; the figure for Hara was calculated on the data provided by Kodama Kota[*] , "Kinsei ni
okeru mura no zaisei," NKR 12:353. In 1840, Hara village (79 households, 72 of which belonged to titled peasants; total
population 325; 470 koku) paid 155 koku tribute rice, and its village expenses, computed in koku at the then going price for
rice, came to 131 koku, or the equivalent of 81 percent of the rice tribute; in 1845 the two amounts were equal (155.8 and
154.6, respectively, which Kodama wrongly reported as 44 percent instead of 99 percent) (ibid., 346, 353).

― 119 ―

munity, who advanced the cash. Then at the end of the year, sometimes to the surprise of the villagers, the debt was divided up
among them according to one formula or another, and each household paid its share to the moneylender, usually with interest
and often at outrageous rates of 30-40 percent.[89] One of the reasons for this arrangement was the lack of independent village
income from common property, which was the case in corporate villages especially in pre-Tokugawa times.

Villages and other communities have always had expenses, but written budgets like the ones described above, like so many
other practices of rural administration, seem to have appeared only in the mid 1640s. This was the time when limits were
being set to the headman's arbitrary powers. It was also the time when some of the corvée, such as the "national" corvée, was
being convened into cash or rice contributions.[90] In other words, in earlier decades village expenses probably were
predominantly in the form of labor. The headman, who was entitled to some of that corvée and as dogo[*] also received corvée
from a number of peasants as his dependents, did not always keep these two kinds of labor, official and private, apart.[91]
Remunerations for village officials in the form of labor or tax allowances easily led to abuses and disputes. This explains the
drive by villagers to substitute them with payments in rice and with salaries pegged to the total kokudaka of the village. In
addition, the line between corvée ordered by the overlord and that generated locally could easily be blurred to the benefit of
the decision makers in the village—and it was. Numerous complaints forced the bakufu to take action. The bakufu directive of
1713, quoted at length above, ordered that not only taxes but also village expenses be made public. This was not the first nor
the last legislation on this subject.

In 1642/5, three months before the bakufu issued the directive with regard to "calculation of the yearly tax, etc.," it had
identified this "etc." as "expenses related to the various corvées" in a similar way: the corvée was to be specified for the small
peasants in writing; the headman and kumi heads were to affix their seal to it; and official approval

[89] See KDJ 13:676, s.v. "murazaisei"; for village budgets in general, see also ibid., 683, s.v. "muranyuyocho[*] ." In the Kinai
region the interest rates were fixed at 15 percent in the late seventeenth century (see Fujiki Hisashi, "Ikoki[*] sonrakuron,"
214).

[90] Kanzaki Akitoshi, "Muranyuyocho[*] ni tsuite," NKR 12:415.

[91] In some places this practice persisted until the end of the seventeenth century. For three cases, of 1684, 1693, and 1694, see
ibid., 419-20.

― 120 ―

by seal of the intendant's helper (tedai) was to be secured.[92] A year later all peasants ' seals were required.[93]

With such vague specifications, and with no sanctions provided for noncompliance, abuses at the intendant and village level
continued, partly because tribute and corvée were not differentiated in these village budgets. In 1666/4 it was ordered that two
lists be drawn up, one specifying each household's tribute, the other its corvée.[94] In 1694/1 the intendants themselves had to
provide two copies of the list of items to be used for drawing up the budgets, one for the headman and one for the peasants.[95]
During the Kyoho[*] Reforms of the 1730s the intendant's responsibility in this matter was stressed, but ultimately the burden
of enforcement was put back on the village officials in 1744, when for the first time punishments were specified for not
informing peasants of the amounts of tribute, corvée, and the village expenditures. Headmen would lose their office, and kumi
heads would be fined. If there was blatant greed, the former was to be expelled and his property confiscated; the latter was to
relinquish his office and be fined.[96] Village headmen had, of course, been punished in the past, but this was the first time that
these punishments were determined as a specific category.

Also in 1744 the bakufu clarified the formulae for calculating tribute and village expenses:
 corvée for the shogun and the intendants plus village expenses and expenses related to suits and disputes, and so
on, have to be computed per taka; this holds also for outsiders [with fields in the village];
dwellers in the mountains, the wilds, bays or salt producing shores, and so on, or places with many households
without taka or with

[92] Sugawara Kenji, "Muranyuyocho[*] no seiritsu: Kinsei muranyuyo[*] no kenkyu[*] josetsu," NKR 12:386. The discussion of
this legislation that follows relies on this article.

[93] TKKz 5:158 (no. 2788, art. 14). The same article appears again in a directive of 1644/1 (ibid., 4:122 [no. 2105, art. 15]). In
1652 noncompliance was seen as a major source of administrative problems (ibid., 125 [no. 2108, art. 7]).

[94] Ibid., 4:127 (no. 2110, art. 3).

[95] Ibid., 5:140 (no. 2772, art. 4).

[96] This stipulation is article 98 of Yoshimune's famous Kujikata osadamegaki (TKKk 4:210). For the additions to this text that
specify the punishments, see John C. Hall, "Japanese Feudal Laws III," 788. The German translation also lists the punishments:
Rudorff, "Tokugawa-Gesetz-Sammlung," 98-99.

― 121 ―

small taka, have to compute by the number of persons and include dependent servants (excluding wives and
children);
shares in the common use of mountain forests and moors have to be computed per taka; this applies also to
outsiders;
expenditures for festivals, donations for religious purposes, and so on, can be decided in common agreement as one
sees fit;
for all the above computations, however, prior custom can be followed in places that have been free from quarrels
in the past.[97]

The bakufu's hesitant approach toward intravillage political matters stands out clearly. Forced by circumstances to sanction a
method with its authority, the bakufu suggests the taka system as a safe guidepost but then lists a number of exceptions and
finally admits implicitly that the taka system should be used (only) if everything else fails.

Thus, in the mid seventeenth century the bakufu insisted that official "budgets" be drawn up to prevent village disputes about
corvée distribution. A century later it gave minimal guidelines on how to achieve fairness and prescribed serious penalties for
negligent village officials. In the late eighteenth and early nineteenth centuries legislation concerning village budgets was
limited to efforts to shrink expenses through sumptuary measures. Although the bakufu had repeatedly stipulated since the
early 1640s that "small peasants" should be involved in decisions concerning tribute and village expenses, it was only in 1744
that it issued guidelines to deal with disputes and stipulated penalties. During the century in between, peasants agitated
around these issues on their own and broadened their political base within the village. In this sense, the stronger stance taken
by the bakufu in 1744 sanctioned peasant gains.

"Representing" the Peasants

One of the most significant institutionalized intramural checks on village leadership, at least in principle, was the "peasant
representative," or hyakushodai[*] . He was to keep an eye on village governance for all the peasants, which in concrete terms
meant foremost on decisions made by the titled peasants with regard to tribute allocation and village expenses. This
institution should have been of obvious interest

[97] TKKk 2:67.

― 122 ―

for scholars exploring the question of village self-governance. Yet they have not paid much attention to it, focusing instead on
the question of village autonomy as reflected in village laws.

In an extended study of land ownership published some thirty-five years ago, Oishi[*] Shinzaburo[*] discussed this position
briefly as one of the three official village offices, murakata sanyaku (next to the headman and the kumi heads), the way it is
described in the famous late-eighteenth-century handbook for local administration, the Jikata hanreiroku .[98] The earliest
instance reported by Oishi[*] dates from 1714 (in Oiwake, Kita-Saku district). It is generally accepted that the office was
institutionalized around that time, although not universally: not all villages had peasant representatives in the second half of
the Tokugawa period.[99] "Institutionalization," therefore, depended on local conditions, that is, on whether or not overlords
acknowledged such an office or whether villages established one on their own. Villages here stands, however, for titled
peasants, who, for obvious reasons, were not eager to be monitored. Here again, therefore, the establishment of peasant
representatives was often a response to a local crisis.

In 1724 in Gorobe-shinden (Kita-Saku district), for example, when the headman's retirement, officially for health reasons,
caused strife, a peasant representative was installed.[100] Demands for creating peasant representative posts were quite
numerous in the first decades of the eighteenth century. Once they were recognized by overlords, the peasant representative's
seal was required on all official documents (including suits), next to that of the headman and the kumi heads.

Sakai Uji recently examined the history of this office in the Kanto provinces and concluded that its widespread "officialization"
was nothing but the sanctioning by the bakufu and daimyo of a much older peasant practice.[101] According to Sakai, one has
to distinguish between sobyakushodai[*] ("all-peasant representative") and hyakushodai[*] ("peasant representative"). In
general, the former is found in the records only

[98] Oishi[*] Shinzaburo[*] , Hokenteki[*]tochi shoyu[*]no kaitai katei: I Kinsei jinushiteki tochi shoyu[*]no keisei katei
(Ochanomizu shobo[*] , 1958), 177-85, esp. 184-85; Oishi[*] Hisakata, Jikata hanreiroku 2:93.

[99] KDJ 11:996, s.v. "hyakushodai[*] ."

[100] Oishi[*] , Hokenteki[*]tochi shoyu[*] , 180-81.

[101] Sakai Uji, "'Sobyakushodai[*] ' kara 'hyakushodai'[*] e: zenki hyakushodai[*] no seiritsu o megutte," in Ronshu[*]chu-
kinsei[*]no shiryo[*]to hoho[*] , ed. Takizawa Takeo (Tokyodo[*] shuppan, 1991), 373-406.

― 123 ―

between the 1660s and 1770s; the latter from the 1680s until the end of the Tokugawa period. Thus, "all-peasant
representative" was the only title until the 1680s, when the second title started to appear, but it remained the predominant one
until the 1720s, when it began gradually to lose ground, and it finally disappeared after the 1770s.[102] "All-peasant
representative" was first used for a position that had sprung up within the villages, whereas "peasant representative"
ultimately became the title used officially by the overlords and came to supersede the former. Sakai puts the official
institutionalization of this office in the Kanto provinces in the 169.0s.[103] This office thus developed within the context of the
small landholders' struggle for access to information and power discussed earlier. Although the first written trace of "all-
peasant representative" dates from the 1660s, the possibility or necessity of this village position is already contained in bakufu
directives of the years 1641-43.[104] The second of two copies of the village budget the bakufu ordered made in those years was
for the small peasants. This copy must have had a recipient. (The same goes for the tribute roster; see appendix 3, addendum
to art. 35.) So it is very likely that, insofar as this directive was implemented, there must have been at least an informal
peasant representative.

This "representative," however, was not necessarily a small peasant. It was more likely that he was a titled peasant and that if
genuine peasant representation by a non-elite peasant was to occur, this had to be achieved by internal pressure, as the
following case of Iribuse village (next to Ken's Makibuse) in Kita-Saku district indicates. There, the offices of kumi head and
peasant representative had always been filled "since the ancient past" by some of the twenty-four titled peasants. The
nontitled peasants, however, had agitated to have some say in who was elected, and they wanted to be candidates for office
themselves. In 1760 an agreement was worked out, on bakufu orders, by two officials from villages in the area (see appendix 1
for the text). First, the rule that the kumi head always had to be a nontitled peasant was abolished. However, nontitled
peasants could be elected, which must have been a con-

[102] Ibid., 377, 380.<

[103] Ibid., 396.

[104] Sakai refers to three directives in ibid., 397-98. One can be found in TKKz 5:153-54 (no. 2782, art. 10). The others are
directives by the Kanto intendant (gundai) Ina Hanjuro[*] that Sakai found in the Saitama-ken shi: Shiryohen[*] , vol. 7
(Saitama-ken, 1985), 46, 47.

― 124 ―

cession, although the eventual transfer of power from a titled peasant to a nontitled one was still conditioned, it seems, by the
willingness of the titled peasant to relinquish his post. Henceforward, however, the nontitled peasants would elect their own
representative from among their own ranks.

Thus the genesis of the office of peasant representative was prompted partly by the movement from below whereby peasants
insisted on access to crucial information on matters that profoundly affected their material well-being. In order to avoid
trouble, the bakufu incorporated their demands by ordering the diffusion of information. Often, we surmise, this information
came to be monopolized by the titled peasants. In a further move, then, the nontitled "small" peasants wanted a representative
of their class to be the one who checked the village elite's governance.

That power remained with the powerful in the villages should perhaps come as no surprise. For a considerable time, however,
the thesis that a small peasantry was established at the beginning of the Tokugawa period has prevented this view from
gaining full acceptance. Even though the original stratum of village power holders split into two, some of them becoming
salaried vassals, others remaining landed, the two new strata were able to make the transition into the seventeenth century
quite successfully. Economic expansion, especially the development of new arable, and political agitation from within the
villages, together with the overlords' responses to these developments, modified the traditional power of this old rural elite,
which, however, succeeded in reproducing itself well into the eighteenth century and in some cases even until the end of the
Tokugawa period.

Analytically speaking, one can refer to those who found themselves in this stratum as an objective class, because they
monopolized either economic cure political power (in the beginning) or political power only (later on, when economically they
became indistinguishable from "small peasants" in certain regions). When they lost their economic prominence, however,
many were able to hold on to their social position by manipulating a special kind of power, namely, that of status. Within the
villages, they closed ranks, forming a subjective class, to defend their privilege, a development taken up in the next chapter.

― 125 ―
3
Status Power
Samurai fight with weapons, peasants with lawsuits.
Tanaka Kyugu[*] , 1721

American students of the sociopolitical or socioeconomic (as opposed to the purely demographic or purely economic) side of
Tokugawa peasant society have, with the exception of Thomas C. Smith, framed their interpretations within a particular
trajectory of peasant protest spanning the whole period. In the first century, overlord domination met resistance by solidary
villages. By the nineteenth century, however, interclass confrontation had turned into intraclass struggle within the peasantry,
its solidarity now fractured along lines of divergent economic interests.[1]

This picture of late Tokugawa peasant society is not to be questioned even if these scholars take different positions vis-à-vis
class theory. Herbert Bix perceives a growing class struggle, while Stephen Vlastos works with a widened Marxian concept of
oppression. William Kelley, on the other hand, does not discuss the early period and firmly rejects any notion of group
solidarity to focus on shifting coalitions around temporary interests that cut across economic strata. That these scholars do not
seriously engage the notion of social class as such is irrelevant

Epigraph: Tanaka Kyugu[*] , Minkan seiyo[*] , 335.

[1] Bix, Peasant Protest ; Burton, "Peasant Struggle"; Irwin Scheiner, "Benevolent Lords and Honorable Peasants: Rebellion and
Peasant Consciousness in Tokugawa Japan," in Japanese Thought in the Tokugawa Period , 1600- 1868, ed. Tetsuo Najita and
Irwin Scheiner (Chicago: University of Chicago Press, 1978), 39-62; Vlastos, Peasant Protests ; James White, "Economic
Development and Sociopolitical Unrest in Nineteenth-Century Japan," Economic Development and Cultural Change 37 (1989):
231-59. William Kelley limits his analysis to the nineteenth century in his Deference and Defiance in Nineteenth-Century Japan
(Princeton: Princeton University Press, 1985).

― 126 ―

for the point I want to make, namely, that they all posit in some way a solidary village at the beginning of the Tokugawa
period. Admittedly, the seventeenth century is not the focus of any of these studies, yet one still may raise the question whence
and why this image of early village solidarity.

This question is also relevant with regard to the research by Thomas C. Smith, who stands alone in having highlighted, very
early, the seventeenth century's great economic discrepancies.[2] For Smith, looking for the agrarian origins of modern Japan,
the seventeenth century as origin of these origins was ultimately far removed from his interests. Moreover, the modernization
paradigm within which he was working, although anti-Marxist in one way, shared with economistic Marxism a slighting of
political factors. Smith's research agenda led him away from the questions taken up here.

These other scholars took as their subject the most salient features of peasant protest, the ikki, or confrontations of
supravillage authority by whole villages or groups of villages. These confrontations occurred only in mid and late Tokugawa.
When Anne Walthall and Herbert Bix touch upon intravillage protests, these are predominantly eighteenth-century protests.[3]

Status and Class—Mutually Exclusive?

Perhaps a more fundamental question underlies the assumption that economic discrepancies as source of class tensions
developed only fully in the eighteenth century, turning seventeenth-century villages into communities that either were
harmonious or harbored status conflicts, which were considered to be less significant than class conflicts. This question
concerns the relationship between class and status and, ultimately, how economic and symbolic matters weigh in on one's
scale of reality.

We know that status in Tokugawa Japan was fixed by laws, starting with Hideyoshi's famous Edict Restricting Change of Status
and Residence of 1591, which thus created a "society of orders." Actually, that edict did not aim at status as such: its immediate
purpose was to secure

[2] Smith, "Japanese Village."

[3] See Anne Walthall, Social Protest and Popular Culture m Eighteenth-Century Japan (Tucson: University of Arizona Press,
1986).

― 127 ―

a stable, productive agricultural labor force at home while Japan was engaged in a war abroad. Nevertheless, in combination
with the Sword Hunt Edict three years earlier, its sociopolitical effect was the great divide between rulers and ruled, samurai
and (mainly) peasants. While this nationwide divide certainly was not without obvious economic underpinnings, namely, a
division between producers and extractors of economic surplus, scholars have too readily assumed that it over-determined
people's preoccupations and, hence, that all serious conflict centered around the opposition between dominators and
dominated, or orders of rulers and commoners.[4] Status thus subsumed class before the two started to separate, that is, before
class divisions as such appeared, which would have occurred only sometime in the eighteenth century.

According to Herbert Bix, "Generally throughout Japan from the 1760s onward, the division of peasants along economic lines
was slowly beginning to override their unity along status lines.... One may reasonably hypothesize, therefore, that class and
income stratification was starting to introduce distinctions between households, which, in turn, reduced the likelihood that
mere status could continue to serve as a firm basis of solidarity."[5] And Nakane Chie writes that as a result of "the Tokugawa
policy," " in theory there was no social differentiation among the households of a village community," but by "the eighteenth
century ... differentiation in peasant statuses became common in most villages, resulting in titled farmers (owners) and
dependants (tenants)."[6] Stephen Vlastos, who writes that "the juridically determined social order in which peasants were
subsistence agriculturalists, as indeed they were in the early Tokugawa period, bore little relation to functional relationships
within the market economy of late Tokugawa—circumstances which profoundly affected conflict and collective action at the
end of Tokugawa rule,"[7] evokes an image of an early peasant class unified by status that developed intraclass conflict only
under the impact of a commercial economy.

[4] Bix, Peasant Protest , xvi-xvii, xxvii; Vlastos, Peasant Protests , 11-14, 159, 166-67; Burton, "Peasant Struggle," 138, 151, 166;
James White, "Rational Rioters: Leaders, Followers, and Popular Protest in Early Modem Japan," Politics and Society 16, no. 1
(1988): 46-47.

[5] Bix, Peasant Protest , 103, 104; on p. xvii he writes that "the degree of peasant 'classness' tended to increase while their
'statusness' decreased."

[6] Nakane, Kinship , 47, 49.

[7] Vlastos, Peasant Protests , 7-8.

― 128 ―

These interpretations contradict the argument presented in chapter z and have two points in common. One, they work within
the framework of a legally defined status order, implying that legislation was the only source of status division (which effected
class solidarity) and that no further status divisions existed in the village, because of the absence of legislation at that level,
assuming, in addition, that villages did not generate status differentials on their own. Two, since the legally defined status of
the peasantry gave way to internal class divisions, there were no internal class or social divisions worthy to be referred to as
such prior to the development of a market in the eighteenth century.

Class and status are dichotomized historically here, to the point of occupying mutually exclusive sequential positions: status
societies are succeeded by class societies.[8] A quick reading of Max Weber may give that impression, for he is quite explicit in
saying that class situations are market situations, while status orders (Stände ) are not, on this point approximating Karl
Marx's historical view.[9]

Class societies are dominated by "functional" interests, the rational dictates of the market, which does not know personal
distinction, and stratified through pure relations of property and production, generating classes that are unaware of their
shared objective conditions unless they resort to social and political action. Status societies, in contrast, are organized in
groups consciously distinguished from one another in terms of social honor, ascriptive privileges, lifestyles, and hence
prescriptive modes of consumption that abhor and suppress the purely economic, which is typical of class societies.[10] In class
societies, one may add, the ruling class owns the means of production (and its own means of consumption); in status societies,
an elite also directly controls the means of consumption of others through extraeconomic means.

Status, which Weber defines as "an effective claim to social esteem in terms of positive or negative privileges," entails a clear
social recognition of belonging, while property and social classes are constituted foremost by objectively shared interests,
which may or may not lead to forms of associations. For the sake of completeness, I should further

[8] For a similar dichotomy that has governed scholarship on seventeenth-century France (was it a "society of orders" or a
"society of classes"?), see the discussion in Beik, Absolutism , 6-9.

[9] Weber, Economy and Society , 2:927-28.

[10] Ibid., 1:306-7; 2:929-30, 932, 936-38.

― 129 ―

note that for Weber, property classes are "determined by property differences," while social classes are constituted by "the
totality of those class situations within which individual and generational mobility is easy and typical," class situations being
the "typical probability, within a given economic order, of procuring goods, gaining a position in life and finding inner
satisfactions."[11]

Dichotomies, such as that of class and status, are put forth by Weber to construct ideal types that, although mutually exclusive,
can serve as flexible, descriptive tools for analyzing social realities, which are always hybrids somewhere between the pure
extremes. A closer reading of Weber, therefore, reveals that the ideal types are not essences functioning as monocausal
explanations. His thought is more fluid; he thinks in more relational terms.[12] Weber suggests the following possible
relationships between status and class: "Status may rest on class position of a distinct or an ambiguous kind. However, it is not
solely determined by it .... Conversely, status may influence, if not completely determine, a class position without being
identical with it." And "class distinctions are linked in the most variegated ways with status distinctions. Property as such is
not always recognized as a status qualification, but in the long run it is, and with extraordinary regularity." On the relation
between power and status, he writes: "Quite generally, 'mere economic' power, and especially 'naked' money power, is by no
means a recognized basis of social honor. Nor is power the only basis of social honor. Indeed, social honor, or prestige, may
even be the basis of economic power, and very frequently has been."[13]

Pierre Bourdieu, in his ongoing reflections on the theoretical implications of his own vast research, has spelled out more
systematically the various ways in which class and status (in the modern sense rather than the historical sense of orders or
estates) are linked and how power and honor can be the basis for each other. Bourdieu has thus developed notions of various
types of capital (material, social, cultural, and symbolic) that can be converted into one another, which entails a critique of
substantialist thinking and realist views of class. Hence the importance,

[11] Ibid., 1:305, 302.

[12] Pierre Bourdieu stresses the fundamental importance of thinking in terms of relationships rather than essences (see his
Logic of Practice , 4; and In Other Words , 40, 126).

[13] Weber, Economy and Society , 1:306; 2:932, 926.

― 130 ―

as an analytic starting point for this relational approach, of the notions of field and social space, where people occupy certain
positions, through which one "can understand the logic of their practices and determine, inter alia , how they will classify
themselves and others and, should the case arise, think of themselves as members of a 'class.'"[14] Social spaces consist of
differential power positions, determined intrinsically by material conditions and relationally by their distinctive distance from
other positions. Agents thus identified as sharing common positions constitute only "theoretical classes" and not "groups which
would exist as such in reality," unless they are actually organized politically[15]

This objective truth of social classes has to be supplemented by considering what provides domination the surface of
legitimacy that allows relations of exploitation to function as such. This "supplement" (from an objectivist standpoint) is
provided by signs of distinction that effect a social misrecognition or, more precisely, at least a socially accepted
misrepresentation of those relations. Such signs of distinction function as a rare symbolic capital whereby one distinguishes
oneself foremost

[14] Bourdieu, In Other Words , 50; see also 49 and 126. Japanese scholars' discussion concerning status, which centered
around Minegishi Kentaro's[*] articles, some of which he collected in his book Kinsei mibunron , has been hopelessly stock in
the kind of realist, reifying perspective Bourdieu criticizes. Bourdieu advises against "questions ... about limits and frontiers"
when studying classes (In Other Words , 50), which is precisely what Japanese historians were preoccupied with in the 1980s,
according to Minegishi (Kinsei mibunron , 56). In addition, Minegishi argues (13-21) against the Marxist views of two
prominent medievalists: Kuroda Toshio's instrumentalist class interpretation of status as a means of "extra-economic
coercion" and Ishimoda Sho's[*] emphasis on the mediation of the state. Takagi Shosaku[*] (see chap. 2, n. 44) agrees with
Minegishi's anticlass stance but emphasizes the role of the state (dissociated from class) in contrast to Minegishi, who argues
that the origins of status were social. Recently, Asao Naohiro has pointed out that during the last decade a number of
historians have avoided the substantialist dilemmas that framed the earlier debate and started to analyze the flexible and
instrumental relationships between status and the evolving world of occupational diversification and competition, in other
words, how public sanction of status served interest groups not readily identifiable as classes ("Kinsei no mibun," in MK, 35-
38). Class has thus all but disappeared from the discussion of status among Japanese historians.

[15] Bourdieu, In Other Words, 117, 118. The simple distinction between objective and subjective classes proposed by Bourdieu
is none other than Marx's classes "in themselves" and "for themselves" or Berreman's "etic" categories and "emic" groups (see
Gerald Berreman, "Social Inequality: A Cross-Cultural Analysis," in Social Inequality: Comparative and Developmental
Approaches , ed. Berreman [New York: Academic Press, 1981], 18).

― 131 ―

from those holding immediately lower positions, whose aspirations constitute a direct threat to one's social identity. "The logic
of the symbolic," therefore, "makes absolute 'all or nothing' differences out of infinitesimal differences."[16] Thus, struggles
within social fields will always also be at least (and exclusively, unless the objective conditions are clearly foregrounded)
struggles for specific differences to thereby modify one's position within the established and accepted order of classification,
an order of legally or customarily institutionalized "indices of consecrations."[17]

Bourdieu's reflections on class and status are the result of extensive sociological research in modern society, where, as he has
noted himself, sumptuary laws do not set legal limits on the deployment of symbolic strategies, as they did in Tokugawa Japan.
Nevertheless, what he has to say further about the relation between the two is very helpful for the present study. Status
groups, Bourdieu argues, structure their strategies of distinction through a social logic of union and separation, seeking
thereby "to make de facto differences permanent and quasi-natural, and therefore legitimate, by symbolically enhancing the
effect of distinction associated with occupying a rare position in the social structure"; these strategies are therefore "the self-
consciousness of the dominant class."[18] Hence Bourdieu's critique of Weber, to whom he acknowledges a great debt: "'Status
groups' based on a 'life-style' and a 'stylization of life' are not, as Weber thought, a different kind of group from classes, but
dominant classes that have denied or, so to speak, sublimated themselves and so legitimated themselves."[19]

The Great Status Divide

Bourdieu's insights are relevant for the study of the macrosocial space of Tokugawa society and most pointedly for the micro
fields of power that

[16] Bourdieu, In Other Words , 136-37; the quotation is from 137.

[17] Ibid., 138. Bourdieu further elaborates on these "indices of consecration" as "objective marks of respect calling for marks
of respect, a spectrum of honours which have the effect of manifesting not only social position but also the collective
recognition that it is granted by the mere fact of authorizing such a display of its importance."

[18] Ibid.
[19] Ibid., 139. Bourdieu has mentioned and discussed his indebtedness to Weber on numerous occasions; see, for example,
ibid., 21, 27-28, 46, 49, 106-7.

― 132 ―

were the villages. From this perspective, Tokugawa feudalism's ever-multiplying status stratifications and hierarchies, literally
embodied through codes of dress and address attached to hereditary households especially in the dominant class (thereby
endlessly split into dominant and dominated fractions), can be seen as defense strategies of particular groups against the
aspirations of neighboring groups. Although conceivable and indeed real because realistically possible in the turmoil of the
sixteenth century, skipping several rungs on the hierarchy became gradually unthinkable, which strengthened domination in
that the range of people's aspirations became more and more circumscribed. Those occupying the intermediary positions
would certainly resist such ambitions by people below them.

Status legislation with regard to the peasant class developed over time and aimed mainly at reinforcing the distance between
peasants and warriors. Peasants were not allowed to use surnames in public documents or to carry long swords, and they had
to wear cotton, dismount when encountering samurai, use respectful forms of address, and so on.[20] Overlord authority,
which through its land surveys had identified certain peasants as "titled peasants," did not differentiate further among this
peasant elite except by lifting some of the prohibitions, allowing the use of surnames and the wearing of one sword by peasant
officials such as village headmen and village group headmen.

It should be understood that the prohibition was not against having surnames but against using them in public, such as on
official documents, and it was not a Tokugawa innovation, since it had existed already in the mid Muromachi period, as a
prohibition of 1485 reveals.[21] And exceptions during the Tokugawa period were perhaps less frequent than one might think.

In the 1820s in Tsuyama domain (Mimasaka province), only 44 commoners from 266 villages were granted the privilege of
using their surname; in Katsuyama domain (also in Mimasaka province) only 59 from 106 villages; and in Matsushiro domain
(Shinano) only 2 out of 871 village officials were granted the privilege. Moreover, the prohibition on surnames was directed
most specifically at peasants. Doctors

[20] Minegishi reports such legislation for clothing in 1628 and for behavior in the 1640s in bakufu territories and in Choshu[*]
domain (Kinsei m ibunron , 114).

[21] Kobayashi Kei'ichiro[*] , "Shomin no myoji[*] wa itsu goro kara tsukerareta ka," Nagano , no. 99 (1981): 1-2; see also
Sekigawa Chiyomaru, "Shomin no myoji[*] ni tsuite," ibid., no. 3 (1965): 14-15, 16.

― 133 ―

and Shinto priests were exempt, and exceptions made for other commoners seem to have been far more numerous among
townspeople than among peasants.[22]

Overlords identified peasants with the land, rather than the other way around. The peasants' generic (and public) identity was
not related to their family, their household, or their lineage, but to the land, an identity that was sometimes inscribed in
documents, next to their first name, where, in "sidescript," the size of their field was added, such as "Yohachi2.785 koku ." That
this prohibition spilled over, to various degrees at different points, into the "private" sphere is clear from the way documents
of religious organizations for the management of festivals or membership lists of confraternities were signed. Only in the mid
or late Tokugawa period did all the signatures on some membership lists analyzed by Kobayashi Kei'ichiro[*] include
surnames.[23] This is an indication that the status markers created by the overlords were being accepted by and large within
peasant communities.

As we have seen, the political concerns of seventeenth-century peasants were intravillage divisions rather than the macro
divide between samurai and peasants. In the villages the confrontations were between economically and socially neighboring
groups, which is where such struggles always take place, as both Max Weber and Pierre Bourdieu have pointed out.[24] In
premarket economies, according to Bourdieu, economic power must be partly converted into symbolic capital, "a

[22] Sekigawa, "Shomin no myoji[*] ," 15.

[23] One complete list, covering the years 1592-1902, of the organizers of a yearly festival at a Hachiman shrine reveals that
until 1612 only first names were entered; a few surnames were included in some years between 1613 and 1669, when the list
reverts back to first names for the next eighty years; and after 1748 all signatures include surnames. A Koshin[*] confraternity
list (1693 to the present) includes the surnames of all its members starting in 1858 (Kobayashi, "Shomin no myoji[*] ," 3).

[24] As Weber put it, "It is not the rentier, the share-holder, and the banker who suffer ill will of the worker, but almost
exclusively the manufacturer and the business executives who are the direct opponents of workers in wage conflicts. This is so
in spite of the fact that it is precisely the cash boxes of the rentier, the share-holder, and the banker into which the more or
less unearned gains flow, rather than into the pockets of the manufacturers or of the business executives" (Economy and
Society , 2:931). And Bourdieu writes that "minimum objective distance in social space can coincide with maximum subjective
distance. This is partly because what is 'closest' presents the greatest threat to social identity, that is, differences (and also
because the adjustment of expectations to real chances tend to limit subjective pretensions to the immediate neighborhood)"
(Logic of Practice , 137)

― 134 ―

legitimate possession grounded in the nature of its possessor," or symbolic power, "the power to secure recognition of power,"
before it is accepted as legitimate by those subjected to it.[25] In Tokugawa villages this symbolic capital, grounded in but not
identical to material conditions, consisted of status and was largely generated and manipulated from within.[26]

"Shared" Communities

Several times in chapter 2 we came across various ways of dividing power, tribute, corvée, and so on, within the village.
Households were corvée households (yakuya) or not, and some were only half so (hanyakunin). Official obligations could be
distributed per household (iewari , tsurawari ), per assessed yield (takawari , tanwari ), or as combinations thereof. Every
household (ie) had some share of privileges and burdens. A good part of village life was literally a matter of "shares," kabu .
Tokugawa villages were "shared" communities to a very high degree, making them probably quite unique.

This term conveys well, through a concrete image, the concept behind this structure. Kabu literally means "stump" or "roots" of
a tree or plant, and kabuwake means "dividing roots (shares)," as when one separates the roots of a plant into parts so that it
can multiply and be divided further into share portions or recombined into bundles. Kabu is a quantitative and combining
form that could also be, and was, applied to nonmaterial things such as status and power. (Modern stockholding companies,
kabushiki kaisha , deal in shares or stocks, kabushiki .)

The concept of shares, while seemingly modern, is intimately related to the "refeudalized" setup of the early Tokugawa period,
when over-

[25] 25. Bourdieu presents this argument in "Modes of Domination," chap. 8 of his Logic of Practice (122-34); the quotations are
from 129 and 131.

[26] Smith mentions both class and status when writing about the great contrasts in seventeenth-century villages, and he says
that the documents leave "little doubt concerning the general picture of inequality among holdings .... Not infrequently we
encounter peasant holdings that can only be described as estates." He goes on to say that it is "not surprising that such
extremes of wealth and poverty among hyakusho[*] of the village were accompanied by marked social distinctions." There
were, he says, "two distinct economic classes among the peasantry," and "not infrequently the existence of distinct economic
classes in the village was explicitly acknowledged" ("Japanese Village," 266, 268 and n. 28).

― 135 ―

lords established villages as tribute units for a certain quota of produce, goods, and services based on the village kokudaka, or
assessed yield. Although land cadasters and population registers recorded individual plots and the people on them, the yearly
tribute expected by the overlords was the village quota. This relieved the rulers from keeping track of the ever-shifting land,
population, and relations between the two within villages. They did not want to deal with the complex effects of time and
change. Consequently, villages were left to deal with this reality on their own.

For the peasants, in other words, cadasters and registers were outdated almost as soon as they were compiled: households
multiplied or died out, acquired or lost land; new fields were developed, others abandoned; changing weather conditions
resulted in average or bumper crops, or total or partial crop failures. All these variations had to be juggled locally in a
coordinated effort to produce the fixed quota (which only rarely was adjusted downward in bad years), in much the same way
that factory workers have to adjust their schedules, breaks, sleeping, gestures, postures, and breathing to the set production
quota at the fixed speed of the conveyer belt. A coordinated group effort is essential in such systems.

In the beginning, villages had to struggle with the discrepancy between real conditions and productivity expectations without
any directives from above. Then, most of the nengumenjo[*] , the yearly village tax bills, simply included the total amount due
and the order that (1) all of it be forwarded after a meeting of "the headman, the elders, the small peasants, and outsiders who
owned land in the village," and (2) "since the rate (men ) is a global one for the whole village (souke[*] ) and it is natural that
some peasants go bankrupt or abscond, the community make up the forfeited amount."[27] The question thus faced every year
was how to effect an equalization (domen, jinarashi , literally, "land

[27] Mizumoto discusses these intravillage adjustments in pt. 2, chap. 2, and pt. 3, chap. 2, of his Mura shakai. The following
discussion, however, is based on Nishiwaki Yasushi's study of Niremata village, Anbachi district, Mino province (in what today
is Gifu prefecture) (see "Kinsei zenki no nengu sanyo[*] to 'mura' chitsujo: 'Narashi' sanyo[*] o meguru murakata sodo[*] no
bunseki o toshite[*] ," Shikan , no. 106 [1982]: 19-39 [hereafter "Nengu sanyo[*] "]; and "Kinsei zenki Mino Waju[*] chiiki no
'kono[*] ' to sonraku: Bakuryo[*] Anbachi-gun Nirematamura ni okeru kisoteki bunseki," Gifu shigaku , no. 76 [1982]: 41-82
[hereafter "Mino Waju[*] "]. The quotation, from a directive of 1621 by an intendant's assistant (tedai), is taken from idem,
"Nengu sanyo[*] ," 20.

― 136 ―

exemption," "land equalization") of this burden among all members of the village, which, if for no other reasons, thereby
became a community with vital interests at stake, in other words, a matter of sharing.

There were two alternatives: either to apply a new rate to each plot according to its officially assessed grade to produce the
supplementary tribute necessary for meeting the village quota or to check the actual harvest of each field and allocate tribute
accordingly within or without the framework of graded fields. The latter practice, devised by the peasants, was ultimately
sanctioned by law in the mid 1620s (in the district studied by Nishiwaki Yasushi, on whose research I rely). The tax bills
ordered equalization based on the actual harvest and according to the grades of the arable.[28] The bakufu thus incorporated
into its policy peasant practice that in fact was dealing with the shortcomings of the kokudaka system. This was one way in
which the original system was supplemented by new directives to make adjustments for changes that inevitably occurred over
time.
The amounts involved were sometimes considerable. In a village studied by Mizumoto, the village as a whole was 34 percent
under par in 1623, but the individual fields' shortfalls varied greatly from a low of 9 percent to a high of 77 Percent.[29] It is not
hard to imagine the tensions this produced within the village, tensions between villagers and the headman (especially in the
early period, when the headman monopolized both precise knowledge and decision-making powers) as well as between
tenants and landowners (because landowners and not the tenants paid the tribute for rental parcels).[30] And yet, each year
there was a deadline to be met.

[28] Nishiwaki, "Nengu sanyo[*] ," 21-24. Besides the two alternatives mentioned in the Niremata documents, there actually
existed a third one, which was practiced widely elsewhere: jiwari , or redistribution of the land (without changing the
proportion of the total acreage held by each household). Often (but not only) applied when land was lost as a result of floods or
other natural disasters, the land in the village was redistributed—sometimes by village initiative, sometimes by overlord order
—only once or regularly at intervals of several decades. Unlike earlier scholars, Aono Shunsui considers this system an
integral part of the Tokugawa murauke system (see his Nihon kinsei warichiseishi no kenkyu[*] [Yuzankaku[*] , 1982], 14-18). In
some locales, this system operated until the 1950s (KDJ 7:769, s.v. "jiwariseido"). For the only discussion of the system in
English, for Kaga domain, see Brown, Central Authority , 94-112.

[29] Mizumoto, Mura shakai , 41-43.

[30] Ibid., 52-63. In Saku district, Shinano province, tenants usually paid two-thirds of the harvest to their landlords, retaining
one-third for themselves (Ichikawa Takeji, "Saku chiho[*] ni okeru reizoku nomin[*] 'kakaebyakusho[*] ' no jittai," Nagano , no.
128 [1986]: 15).

― 137 ―

Initially, the bakufu had a simple way of dealing with those peasants who fell in arrears: it took their land, gave it to someone
who could pay the tribute share, and banished the culprits. We find such measures both in intendants' directives of 1640
("transfer the land") and in bakufu rules for villages issued in 1642 ("banishment").[31] With the famines and crop failures of
the 1640s threatening the very existence of many small landholders, however, the bakufu reversed course and instituted its
famous policies to protect the "small peasants" from bankruptcy: it required, among other things, that all peasants approve the
tax distribution by affixing their seal, instituted a system of receipts for payment of tribute, and forbade the sale of land.[32]

Mizumoto has heralded these equalizing village practices as victories by peasants for the causes of justice and equity,
especially against the arbitrary rule of headmen, and they generally were.[33] In order to indicate, however, how local
conditions made such struggles very complicated, and their immediate outcome at times less than an emancipation of small
landholders, we shall follow in some detail Nishiwaki's case study of one such struggle in Niremata village, Mino province
(Gifu prefecture).

Factious Peasants: Niremata, 1654

There are various overlapping dimensions to this case in Niremata village: struggles of elders to participate in village
government, tensions resulting from the uneven impact of a drought on the harvest, a consolidation of lineage power, and an
attempt to split the village into two. A partial apportioning of well water during the drought of 1654 had mixed results in
Niremata. Upland fields could not be irrigated and thus

[31] Nishiwaki, "Nengu sanyo[*] ," 22. For the bakufu directive, see TKKz 5:155 (no. 2784, art. 8).

[32] The famous prohibition on the sale of land was issued in 1643/3 (TKKz 4:121 [no. 2104, art. 3], 5:157 [no. 2786, art. 13]). For
an English translation of this important law, see David John Lu, Sources of Japanese History , 2 vols. (New York: McGraw-Hill,
1974), 1:206-7. For a discussion of the law, see Oishi[*] , Kyoho[*]kaikaku , 9-43. The requirements of seals and receipts were
legislated in 1644 (TKKz 4:122 [no. 2105, arts. 14, 15]).

[33] Mizumoto, Mura shakai , 45, 56-60.

― 138 ―

registered a serious crop loss. A denied request for a tax cut was followed by a request for an on-the-spot survey, addressed to
the intendant's assistant (tedai) in charge of crop inspections. In this request the peasants promised either to accept and pay in
full whatever lower amounts the inspector would set or, if the inspector did not act, to lower the tribute on the affected fields
themselves and make up the deficit by additional levies from fields that had escaped damage from the drought. The inspector
saw no reason to lower tribute rates, and so the trouble began.

A faction formed around Genzo[*] (with 28 koku, only 5 less than the headman), the second largest landholder and one of the
four kumi heads (there was also an elder). Niremata had a kokudaka of 526 koku and counted thirty-five resident landholders,
eighteen of whom were independent tribute payers.[34] Genzo's[*] faction held a total of some 120 koku. His followers tore up
the agreement, insisting that tribute and corvée for the overlord be allocated across the board according to the regular rate
regardless of losses. The other faction, led by the elder, the other three kumi heads, and the headman (the author of the
agreement), was forced to borrow cash in various places to meet the village quota.

Around harvest time, in the tenth month, Genzo's[*] faction wanted to split off from the village altogether and have its own
headman, storehouse for tribute rice, runners, and so on. The intendant's assistant agreed only to partition the existing
storehouse, providing two separate entrances, but Genzo[*] rebelled: he destroyed the storehouse, scattered its instruments,
stored the rice of his followers in his own granary, and again applied to secede. The headman argued that there was opposition
from all the peasants because of the increase in administrative costs. Ultimately, the village split into two Kumi (not to be
confused with the kumi of five-household groups) (see table 11).
Obviously, those who suffered losses beyond the village's average would benefit most from an adjustment of tribute amounts
per field based on the actual condition of each field. These lowered amounts would be supplemented by others, who would
pay above their normal share. We cannot know whether the secession drive was fueled by the uneven distribution of crop
losses among the two factions, for we only have data on the headman's group (see fig. 3).

[34] Nishiwaki, "Nengu sanyo[*] ," 24-38.

― 139 ―

Table 11.
Holdings of Niremata's Two Kumi, 1654

Headman's Kumi Genzo's[*] Kumi

33

28

24

20

19

15

14 14

13 13 13

12

11 11 11

10 10

99 9

888 8

777 777

5 5

33 3

22
 1

SOURCE : Computed from Nishiwaki, "Nengu

sanyo[*] ," tables 2, 3, and 5 (pp. 28, 29, and
32).

NOTE : Each figure represents one

household's kokudaka. The figures for the
Genzo[*] faction are approximate figures.
For a different but very close set of figures,
see Nishiwaki, "Mino Waju[*] ," table 10 (p.
56).

It is worth noting, however, that all thirteen landholdings on Genzo's[*] side were 15 koku (one) or less (the other twelve),
while the headman's group included, besides some twenty-one peasants in that category, all four large holders of the village
excluding Genzo[*] (with 19, 20, 24, and 33 koku, the last one the headman's). Losses in the headman's faction

― 140 ―

Fig. 3.
Crop Loss per Holding Size in Niremata's Headman's Kumi, 1654

Source: Nishwaka, "Nengu sanyo[*] ," table 3 (P. 29).

averaged 13.3 percent of the quota.[35] As it turns out, sixteen households (among whom there were nine dependents) had
losses above the average, and nineteen households (including eight dependents, among whom were the only three that had no
losses) had losses below the average. The bulk of the "small" peasants (seventeen out of twenty-two

[35] Aono (Warichiseishi , 18, 51) estimates that if any adjustment of over 10 percent was needed to bring the actual yield in
line with the putative yield, then a proportional distribution of the burden based on the land cadasters became irrational.
Then villages had to resort to an equalization (based on an internal land survey they conducted themselves), or levy a fee per
bale (todai , "a fee per to [one tenth of a koku or 18 liters]"), or, ultimately, perform a redistribution of land (jiwari).

― 141 ―

households, with 5 to 15 koku) had less than average losses, but all six "very small" holders, of less than 5 koku (five of whom
were dependents), lost more than the average (excluding one who had no losses).

What can we learn from this hard-to-follow detail? Its relevance has to do with the way tribute was paid: only independent
landholders paid it, including the portion of their dependent holders' parcels if they had dependents. In the present case this
applied mostly to landholders of over 15 koku, especially the larger landholders, among whom are the village officials. Six out
of the seven such landholders in the headman's faction, it is worth noting, were also among those who suffered above average
losses. While very small and small peasants undoubtedly could be ruined by a crop failure averaging only 13 percent (one
holder of 1.9 koku registered a 52 percent loss; another of 4.4 koku, one of 42 percent), it was the large holders who lost most
in absolute terms, since they also shouldered their dependents' tribute payments. Hence it was the large landholders who
stood to gain most from a tax formula that supplemented their shortfalls with extra contributions from those with lesser
losses, that is to say, from the "small peasants," who were mostly autonomous and without any dependents themselves. The
loss of this stratum of small peasants alone constituted 38 percent of the headman's faction's total loss.
From a rational economic point of view, one may thus hypothesize that the wealthier peasants, led by the headman, sought to
defend their interests through a tax formula that was most beneficial to themselves because it made the small independent
peasants shoulder the burden. From this view also, Genzo[*] would then represent the plight of the small peasants, which
makes one wonder why more small peasants did not side with Genzo[*] .

Before we turn to this question, a few observations are in order. This is a good illustration of how extramural authority
affected village politics through the sheer mechanism of the village tribute quota. Weather conditions affected individual
peasants unevenly, but the adjustment was communal, which in turn brought about a new unevenness, this one, however, the
result of human decisions. While ultimately the overlords were the enemy, the immediate target of these struggles was within
the village, where greater or lesser fairness could be achieved. And finally, the factions that formed were not mere collections
of autonomous individuals differentiated only by holding size. The partic-

― 142 ―

ular relation between dependents and their bosses in matters of tribute payments gave a special character to these factions.

Villages were structured by what one can call "lineages," although these were not exclusively blood lineages, since non-kin
were among its members. Cognates were often set up as branch houses with various degrees of autonomy with regard to title
to land. Even when they became independent, however, tribute-paying landholders remained tied to the main house through
various obligations. Bond servants (nago, genin, and fudai, about which more will be said later) were even more dependent,
even if they worked specific parcels of the main house land almost as if they owned them. This dependency related very
concretely to the use of tools, equipment, and so on, but also to access to fertilizer, which in the early Tokugawa period was
secured almost exclusively by cutting grass on the mountainsides. This access was not free. Not that extraordinary fees had to
be paid—they were very modest—but this also was a matter of shares. And these shares were controlled by the lineages.
Indeed, if one was not a member of a lineage, one could not have access to the mountains for fertilizer, wood, and so on.
Outside the lineage there was no way to survive.

Genzo's[*] village, Niremata, counted three lineages: the headman's, one headed by one of the kumi heads, and one headed by
Genzo[*] . Genzo's[*] holding was second to the headman's and far larger than that of any of his lineage members. When
Genzo[*] decided to secede, he certainly had ambitions of becoming a headman in his own right, but his following was limited
to his own lineage. The dispute over tribute allocation was thus used by Genzo[*] as an occasion to play out lineage rivalries
and increase his own power by adding to it the official dimension of a headmanship.[36]

Lineage cohesion was often reinforced ritually through ancestor worship, by members' belonging to the same ancestral
Buddhist temple for annual or memorial rites for the dead. Niremata is an exemplary case of such practice. There were three
temples in the village, one for each

[36] One should recall that lineages were not sanctioned, either negatively or positively, by higher authority. For the overlords,
the only official subdivision of the village was into kumi of five households, each with its own head. The authorities, however,
may well have aimed at breaking lineage power indirectly, since intendants often required that kumi be organized strictly on
a contiguous geographic basis; that is, they should not include just "kith and kin" (Ichikawa Yuichiro[*] , Saku chiho[*] , 25, 33.
See also article 1 of the goningumi rules for Shimo-Sakurai village in Kita-Saku district [Shinano] of I640 in appendix 2).

― 143 ―

lineage: temple and lineage membership overlapped almost perfectly (only few households belonged to outside temples), even
to the point that at the time of the split into two Kumi the two blood-related households of Genzo's[*] group who shared the
same temple as the headman defected to the headman's Kumi.[37] This triggered a stern warning from Genzo[*] that any
further defections would result in the confiscation of the property by his Kumi! Effective or not, this threat illustrates the kind
of power lineages presumed to wield even in matters of landownership.

Why did Genzo[*] choose this moment to secede? The number and composition of Niremata's neighborhood kumi was not
stable in the 1630s and 1640s: there were five in 1630, ten in 1634, and five again in 1641. It appears that while certain of these
kumi were made up of lineage members only between 1638 and 1669, the smallest of the three lineages, Genzo's[*] (who had
been an elder and then kumi head all along), was dispersed over several kumi until 1649.[38] In that year, however, he was
registered as head of a kumi in which all but one of the members of his lineage were gathered. The secession group that
formed a separate Kumi five years later was thus nothing else but a singlelineage kumi. As in the case of Makibuse village in
chapter 1, as well as in the case of Kodaira below, the bakufu rule concerning kumi formation seems to have been widely
ignored, seriously affecting intravillage authority relationships.

The secession petition was signed not only by the five full members of Genzo's[*] kumi: it bore sixteen signatures. Thus, within
the secession move the voices of those without full political membership in the village suddenly became important, and to that
extent the power of their leader and lineage head can be said to have been conditioned by their assent. It is clear, however,
that such "emancipation" of dependents was circumscribed by lineage control of the lives of its members. This is also evident if
one compares the kumi membership list with the population register (shumon[*] aratamecho[*] ). Dependents do not figure at
all on the first list, but they are marked as such on the second. On the latter their status as kerai , or vassals of their patron
households, is clearly indicated. Thus the population registers, a neutral recording device in the eyes of the overlords, came to
be used not only to record people but also to make their status official.

[37] Nishiwaki, "Nengu sanyo[*] ," 34, 35.

[38] Ibid., tables 7 and 8 (pp. 36, 37); idem, "Mino Waju[*] ," 44.
 ― 144 ―

Fractioned Peasants: Niremata, 1615-1680

A document of 1651 lists 33-7 titled peasants in Niremata. What could be the meaning of a fraction of a peasant? Here we face
again the operation of shares, this time shares in the status of "titled peasant." Such a fractioned status, however, was not
instituted by the higher authorities.

Niremata's household register (iekazu aratamecho[*] ) of 1635 reflects the overlords' perspective.[39] It lists 30 ie: 18 titled
peasants and 12 "others," which include the headman, three temples, and a tobacco cutter, among others. This document,
addressed by the headman to the intendant, is an official report of Niremata's corvée capacity. Another document, of 1651, lists
a total of 42 ie (an increase of 40 percent over sixteen years), 33.7 of which owed corvée: 26 titled peasants and 16 others who
held fractions (of 30, 50, 60, and 70 percent) of titled peasant shares with a total value of 7.7 shares, making a grand total of
33.7 titled peasants.[40] This document is an internal village document addressed by the five elders to the headman concerning
the distribution of corvée due to the overlord. Some of these corvée-owing families were among those listed separately sixteen
years earlier as "others." Many of the new full or fractional titled peasants were recently established branch families (with or
without blood ties to their main families). The distribution of the required corvée among these ie reflects the village's
adjustment to the establishment of new autonomous landholders.

The population registers (shumon[*] aratamecho[*] ) record this trend: between 1638 and 1646 all small landholders, even
those of nonblood dependents (genin), are listed as independent units. From 1649 on, however, the genin are dropped, and we
find only extended families listed. How does one make sense of these shifting ways of recording households, first as dependent
(1635), then as independent (I638-46), and then again as dependent (1649)? This does not fit the pattern of gradual and
progressive emancipation of dependents as a result of economic circumstances that Smith's description in his Agrarian Origins
of Modern Japan leads one to expect.

[39] Nishiwaki, "Mino Waju[*] ," 43-44.

[40] These figures are not computed from Nishiwaki's "Nengu sanyo[*] ," tables 2, 4, and 5 (PP. 28, 31, 32), which would lead to
slightly different numbers; instead, they are taken from his "Mino Waju[*] ," 43.

― 145 ―

We get a first clue from two population registers from 1646 that differ in two major respects: none of the figures in the two
documents are the same, and only one of the two hears a seal. This is an example of double bookkeeping. One register was
destined for extramural eyes , namely, the intendant's, and was thus an official document; the other was for intramural use ,
reflecting genuine practice. The former lists 26 ie grouped in five kumi with a population of 224, the latter 41 real ie divided
among eight kumi with a population of 267.

Nishiwaki explains the discrepancy as follows: the village under-represented by fifteen the number of landholders in order to
avoid an increase in tribute, especially corvée, which was assigned per ie, distributing this corvée among all the households,
even those not registered as official tribute units. This created an ambivalent situation for newly established small holders. If
they became full-blown titled peasants, their tax burden would increase, but with it would come full political membership. On
the other hand, the titled peasants of old could only lessen their corvée burden by setting up small landholders to share this
burden. The compromise was to limit the number of titled peasants but grant branch houses fractions of a "title," perhaps
increasing the fraction as time went by and they became economically more secure, until they were "fully vested."

The bakufu was caught between two contradictory policies. On the one hand, especially after the famines and crop failures of
the early 1640s, it had to help small landholders survive, which implied establishing them as autonomous tribute producers,
as they were recorded on the land cadasters (the survey of 1623 showed fifty-two peasants, 30 percent of whom had holdings
of less than five koku).[41] On the other hand, the real tribute payers were the wealthier peasants, those in the 10-30 koku
bracket, to whom many of the smaller peasants were attached in some form of dependency. Nishiwaki's analysis of which
bracket of holders held the highest percentage of taka (and hence produced the most tribute) in Niremata between 1637 and
1684 reveals an interesting pattern (see table 12). Before the 1630s the percentage of those with less than ten koku gradually
increased to 13 percent. This trend was then interrupted during the crisis decade of the 1640s, when it dropped to between 3
and 7 percent. It picked up again in 1648, when it jumped to 35 percent, and it did not drop below 30 percent for

[41] Nishiwaki, "Mino Waju[*] ," table 7 (52).

― 146 ―

Table 12.
Categories of Holdings in Niremata, 1623-
1684

Taka Holdings (in koku )

0-10 10-20 20-30 30 +

1623 10 25 55 8

1630 13 23 39 22

1642 3 32 26 37

1643 3 32 26 37

1644 5 8 38 47

1646 5 6 38 49

1647 7 16 33 42

1648 35 44 14 1

1651 39 36 18 6

1653 12 27 26 34

1657 29 55 6 9

1660 34 42 5 17

1669 39 39 11 9

1684 16 39 22 21

SOURCE :Adapted from Nishiwaki, "Mino

Waju[*] ," table 12 (pp. 61-62).

the next forty years, except in 1653, when it was 12 percent, and 1684, when it was 16 percent.

This curve seems to reflect the slow growth of small landholders, interrupted by the natural disasters of the 1640s, and their
subsequent resurgence, perhaps as a result of bakufu policies. The question is, what happened to them during the half-decade
or so when they disappeared from the statistics, and what was behind the sudden jump in 1648? The answer to the first
question is simple: many of the small landholders were reabsorbed into their parent households. The larger members of the
lineage tided them over during the bad years. This is one instance where one can observe, not an irreversible "emancipation"
of dependents, but a circular movement whereby "independent" holders, in various ways still in need of the support of their
parent households, returned to a position of full dependency.[42] It is important to note,

[42] For instances of movement back and forth between the status of fully dependent bond servant (genin) and house-owning
lifelong servant (fudai), and between the status of fudai and yakunin and back to fudai over two generadons, see Minegishi,
Kinsei mibunron , 142-46.

― 147 ―

however, that during this period far more non-kin than kin dependents were set up as independent branch houses. In other
words, although non-kin dependents were "treated as kin," they seem to have been left to their own devices in times of crisis.
[43]

This practice of absorbing small units into larger ones went directly counter to the intendant's policy of establishing small
independent landholders. This is clear from the difference between two tax rosters (menwaricho[*] ) submitted to the
intendant ten days apart, which explains the sudden jump in the numbers of small holders. The first roster, dated 1648/11/11,
listing the names of twenty-nine tribute owners with their individual assessed values, was obviously returned by the
intendant to be replaced by a second roster, dated 1648/11/21, which was almost twice as long, listing fifty-two names.[44]
Although the total villageassessed value and the number of its tribute payers is the same on both lists, the second list certainly
reflects the intendant's effort to identify, and thereby perhaps empower, the individual tribute producers and increase the
number of households available for corvée.

The additional producers on the new list were overwhelmingly family members, especially direct descendants: thirteen sons
and grandsons, two brothers, and two servants (genin). This does not mean that they were independent, because the original
parent families continued to use their own seals for them on official documents. Neither were they all small owners: the
headman's 72 koku of the first list was broken down on the second into 33 koku for himself, 23 for his son, 13 for his grandson,
and z for a bond servant. It is among these new additions, however, that the "fractional peasants" appear. For example, the
servant (incidentally, the younger brother of a 13-koku holder) is a 0.3 titled peasant; two others, with 8 and 10 koku, are both
0.7 titled peasants.[45]

[43] In 1644 there were thirteen branch households in Niremata, eleven of which were not related by kinship to their main
households (nine had no holdings). Smith (Agrarian Origins 15 ff.), among others, stresses the fictional kin relationship that
servants had with their masters' families and that they "were often maintained by virtue of an obligation w keep them rather
than for strictly economic reasons" (29).

[44] For the two lists, see Nishiwaki, "Mino Waju[*] ," table 13 (64).

[45] Data computed from ibid., table 15 (64); and idem, "Nengu sanyo[*] ," table 5 (32).

― 148 ―

These gradations reflect to some degree holding size, although as many as nine fully titled peasants had smaller holdings than
that of a 0.7 peasant, and also degrees of dependency to the main house as well as shares in political power. This status was
not reflected on the tax rosters, but on the population rosters, where most of these peasants were also recorded as heya-sumi ,
or "living in a room" (of the main family).

Although the authorities recognized any taka holder as a peasant, they ultimately came to recognize, via the population
registers, the lineage-based hierarchy among holders within the village. Here the trend was to cap the number of full-status
peasants (eighteen in 1651; fifteen two years later; then fourteen; and finally thirteen in 1669), limit the number of official
branch houses and restrict these to kin, and ultimately, in 1684, to absorb newcomers into the village (mainly craftsmen and
nonpeasants) within the lineage framework explicitly as retainers. Only in the 1680s was the status of branch families as
autonomous units acknowledged within the village.[46] By 1684 the heya-sumi category had disappeared, as had all extended
families (see table 13).

The Niremata case illustrates the impact of population growth on village life in seventeenth-century Japan: how the village
adjusted its internal social and political structure and how this adjustment had to take into account the overlord's own
reactions to this growth and, in a feedback movement, to the village adjustments to his reactions. The political, social, and
economic status of new households shifted several times over a sixty- to seventy-year period. Let us now closely follow an
effort by some dependents to change their status.

From Tribute Producers to Tribute Payers: Shimo-Kaize, 1687

Peasants were hierarchized according to degrees of autonomy, as Thomas C. Smith reported in the first chapters of his book.
Titled peasants with extended households had a number of dependents aside from kin: lifelong servants (fudai), without
separate living quarters and mostly single; and bond servants (genin), with or without separate living

[46] Data are from Nishiwaki, "Mino Waju[*] ," tables 10, 13, and 17 (56, 59, 78).

― 149 ―

Table 13.
Household Patterns and Branch Houses in
Niremata, 1669 and 1684

Types of Households 1669a 1684b

Linealogical
composition

Titled peasants 13 —

Kin branch houses 2 —

Non-kin branch 2 —
houses
Total 17 34

Kin co-residents 22 0

Non-kin co-residents 2 0

Total 24 0

Nuclear family

Main family 5 17

Kin branch houses 4 9

Non-kin branch 0 4
houses

[unclear] [2] [9]

Total 11 39

Extended family

Main family 7 0

Kin branch houses 0 0

Non-kin branch 2 0
houses

[unclear] [0] [0]

Total 9 0

a Headman's Kumi only.

b Both Kumi.

SOURCE: Compiled from data in Nishiwaki, "Mino

Waju[*] ," tables 10 and 17 (pp. 56 and 78).

quarters, single or married, with or without a plot more or less their own. One step above them were semiautonomous branch
houses (kadoya), which were servants set up in a separate "dwelling," or shack. These structures, often located at the entrance
to the compound, quite commonly were no more than hole dwellings with a pointed, thatched umbrella roof reaching to the
ground, identical to prehistoric Yayoi dwellings. Also considered as dependents were indentured servants, who were almost
always unmarried. Extended families set up

― 150 ―

branch households (bunke or kakae) as titled peasants or not. They all remained incorporated into a lineage structure.
The ladder of emancipation and prestige extended through the various ranks of bond servants to semi- and fully established
branch houses, titled peasants, and eventually patron household if one established one's own branch family within the old
lineage, or beyond if one broke off from it to establish a separate one. Much of this classification was based on ownership of
house and land, but ownership was a very fuzzy notion, as we shall see. Additional factors intervened. One was status, which
was not unambiguously connected with economic wealth. A second factor was political power in the village as lodged in the
village council and the lineages. A third was whether or not one paid one's tribute portion of the village quota directly or not.

All these factors interacted, and there were no strict rules that applied automatically. Moreover, if tradition or custom were
invoked, the question often became whether or not the particular norm applied, as Wittgenstein has acutely observed. Hence,
disputes led frequently to suits settled either through conciliation, often by Buddhist priests and elders from neighboring
villages, or by an intendant. The supporting data in these suits were official documents. Composed for the lordly purposes of
controlling population and extracting economic surpluses, these documents were used politically within the villages in status
disputes.

Disputes concerning status, position, and hence power in the village as expressed in these documents often had a certain
degree of urgency added to them when one party refused to sign the population register that was just about to be turned over
to the intendant (as we saw in Ken's case, in chapter 1). The refusal to sign under one's entry because of the status attached to
it drew the higher authorities into these disputes. The examples of this strategy that follow are from the eastern part of
Shinano province, mostly from the Kita-Saku district.

The first case is a lawsuit from Shimo-Kaize (1687-88) submitted to the intendant by three plaintiffs, Tokubei, Kichibei, and
Hikozaemon, against their patron, Hanbei, concerning their disputed status as fully established branch houses (kakaeya) or
semiautonomous branch houses (kadoya). There are six documents: the original suit; an official written response, which the
defendant was invited to submit to the intendant according to custom; and four additional recorded testimonies (kuchi-

― 151 ―

gaki ) by the headman, the tided peasants, and other witnesses, the last three dated eight months after the suit was initiated.
[47] The texts that follow are edited translations, modified for clarity.

1. Tokubei, Kichibei, and Hikozaemon's suit (1687/7/11)

1. In past years, we were recorded every year in the population and kumi register as Hanbei's fully established
branch houses (kakaeya ), but this year Hanbei registered us according to a new procedure as [only]
semiestablished branch houses (kadoya ). We had a conference with Hanbei and told him that we would not affix
our seal to the registers unless we were entered as kakaeya as in the past. Then the headman and the village council
urged us to affix our seal; otherwise if the entries were changed, the rosters could not be presented to the shogunal
authorities [in time]. Since we did not wish to go as far as to cause trouble to the authorities, we affixed our seals.

Twice we went to the higher officials about our registration in the population roster as kadoya, and we explained
our ancestry to them. They suggested that Hanbei change the entries back to kakaeya. Now, because of the selfish
(wagamama ) behavior of Hanbei, we request to change kumi away from Hanbei. Please summon him and
communicate our request.

2. The three plaintiffs and the defendant share a common "ancestor" three generations back through their "great-
grandfather," Umanosuke [the three plaintiffs were related through Umanosuke's adopted son, originally the son of
an uncle (see fig. 4)]. When Umanosuke retired, he gave half of his 44 koku to his son Goemon, keeping the other
half for himself. At his death in 1645, this latter portion was divided into three between Tokuzaemon (Goemon's
son), Shige'emon (Umanosuke's adopted son), and Shin'emon (Umanosuke's son by a woman other than his wife).
This was witnessed by two villagers (Daitokuin and Chuemon[*] ) and a peasant from the new branch village
(Jiroemon[*] ).

3. Shiroemon[*] and Bun'emon were at Umanosuke's funeral, but all memorial services down to the thirty-third year
were conducted by the three inheritors (Shige'emon, Tokuzaemon, and Shin'emon), to which people in the village
can testify. Now, Hanbei, in his selfish manner, maintains that because we are within the (undivided) 44 kokudaka,
we are only kadoya. But Bun'emon, Shiroemon[*] , Shige'emon, Shin'emon, and Tokuzaemon, all five of them
(individually), have a part of this 44 koku and have contributed the yearly tribute and corvée separately according
to their taka.

[47] The six documents of this case can be found m NAK-KS2 (1): 946-52 (no. 476). The first three are also reprinted in Aoki
Koji[*] , Hennen hyakusho[*]ikki shiryo[*]shusei[*] , 17 vols. (San'ichi shobo[*] , 1979-93), 1:568-71.

― 152 ―
Fig. 4.
Umanosuke's Lineage, Shimo-Kaize, 1630s-1680s
a present at Umanosuke's funeral m 1645
b inheritors of Umanosuke's 22 koku and sponsors of all memorial services

Note: Based on the plaintiffs' argument in paragraphs z and 3 of their suit.

Some data in this chart, especially the relationship of the plaintiffs to the lineage,
were challenged later in the sum Moreover, Shin'emon had a son by
the same name. Bracketed numbers represent koku.

We testify to the truth of the above and request that Shin'emon and Tokuzaemon be summoned and that you
inquire with them about this matter, and that you summon Hanbei and reach a decision.

1687/11
Shige'emon's children from Kami-Kaize,
Tokubei
Kichibei
Hikozaemon

The plaintiffs build their argument about their status as full-fledged branch houses on genealogy, or more properly linealogy
(as we shall see via Shige'emon), and on the performance of memorial services. They maintain (against Hanbei) that the
original estate was divided into separate tribute units. Thus, as "children" of one of the owners (Shi-ge'emon), they cannot be
considered Hanbei's dependents (kadoya). As full branch houses, they would not be on an equal footing with Hanbei, their
patron, but as such they certainly would be full lineage members in their own right.

― 153 ―

Hanbei's counterargument in the reply requested by the intendant follows. He opens with an attack on the alleged linealogy
and responds to the allegations one by one in great detail. Without the necessary local knowledge of the kin and power
relations, which constituted an invisible subtext, the detail must have been as confusing for the intendant as it is for us.
Reading Hanbei's reply, one can thus put oneself without much effort into the sandals of the intendant; reader positions are
always distant ones, though not always equidistant.

2. Hanbei's Response to the Intendant (1687/12/4)

1. Re : The reasons why the three plaintiffs are [only] kadoya. Their grandfather was Bunzo[*] [and not Umanosuke],
who had an older brother, Bunshiro[*] . My great-grandfather Umanosuke bought them as lifelong bond servants
(fudai ), and hence they were supposed to serve him forever. However, because they had served since their youth,
they sued for their freedom, which was granted them, and so they were established under Umanosuke as his
kadoya. In addition Umanosuke gave them the land they had brought under cultivation during their service years.
Bunshiro's[*] corvée obligation as kadoya was twenty days of labor every month; Bunzo's[*] was fifteen. Bunshiro[*]
married Tsuru, one of Umanosuke's bond servants [gejo , the term for a female genin], having two daughters by her:
Miya, who left Umanosuke's household to work in a temple when Umanosuke's wife died, and Musu, who bought
her freedom with cash and married someone from another place. When Bunshiro[*] died, he gave his land m
Bunzo[*] , who became a kadoya.

2. This year, at the transplanting of the rice seedlings (taue ), only Kisaburo's[*] [another recalcitrant kadoya, who
was not one of the three plaintiffs] and Tokubei's wives showed up, while they themselves did not come and help [as
they should have as kadoya]. So I went to Chuemon[*] , the headman, to officially request that he inform the higher
authorities and seek a decision, to which Chuemon[*] replied m wait until he got more information about the
matter. After some postponement, Chuemon[*] , Zenzaemon, and Heizaemon [the headman and two village officials,
presumably] communicated that the men in question offered their apologies for not showing up at the
transplanting of the rice. The same three officials also offered some sake with their own apologies, which I accepted,
and I was asked to forget the whole thing, which I did.

3. To record them as kadoya is not a new practice. In 1644, under intendant X [named], the population registers and
household rosters recorded Yaemon [?] as Umanosuke's kadoya and Shige'emon as Shiroemon's[*] . [Here
Shige'emon, of crucial importance to the plaintiffs, is argued by Hanbei to have been a kadoya, certainly not on an
 equal footing with the other four inheritors of the estate.] These documents, written by Daitokuin with all seals of
the village attached after a full

― 154 ―

meeting, list the division of fields. The following year, Umanosuke passed away, but because Shiroemon[*] was a
grandson-in-law, he could not inherit the kadoya [Shige'emon, who had been his kadoya thus far].

Subsequently, under three different intendants [named], the entries are still kadoya, but as headman [Hanbei had
been headman, as Umanosuke had been] I had difficulties with that. In the current population register submitted to
the intendant, there are also kakaeya entries. However, given the fact that these persons are in that situation [of
dependency toward me] and cannot but obey me, ultimately it does not matter how they are registered.

Now, you may ask, why [after just saying that the label does not matter] I registered them, nevertheless, on this
year's registers as kadoya. The reason is as follows. A decision was made in the village council to [upgrade and]
register as kakaeya relatives (uncles, brothers, etc.) of full kumi members [only titled peasants were full kumi
members]. So we registered them all equally as kakaeya. The above persons, however, have been kadoya for
generations, and they are therefore not like the others and should not be registered as if they were their equals;
hence their registration as kadoya.

4. Re : The signing of the goningumi register by these three as kadoya. This whole thing was triggered by
Kisaburo's[*] refusal to sign as kadoya, and the dispute was brought before the headman, who came to me. I stressed
to him again and again that I could in no way remove the kadoya entry from Kisaburo's[*] name, and I also gave
him reasons why the others, Kichibei, Hikozaemon, Tokubei, and so on, should remain registered as kadoya and
told him that each of them should affix their seals in the presence of the others. So the headman called them all and
explained to them in detail why they should be registered as kadoya, to which they all agreed, and thus they affixed
their seals to the goningumi roster. This is the truth, which you can verify with the headman and the titled peasants
of the village.

5. Re : The adoption. The 22 koku of the retired Umanosuke were inherited by Tokuzaemon, Shin'emon, and
Shige'emon. The land division was witnessed by Jiroemon[*] , Chuemon[*] , and Daitokuin. You can ask the latter two
and have them testify.

6. Re : The funeral and memorial services. Because there was no unhulled rice for the funeral, my father,
Shiroemon[*] , who was employed by the intendant's assistant (tedai ), Ichikawa Gohei [from Gorobe-shinden?],
borrowed six bales from the storehouse. All other expenses were divided into two by Tokuzaemon and Shin'emon,
and they were also responsible for the memorial services from the first through the seventh year and did not collect
any contributions from Shige'emon. [However,] after that, Shin'emon (Umanosuke's son by a woman other than his
wife) and my uncle Bun'emon did not set them [Bunzo[*] and Bunshiro[*] ] free. Their only proof that they were set
free is that they, as life-

― 155 ―

long bond servants, made a request to be set free, but what they said [about this] to the authorities is a lie, which
has been the cause for a longstanding grudge and quarrels between us.
As for the remaining memorial services, I do not know about Shin'emon, but as far as my uncle Tokuzaemon is
concerned, because he was only four when Umanosuke died [he did not contribute], my father, Shiroemon[*] , took
care of all the services.

7. Re : The holdings of Kichibei, Hikozaemon, and Tokubei. On the cadasters of the 1629 land survey, Bunshiro[*] and
Bunzo[*] were recorded as Umanosuke's dependents [co-residents?] (uchi ). Next, on the cadasters of the 1676 land
survey, Shigehei, Hikozaemon, and Kisaburo[*] were recorded as Shiroemon's[*] dependents (uchi ).

8. Re : The rewriting as kakae on the new population register. There was a dispute concerning signing the new
register under the heading of kadoya. The matter was brought before the two intendants [named], and their opinion
was that if I by all means insisted on their being kadoya, this would lead to an endless dispute and prevent the
registry from being submitted in due time. Therefore, the headman and the titled peasants consulted with each
other, and they said that it did not matter what was entered as long as the registers were submitted in time, and so I
changed the entry and submitted it with my seal.

1687/12
Hanbei

Hanbei's argument is very complex, but I shall limit myself to the main points. The hierarchical relationship between fudai,
genin, and kadoya is clearly stated. Fudai are bought, usually at a very young age from outside the village. Prices for fudai are
rarely mentioned in documents, but in Hirabara, a few kilometers to the north, a ten-year-old boy was bought for three bu
(worth about 1.5 koku of rice on the market), and a four-year-old girl for six bales (about 3-6 koku of unhulled rice) in the
1650s.[48] They were virtually house slaves for life unless their

[48] The prices are from Ichikawa Yuichiro[*] , Saku chiho[*] , 119; the equivalent in rice for the gold price is calculated from
Ono Takeo, Edo bukka jiten , Edo fuzoku[*] zushi, 6 (Tenbosha[*] , 1989), 451; the bales of unhulled rice are calculated at the
rate of one bale per 0.6 koku (NAK-T 5, "Furoku," 4). Brutal as this practice may seem, these children were seen, perhaps all too
euphemistically, as rescued (quite a few of them had the name Kaisuke, "bought and rescued" [Ichikawa Yuichiro[*] , Saku
chiho[*] , 119]). They were certainly given a better livelihood than children who are recruited into prostitution for foreigners in
Southeast Asia; nowadays recruiters exchange television sets for children, or they buy them still in the womb (Murray
Kempton, "A New Colonialism," New York Review of Books 30, no. 19 [1992]: 39).

― 156 ―

freedom was purchased, as in the case of the two female genin. "Freedom," the transition from fudai to kadoya, meant that
they were "given" the land they themselves had brought under cultivation, but most of their labor was still for their former
owner, who also paid the necessary tribute on their land. The refusal by some of the kadoya to help out with transplanting the
rice seedlings at Hanbei's signaled that they did not consider themselves kadoya owing labor to a boss. In addition, Hanbei's
case rests on documents that testify to their kadoya status. As in the Niremata case, a clear line was drawn between
dependents who were relatives of titled peasants and those who were not (most kadoya).

Another source provides social background data on the stratified minisociety of Shimo-Kaize village. In 1643 Shimo-Kaize's
population of 218 lived in thirty-one main homesteads, three kado dwellings, and eleven long houses. Three households each
had 1 "dependent," and thirteen houses had a total of 30 servants (male and female).[49] Eleven of these servants were
recorded as hikan (a rather pejorative term for "serf," common in pre-Tokugawa times). The village kokudaka was 336 koku,
averaging 10.8 per household; the largest holding was 22 koku, the smallest 2.3. Thus we know that Umanosuke was the largest
holder of the village, since he had 22 koku after he retired. He also had a rather large house, because his must have been one
of the three houses of 27 tsubo (930 square feet, or 90 square meters). The kadoya, on the other hand, were housed in buildings
that varied from 425 square feet, to half that size. Most of the long houses measured only some 210 square feet.[50] By the time
this suit was filed, however, many of the kadoya must have already been promoted to kakae as a result of the gradual breakup
of the larger holdings.

Next, the defendant Hanbei submits a recorded testimony that is certified as true by the headman.

3. Hanbei's Recorded Testimony, Certified by the Headman (1687/12)

1. Re : The question that the three were kadoya for generations but now claim that they are kakae. They filed a suit
concerning this dispute. In my response, I claim that my great-grandfather Umanosuke bought Bunshiro[*] and
Bunzo[*] , that they should have worked for him forever, but that they were set free[, etc.]. Now, when I investigated
who received this land that they had developed and what Bunjiro[*] and Bunzo[*] were in rela-

[49] These data are from Ichikawa Yuichiro[*] , Saku chiho[*] , tables 2, 8, and 9 (77, 100, 101).

[50] Ibid., 78, 79, 119, 124.

― 157 ―

tion to the three plaintiffs, I found that the two cadasters had entries as I claimed and that Bunzo[*] was Tokubei's
oyabun [real or relational parent or patron], Hikozaemon's father-in-law, and Kichibei's grandfather; and that this
Bunzo[*] was the one whom the three suitors claim m be Shige'emon. Bunshiro[*] is Bunzo's[*] elder brother. Now,
who bought Bunzo[*] and Bunshiro[*] as fudai, and is there proof that the two performed service for the indicated
days per month? They were bought by Umanosuke from X [named]; there is no proof that they performed the
alleged service, but everybody in the village knows that they did.

2. Re : The yashiki [homestead and compound] of the three plaintiffs. Is it in my yashiki or is it separate [an
important question to determine their degree of autonomy]? Until forty-three years ago, Bunzo[*] and Bunjiro[*]
were put up on Goemon's yashiki, which was part of the inheritance of 22 koku he received from Umanosuke. But
there were difficulties with building a [new] residence for them. Luckily, there was an empty building in a separate
yashiki, which is where they were put up, and that is where the three plaintiffs live now. As far as the yearly tribute
is concerned, until twelve years ago, the 44 koku were recorded as one registered holding, but they had informally
(naisho ni ) divided it among themselves, and each individually paid tribute as allocated by me. Because this was
difficult, the headman divided the yearly tribute as well as the corvée into separate parts.

1687/12
Hanbei
I certify that the above is true
Yahei, Headman

Yahei, the new headman who replaced Chuemon[*] , vouches for Hanbei's version of the three plaintiffs' ancestry: they are
mutually related and tied to a kadoya. Equally important are the location of their living quarters (whether they are within or
outside someone's compound—he confirms that they now live in Hanbei's grandfather's compound) and their independent tax
status. The latter seems to indicate that the village headman had treated them as independent tribute payers.

These three documents were apparently insufficient for the intendant to make up his mind on this question of status, because
seven months later three additional recorded testimonies were produced, the first one by the headman and the thirteen titled
peasants.

4. Recorded Testimony by the Headman and the Thirteen Titled Peasants (1688/7/11)

1. We have been asked whether the three plaintiffs in question are Hanbei's kakaeya or his kadoya. Since we are
unaware of circumstances long past, we do not know whether they are one or the other. They were listed separately
and they paid tribute and corvée separately.
 ― 158 ―

2. The three allege that their grandfather Shige'emon had been adopted by Hanbei's great-grandfather Umanosuke
and that they were attached to the portion of the land on which Umanosuke retired and which he passed on to
Shige'emon. Now, is there certainty that Shige'emon was Umanosuke's adopted son? It is hard to tell whether he was
his adopted son or his kadoya because it is too long ago and the situation is not dear. Of course, as far as the division
of Umanosuke's 44 koku is concerned, the three plaintiffs have today a portion thereof, but on what basis this was
divided, that we do not know.

3. Now the question: Did the three plaintiffs pay their tribute and corvée separately and have they been listed
separately on the name registers since Shige'emon's generation or not; and if not, then since when and for how
many years have they been paying taxes separately? Afterwards [after Shige'emon's time] they made all their
contributions, tax and corvée included, together with Hanbei, but they delivered separately the taxes and corvée of
1676, twelve years ago. Now, if further asked on what grounds this separate payment of twelve years ago took place
or how the division of land occurred, we do not know, but Hanbei was headman then.

4. Hanbei alleges that the three plaintiffs are without a doubt kadoya because every year since his great-
grandfather's generation down to his own they performed kadoya corvée and the villagers all know that. When
investigating the truth of Hanbei's allegation, we found that although we do not know what will happen now, every
year at the time of the transplanting of the rice seedlings, the three, including their wives, without exception helped
on Hanbei's paddies. This is what Kichizaemon said. As for others who may have helped on that occasion, there
were others, although we do not know their number.

5. Hanbei alleges that forty-four years ago, Shige'emon was recorded as kadoya on the population and house
registers, documents certified by Daitokuin and separately by the whole village with the seals of the peasants.
Because this was so long ago, we were not there when these documents were written, but [when we checked them]
we found that it was Daitokuin's writing and that all the peasants certified those documents with their seals.

6. Hanbei alleges that "under the next three intendants, the three plaintiffs were always listed as kadoya; but that
under the last intendant they were also sometimes registered as kakae; ultimately, it did not matter how they were
registered since they were his dependents, who had to obey him." Is this true? In those days population registers
were compiled only every five or ten years. Hanbei in the end could not remember for sure, and he and Zenzaemon
admitted not knowing whether it was kakae or kadoya.

1688/7/11
Yahei, Headman
Thirteen titled peasants

The village leaders are unwilling to pronounce on the status of the plaintiffs. They state, however, that the trio was attached to
Shige'emon's

― 159 ―

land (and not Goemon's). Thus they corroborate the plaintiff's alleged filiation to Umanosuke via Shige'emon. The question
then became whether Shige'emon was an adopted son or a kadoya. They say that they do not know because it was too long ago
but that documents exist that could verify this. What is puzzling is that the plaintiffs seem to have become independent tribute
payers, at least informally, when Hanbei was headman. His argument, however, clearly rests on his assumption that no matter
how they were registered, they were his dependents.

Next is a recorded testimony (with the headman's certification) by Shin'emon, whose father, of the same name, was
Umanosuke's son by a woman other than his wife.

5. Shin'emon's Recorded Testimony, Certified by the Headman (1688/7/11)

1. My father, Shin'emon, is Umanosuke's son by another woman. Of Umanosuke's 44 koku, 22 went to Goemon, son
by his wife; the other 22 went to Shige'emon, Tokuzaemon, and myself. Shige'emon, Tokuzaemon, and I attended
Umanosuke's funeral. Furthermore, Umanosuke's 44 koku are now in the hands of Bun'emon, Shiroemon[*] ,
Shige'emon, Tokuzaemon, and myself; we pay tribute and corvée separately according to our taka. I looked into this;
also into what the three plaintiffs said. My father, Shin'emon, died ten years ago, and because he was non compos
mentis (fuchoho[*] ), he did not tell me anything. As to the division of the 22 koku, I do not know of a document to
that effect, but Shige'emon, the oya [parent or boss, patron] of the three plaintiffs, received 5 sho[*] [0.05 koku] of
paddies and 5 sho[*] of buckwheat dry fields from Umanosuke when the latter died. At the funeral were
Tokuzaemon and myself. The three plaintiffs did not help. Tribute and corvée were paid separately.

2. This is my pedigree. Hence I have been asked whether the three plaintiffs are Hanbei's kadoya, and since I line up
in the lineage at the same [generational] level as the three plaintiffs, I was requested to talk about the times since
Hanbei's great-grandfather Umanosuke, on the supposition that I was knowledgeable about them. But as mentioned
earlier, since my father was non compos mentis , he did not transmit to me any details. However, in Umanosuke's
generation, the three plaintiffs' oya, Shige'emon, worked for Umanosuke fifteen days a month.

1688/7/11
Shin'emon
Certified as true by Yahei, Headman

Everyone is evasive; collective memory seems blocked, this time through the state of non compos mentis of a conveniently
dead man who while still alive might have casually provided his son with the right information. But he did not. A certain labor
dependency of Shige'emon

― 160 ―

vis-à-vis Umanosuke is acknowledged, but this dependency is not spelled out as one typical of a kadoya. Everybody seems to
sympathize with the plaintiffs by not providing the crucial information that would otherwise nail them.

Finally, there is a recorded testimony by two of the three men who arranged Umanosuke's inheritance: Daitokuin (a Buddhist
priest, judging by his name) and Chuemon[*] .

6. Recorded Testimony by Chuemon[*] and Daitokuin (1688/7/11)

1. This concerns the three plaintiffs' allegations that: they are not kadoya; their oya, Shige'emon, is Umanosuke's
adopted son; the 22 koku of the retired Umanosuke were divided into three and given to Tokuzaemon (son of
Goemon, Umanosuke's son by his wife), Shin'emon (son by another woman), and his adopted son Shige'emon; this
division is well known to Daitokuin and Chuemon[*] from this village and Jiroemon[*] from the branch village. We
were summoned and asked to investigate.

Forty-four years ago, still alive, Umanosuke told us that he wanted the 5 sho[*] of paddies and dry fields each to go to
Shige'emon. He died that year. In the spring of the following year the three of us got together and made the transfer
m Shige'emon, but the talk about dividing the 22 koku into three is a lie. As to the adopted status of Shige'emon, the
three plaintiffs alleged that we knew that well, but we do not know whether he had been adopted or not. We have
not heard rumors that he was adopted or that he was a kadoya.

1688/7/11
Chuemon[*]
Daitokuin

Almost a year after the suit began, in a testimony by men who knew Umanosuke, it turns out that Umanosuke's property was
not divided into three as alleged all along. But at the same time, they are ignorant of Shige'emon's adoption or his kadoya
status! It is striking how just about everything is a matter of dispute: the status of kadoya or kakae, the relationship of
adoption, the question of inheritance, the meaning of joint versus separate tribute payments.

We do not know the outcome of this suit, but it seems likely that the result somehow favored the plaintiffs. This absence of a
clear settlement, which would result in one side's admitting defeat, is rather typical, as we shall see. One also senses the
lineage tensions between Hanbei's side and Shin'emon's; village politics were often lineage politics. Overall, one also has the
impression of a status system in transition.

― 161 ―

From Tribute Payers to Full Peasants: Kodaira, 1629-1849

Titled peasants are often understood to be descendants of peasants entered on the early land surveys as landowners and
tribute payers. Ancestry and ownership would thus be necessary to qualify for the title of "titled peasant." Accordingly, one
would expect that as the population grew (especially during the seventeenth century, when the growth rate was 50 percent)
and the number of kin-related small landholders that branched off from the original households increased, the number of
titled peasants also increased. On the other hand, if many new peasants came from the ranks of non-kin dependents (who
were not descendants from original titled peasants) and ancestry was more important than landownership, one might expect
nontitled peasants to outnumber titled peasants. Honbyakusho[*] , "titled peasant," is specifically a title, and its rarity points to
privilege and limited access. Ancestry, however, was only one way to anchor and preserve this privilege against other
claimants, and it was not an absolute criterion: it was only one strategy among others to secure distinction and status.

Thus, a number of questions have to be answered with regard to the status of titled peasant. Why did the number of tided
peasants tend to become limited? Whence the rarefaction? What criteria controlled entry to this status group—ancestry,
ownership, wealth? And did the overlords or the village council control these criteria? How did lineage power intersect with
the power or prestige individuals derived from the fide? Was this symbolic capital—which within the village was political
capital—immune to devaluation (could one lose one's status as a titled peasant) or marketable (could one sell or buy the status
as such)? Could one transfer the title independently from the land, or were the two inseparable? I touched upon some of these
questions in chapter 1. Here I shall discuss them more fully.

In discussing the original titled peasants, identified as homestead owners in the early registers, I have described their function
in rather vague terminology: they were accountable for the tribute, responsible for it, and they channeled it. This language is
intentional, because non-titled peasants also produced and contributed to tribute and corvée. The original titled peasants were
clearly the economically wealthier peasants, if for no other reason than that they had homesteads. The basis for the title was
thus economic and rested on ownership not sim-

― 162 ―

ply of land but of homesteads as well. By the same token, it was also political, since it was the local elite that became titled. The
term "titled peasant" thus has two sides, an extramural and an intramural one, the former having to do with responsibility for
tribute and corvée, the latter (both reinforced by and the basis for the former) with local status and power.

The early titled peasants managed land, not uncommonly estates of more than one hundred koku, and a number of
dependents co-residing on the homestead as well. Gradually these estates were divided and parceled out to dependents, who
thereby became economically independent to some degree from the original owner, who nevertheless retained his
preeminent position within the expanding lineage by registering the various households (ie) and their land as being part of
(buntsuke) or within (uchizuke ) the main holding. Often, but not always, the former referred to kin (affines, descendants, and
relatives), the latter to non-kin.[51] Genealogically, or perhaps more precisely, linealogically , these new units were branch
houses of the main house.

In Shimo-Kaize, as we have just seen, there was a great difference between kadoya and kakae, which I referred to as semi-
established and fully established branch houses, respectively; and the trend was "emancipatory" from kadoya to kakae. The
evolving relationship of these kakae to the titled peasants can be described as follows. The number of titled peasants increased
until the last decades of the seventeenth century, when they appear to have reached a quota in most villages that was
maintained for about a hundred years. During that century the number of kakae grew, but then it decreased and by the end of
the period, in the 1860s, they had virtually disappeared, having become titled peasants.[52] These turning points in the
trajectory of the relationship between kakae and titled peasants need explanation. Why a growth in the numbers first of titled
peasants, then of kakae, and then once again of titled peasants?

To follow this trajectory, the struggles that determined it have to be analyzed and the question of status that constitutes their
center

[51] Ichikawa Takeji, "Kakaebyakusho[*] ," 6.

[52] For instance, in Ozawa village of Saku district in 1704, 33 percent of the peasants were kakae, but in 1864 all kakae had
disappeared. The same drastic drop occurred in nineteen out of the twenty-five villages that constituted the Tanoguchi bakufu
fief, of which Ozawa was a part (ibid., tables 3 and 4 [7, 8]).

― 163 ―

addressed. With this in mind, we shall follow in some detail the history of Kodaira village (presently part of the town of
Mochizuki in Kita-Saku district) as presented to us by Ozaki Yukiya.[53] In 1647 its assessed yield was 175 koku, making it a
small village.

In 1629 Kodaira counted thirteen peasant owners of residences (yashiki), making them "tided peasants" (see table 14) In 1643
there were twenty: the original thirteen plus two branch houses, all holders of between 8 and 21 koku, and five others with
holdings between 4 and 8 koku as yet without homesteads. Since the latter five were entitled to a yashikibiki , an exemption of
tribute owed on their residence, they must have been titled peasants as well.[54] In addition, there were two other peasants
with less than 1 koku who differed from the other twenty in that they did not enjoy the exemption; they were nontitled
peasants.[55] The number of titled peasants thus increased from thirteen to twenty in about a dozen years.[56] Two hundred
years later, in 1849, there were twenty-one Kodaira peasants marked in the registers as osabyakusho[*] ("head peasant"), a
variant of honbyakusho[*] . Nothing seems to have changed, yet this longue durée is filled with colorful events.

The 1677 population roster shows clear status divisions. The peasant population, aside from one outcaste, then comprised
twenty titled peasants (including seven patrons), thirteen clients (including three kadoya), and a number of bond servants
(fudai and genin). Economically, however, the members of the two main groups are not always distinguishable. Although all
holders of more than eight koku are titled peasants, some clients have equally large or even larger holdings and bond servants
as well. The three clients and the one outcaste have holdings larger than that of the smallest titled peasant (under 1 koku).
Nine of the clients, however, have no holdings at all (mudaka).

Over the next twenty years two great shifts occurred. In 1694 one titled peasant, a sake brewer (whose descendant is still in
business

[53] Ozaki Yukiya, "Kinsei sonraku naibu no mibun kaiso[*] ni tsuite: Shinano-kuni Saku-gun Kodaira-mura kakaebyakusho[*]
mondai o chushin[*] ni," Nagano 29, no. 8 (1977): 752-68, no. 9 (1977): 815-30.

[54] In Matsumoto domain this amount was 1 koku (Naito[*] , Honbyakusho[*]taisei , 85).

[55] Although the five holders of 4 to 8 koku had no yashiki they were nevertheless granted the yashikibiki , which means that
they must have been (new) titled peasants (Ozaki, "Kinsei sonraku," 755).

[56] Ibid., table 1 (755).

― 164 ―

Table 14.
Kokudaka and Peasant Status, Kodaira, 1629-1849

1629 1643 1677 1694 1698 1731 1765 1849

Koku H K H K H K H K H K H K M H K M H h K M

120 1
98 1

80 1

54 1

25 1

24 1

23

22 1

21 1 1

20 1

19

18 1

17 1

16 1 1

15

(Table continued on next page)

― 165 ―

(Table continued from previous page)

1629 1643 1677 1694 1698 1731 1765 1849

Koku H K H K H K H K H K H K M H K M H h K M

14 2 3 1

13 1 2 1 1 1 1

12 3 1 1

11 1 1 1

10 3 1 1
9 1 1 1 1 1

8 2 1 1

7 1 1 1 1 1 1

6 1 1 1 2 1 1 2 2

5 1 2 1 2 1 1 3 1

4 2 1 3 1 1 1 2

3 2 2 1 2 1 3 4

2 1 1 3 8 2 11

1 1 1 4 4 1 4 8 4 6

0.5 2 1 1 1 1 3 11 2 7 11 1

0.1 4 1 5 16 1 10 19 1 8

0 9 27 1 15 3 3 1

Total 13 20 2 20 13 20 36 21 47 21 40 21 45 21 48 2

SOURCE : Compiled from Ozaki, "Sonraku mibunsei," tables 1, 2, 4, 7, and 9 (pp. 755, 758, 763, 822, and 830).

NOTE : H = Honbyakusho[*] ; K = Kakae; M = Mizunomi; h = hyakusho[*]

― 166 ―

today) had become very successful, for he had 80 koku, which put him far ahead of the next two holders, with 25 and 13 koku.
The situations of some of the kakae had also improved: the nine taka holders, the largest with 11 koku, each had more land
than sixteen of the twenty titled peasants, but there were now also twenty-seven landless clients. The total number of clients
now stood at thirty-six. By 1731 the picture has changed again. The sake brewer, had nineteen servants and had acquired
another 28 koku, bringing his holding to almost 100 koku. But twelve of the now twenty-one titled peasants had only x koku or
less. Now, there were forty clients, of whom all but one had land. This does not mean, however, that there were no landless
peasants (mizunomibyakusho[*] ): there were eighteen of them, one of whom was a client.

Ozaki does not detail the composition of the goningumi, but it is likely that the kumi were structured along lineage lines. It is
important to note as well that membership in the goningumi was limited to titled peasants, twenty-one households out of a
total of seventy-eight.[57] As we shall see, lineages played an overwhelming role in village politics, one that has not received
much scholarly attention. As lineages grew, the number of temples grew from one temple of z koku to three, one with less than
half a koku, the other two with 5 and 16 koku. It is likely that these temples serviced nor simply "the village" but the lineages.

We should take a closer look at who the titled peasants and clients were. A comparison of three population registers
(shumon[*] aratamecho[*] ), for the years 1694, 1695, and 1698, reveals numerous and frequent changes in the identity of titled
peasants and clients. Between 1694 and 1695 the number of titled peasants increased by one (to twenty-one); however, this
new titled peasant, who had been promoted from the status of client, became a client again three years later, but under a
different patron. In another case, a patron changed places with his client, the latter now assuming the position of lineage
patron and titled peasant. Another titled peasant went into service outside the village for

[57] This was also the case in Tenjinbayashi, a neighboring village. In 1699 there were nine goningumi in Tenjinbayashi, each
counting exactly five members (rifled peasants), except for one group that had only four members. Together with the
headman, this adds up to forty-five rifled peasants; nineteen of them were patrons for a total of thirty-two kakae. In addition,
there were six servants and four outcastes (see Ozaki Yukiya, "Mochizuki-machi no buraku no rekishi [1]," Mochizuki no
burakushi 1 [1975]: 53).

― 167 ―

several years and gave up his title. Other cases in 1698 show a similar instability in the membership of both ranks. One titled
peasant passed his title on to his client when he left the village on a service stint. One Jihei who in 1695 was a client turned up
three years later as not only a titled peasant but also a patron of three clients: one was Jihei's former fellow client; one was
Sajihei, a titled peasant who had two clients in 1694 but only one in 1695 and was himself Jihei's client by 1698; and the third
was the successor to Sajihei's former client.

The situations of a number of titled peasants were precarious, and their clients stood ready to take their place. (By now the
titled peasants constituted a numerus clausus .) These clients were often second and third sons and in some rare cases bond
servants. The fiction that they were all born within the village, however, was maintained; although most bond servants came
from the outside, when they were promoted to client status they were recorded as "born in the village." Also, as we saw in
chapter 1 to be the case in Makibuse, none of Kodaira's approximately thirty bond and indentured servants were married.
Either they could not afford a family (a very widespread phenomenon beyond Japan also)[58] or they were not allowed to have
one (they needed their patron's approval to take a spouse).[59]

In the 1702 population rosters of the neighboring village of Kasuga, one looks in vain for clients. Instead, there are a good
number of mudaka, peasants without taka, recorded as doing "internal work" (naisaku ) for their parents or brothers. They
had families and lived in separate quarters and were thus kin, established as branch families who officially had not been
granted land by their main family.

What qualified one to be a titled peasant, and why was their number restricted at a time when more and more branch houses
were being created? Economic power obviously was not the decisive factor: quite a number of clients held larger holdings
than most of the titled peasants. There was no upper kokudaka threshold that automatically made one a titled peasant or a
lower one whereby one lost one's title. At the same time, all of the clients who were promoted held at least close to

[58] I referred earlier to Pierre Bourdieu's study of the Béarn region in France. I still remember the live-in stable hands on the
farms of the village in Belgium where I was born, all bachelors who were never taken seriously and bore the brunt of many
jokes and pranks.

[59] Minegishi, Kinsei mibunron , 140-42.

― 168 ―

five koku. Unlike the holding, the title usually could not be divided (although in Niremata it was). Hierarchical lineage
structure seems to have determined all important relationships within the village. In society at large the great divide was
between samurai and nonsamurai, in the village between rifled and nontitled peasants; both the nonsamurai and the
nontitled peasants were subjects of governance. The nature of status in samurai society differed from that in peasant society.
Whereas samurai status adhered to all descendants of samurai, for long periods of time during the Tokugawa period the
number of titled peasants was limited; only one of a titled peasant's sons could inherit the title—which was also the case with
samurai offices . The title of "titled peasant" thus seems to have functioned not simply as a marker of status but also as a title of
office. And initially this was the case, since titled peasants were created and set up as such to provide the overlords with the
goods and services they expected from their subjects. But why its scarcity, given that many nontitled peasants obviously
provided the same function? Were there also extramural influences at work in the limitation of the number of titled peasants?

Naito[*] Jiro's[*] answer to this question, which is perhaps too one-sided, is that the titled peasants were those who had a
hyakusho[*]kabu , or "peasant share."[60] In the beginning only the "titled peasants" (homestead owners) had such a share.
While this may sound tautological (titled peasants are those who have a share in the status of peasant as opposed to those who
do not), the term share , as we shall see shortly, was used very early on to refer to title holders.

In accounting for the nearly universal phenomenon of the continued restriction of the number of title owners, Naito[*]
concentrates on extramural factors, thus ignoring the increase in the number of titled peasants in roughly the first half of the
seventeenth century. He cites two cases where daimyo legislated explicitly against an increase in the number of "peasant
shares," one of them using the term kabu .[61] In bakufu legislation there is no explicit reference to limiting the number of
rifled peasants, but many bakufu-initiated village laws (goningu-

[60] Naito[*] , Honbyakusho[*]taisei , 28.

[61] The two cases revolve the Kaga domain and the village law of Iwamoto village in Echizen (ibid., 94-95). The latter
mentions the term kabu and is a document that dates from between 1601 and 1607. For the full text of this document, see
Miyagawa, Taiko[*]kenchi-ron , 2:304. Another case is Ueda domain (see NAK-T 4:464).

― 169 ―

micho[*] zensho) mention the need to provide successors to bankrupt peasant houses.[62] Naito's[*] interpretation of this
injunction seems plausible. He argues that if shogunal policy had simply aimed at establishing small landholders, there would
have been no need for such encouragement: there were plenty of small landholders with lands and homesteads who could
have succeeded to bankrupt titled peasant houses. This injunction makes sense only in a context of limited access to the status
of titled peasant.
Another extraneous influence was undoubtedly the bakufu's 1673 prohibition against dividing land below a certain size: For
peasants the minimum size was ten koku, and for headmen, twice that amount.[63] This law was issued around the time that
the great riparian works of the seventeenth century were coming to an end. In many regions the acreage had been increased,
so that an expanding population could be absorbed. Thus the further creation of branch houses would result in smaller
holdings, and often, for example, in small mountain villages, even in a reduction of acreage of arable by the amount of land
taken up for new homesteads. In many areas the branch houses established during this period of expansion became titled
peasants.

In the registers in Ueda domain, a few kilometers to the north of Kodaira, one finds clients until around 1624, when they begin
to be entered on the name rosters as equal to the titled peasants, and by 1654 they are all registered as such on the cadasters;
that was when the number of titled peasants was frozen. And this seems to have been in response to orders from the overlord.
Subsequently, in the 1670s, clients reappear on the kumi rosters under various labels—"kakae," "elder brother kakae,"
"younger brother kakae," "nephew kakae," and so on—as dependents of titled peasants and yet as owners of homesteads and
lands.[64]

The same thing happened in Gorobe-shinden (discussed in chapter 2), a few kilometers to the east of Kodaira, with the large
development of new fields by a dogo[*] . When all the available flat land was brought under cultivation by the late 1660s, the
number of titled peasants froze at forty-five households, and kakae were entered on the population registers (shumon[*]
aratamecho[*] and goningumicho[*] ) as dependents of

[62] Naito[*] , Honbyakusho[*]taisei , 95-98.

[63] For an English translation of this law, see Lu, Sources , 1:207.

[64] NAK-T 4:463-64.

― 170 ―

titled peasants, hence not as full community members. Half a century later, in 1712, the number of titled peasants in Gorobe-
shinden was still forty-five, but thirty-four of them had a total of eighty-three kakae.[65]

Naito[*] points out that if there was a limitation on the number of titled peasants set by overlord power, as was sometimes the
case, and if, by bakufu order, it had become difficult to establish branch houses (and if, one may add, certain sectional
intravillage interests were served by such restrictions), then new branch houses, even if they were economically independent
and as well off as their parent families, had officially to be registered as still part of, or dependent on, their parent families. In
other words, this limitation maintained hierarchical relations between main and branch families and the importance of
lineage structures.[66]

This arrangement looks very much like another tatemae reconciliation of the contradictions between legal principle (making
branching a virtual impossibility for the majority of peasants) and actual practice (it was not admitted to on paper, but it
happened nevertheless). A tatemae approach to law can perhaps best be rendered as a "nod and wink" approach—the Italians'
perception of their own relationship to law. The authorities had documents showing compliance with the law (the stable
number of titled peasants), and the peasants had a practice they could live with—an eminent illustration of what Pierre
Bourdieu calls a misrecognition of reality that fools nobody.[67]

Thus, the lineages used bakufu restrictions on partitioning holdings to structure themselves and were at the same time
structured by it. In and of itself, however, lineage economic power was not effective if it could not be translated into political
power, which meant acquiring for its members the title "titled peasant." Since the number of titled peasants was restricted,
however, they had to wait for the moment that a titled peasant, for one reason or another, had to give up his share. Another
bakufu institution leveraged by lineages was the kumi system. Kumi organizations were used to strengthen lineage filiations,
because

[65] Ibid., 464.

[66] Naito[*] , Honbyakusho[*]taisei , 50, 175-79.

[67] Bourdieu, Outline of a Theory of Practice (Cambridge: Cambridge University Press, 1977), 133, 171. The Japanese categories
tatemae (face, façade) and honne (real intention), which John O. Haley skillfully uses in his discussion of "law as tatemae "
(Authority without Power: Law and the Japanese Paradox [New York: Oxford University Press, 1991], 186-90), fit perfectly
Bourdieu's analytic notion of misrecognition.

― 171 ―

all heads of the five-household neighborhood groups were patron houses. Intravillage lineage hierarchies and administrative
structures, the latter ordered by overlord authority, thus overlapped.

Let us return now to Kodaira village. By 1731 forty landholding kakae and seventeen landless peasants, all incorporated into
hierarchical lineage relationships, were ruled by twenty titled peasants, half of them virtually landless (i.e., having less than
one koku). The situation was unavoidably wrought with tensions. Some tatemae, or cosmetic, as opposed to substantial,
changes had been made in the entries of the population registers a few years earlier, changes that reduced somewhat the
difference in status between tided and nontitled peasants. The former were no longer entered as having been tided peasants
daidai , "for generations and generations," and to the latter's name was no longer added onbyakusho[*] tsukamatsuri makari ari
, "deferentially in service of titled peasants."[68] But how could these nontitled peasants go about effecting substantial changes?
Some overlords (in Ueda domain at one point) had legislated against such changes; others (often in bakufu territories) were
not interested in such matters.
In 1769 a number of kakae in Kodaira organized themselves as spokesmen for their class, although they did not form a
majority, and raised an issue that was of some interest to the authorities, namely, the structure of the neighborhood kumi. That
year eighteen (out of fifty-eight) kakae filed a suit requesting that the kakae system be abolished. They argued that patrons
should not also function as kumi heads because this constituted a repressive, overlapping authority structure. On 2/29,
however, before the suit was even filed, sixteen of the twenty-one tided peasants of Kodaira submitted a written rebuttal to the
five village officials (the other titled peasants). They argued that there had been twenty-one tided peasants in the village "since
ancient times"; that they, together with the kumi heads and the representative of the peasants (hyakushodai[*] ), made all the
important decisions; and that issues of kumi membership were always discussed by all the peasants. The next day the
remaining forty kakae wrote to the village officials that they disapproved of the suit by their eighteen status mates.

This attempt at staving off the suit failed, and the officials filed a recorded testimony with the bakufu's intendant at Mikage-
shinden,

[68] Ozaki, "Kinsei sonraku," 767.

― 172 ―

stating that the overlapping system was according to longtime village law and that at a meeting of 2/8 "all" the peasants had
decided not to change the entries in the new population rosters that were due within a few weeks. Moreover, they reminded
the authorities that prior to this suit of the eighteen, three among them had gone directly to the office in Mikage-shinden about
this matter and that that office had shown little interest, ordering a private settlement (naisai ). The plaintiffs, however, had
ignored the decision that was subsequently reached in the village. Finally, a few days later, in the third month, the suit was
filed together with a rebuttal signed by the village officials and the titled peasants.

The plaintiffs obviously wanted to use the institution of the neighborhood kumi as leverage to break the power of their bosses,
while the defendants stated that the overlapping system was village law and did not cause problems to the plaintiffs. The
latter, however, argued that while Tokugawa law required that kumi be established by neighborhood and that a head then be
designated, in Kodaira only patrons were kumi heads. The plaintiffs made it clear that they believed the village officials had
been maneuvering to prevent kakae from becoming kumi heads. (The case of nearby Iribuse, discussed in chapter 2, was very
similar.) The defendants, conceding this point somewhat, proposed a compromise whereby the titled peasants would
constitute three separate kumi and the other peasants would be grouped in genuine neighborhood associations and would
choose their own heads. This compromise solution was rejected. The eighteen were intransigent and wanted to negotiate only
if the entry "patron" (kakaeoya) were taken off the population register. They cited the precedent of neighboring Makibuse,
where the overlapping system had been abolished. But the village officials were unimpressed, arguing that Kodaira had its
own village law and that it did not matter what was practiced elsewhere.

This suit describes concretely the trouble patrons caused their clients; some complaints are about the symbolic distancing of
the village officials from the peasants. In the third month the new population register had to be signed. The custom had been
for peasants to drop by the headman's home to certify their entries with their seals, but now they were all summoned to
assemble in the headman's yard, with the officials inside and the representative of the peasants acting as a go-between. The
officials gave three reasons for this hierarchical arrangement. First, since the village had become part of the bakufu domain
(in

― 173 ―

1765) the village law had to be read to the peasants once a year, which was why they were gathered together to hear the
reading and sign the register at the same time. Second, the headman's house was too small for a gathering of all the peasants.
And third, since the weather was fine, it seemed like a good idea to use the outdoors. The last two reasons sound spurious: it is
doubtful that the headman's house was too small, and in the mountains of Nagano in the third month the weather is still
bitterly cold. Obviously, the kakae interpreted this setup as arbitrarily emphasizing the differences in status between the
officials, or patrons, on the one hand, and the peasants, or kakae, in general, on the other.

The eighteen had set out for the bakufu office at Mikage-shinden (some sixteen kilometers away) with their suit once before,
but the village officials had gone after them and persuaded them to return by promising a private settlement. The negotiations
had faltered, however, over the abolition of the patron system, and the eighteen had reintroduced their suit. Then the issue of
a conflict of interests was raised, the plaintiffs expressing their concern that the decision by the titled peasants to have the
village shoulder the officials' travel expenses might have had something to do with the overlapping responsibilities of patrons
and kumi heads. The titled peasants simply denied the allegations and asserted that the decision about shouldering the
defendants' costs had been made not only by the twenty-one titled peasants but by all the peasants. Costs had also been a
factor in postponing the suit of the eighteen for over four years. They had submitted their suit to the peasant representative
several times, but the village officials had ordered prepayment of the costs of the suit. The titled peasants' view of this was that
all these earlier attempts had constituted only informal talks, not formal proposals for a full-fledged suit.

The plaintiffs, conscious of their economic independence, also requested that any household that could not be self-sufficient be
excluded from the kumi. The officials responded by arguing that economics had nothing to do with full kumi membership:
lack of economic self-sufficiency was no reason, they said, to lose one's status as titled peasant and patron. Only if one became
a servant in another village did one lose one's status. But this was not the case with Kyuzaemon[*] , whom the plaintiffs cited as
an example. Kyuzaemon[*] , patron and kumi head with a meager holding of 0.085 koku, was working as a servant (hokonin[*] )
in

― 174 ―

another household. He had in his kumi one successor house and one doshin[*] (a Buddhist priest without a temple). When one
of these two disappeared the field had to be taken over by another peasant, who had himself only 0.15 koku but never lifted a
finger to cultivate the minuscule plot.
The suit contains yet another reference to a previous suit, this one involving a kumi head (0.4 koku) who had expelled one of
his kakae "because not only had he always been insolent but he had also ignored advice from the village and he was selfish
(wagamama )." The village officials had then told the kakae that because he now had no patron he could no longer live in the
village. This kakae also had brought a suit to the Mikage-shinden office. The case had been stopped, however, through
mediators brought in from another village. The same expelled kakae, together with his three co-kakae, were among the
eighteen plaintiffs.

By the time of the suit, the titled and nontitled peasants were economically virtually undistinguishable. Only five titled
peasants had holdings of more than z koku (the sake brewer, with 120; the headman, with 14; and three others, with 5, 4, and 3
koku), four more had x koku, and the remaining twelve had holdings below 1 koku. On the other hand, nineteen kakae had
more than x koku. Both plaintiffs and their ten patrons were spread evenly on both sides of the 1-koku divide. Four patrons
found all their kakae united on the side of the plaintiffs, and eight had succeeded in rallying all theirs in opposition, while the
kakae of the remaining eight were divided between plaintiffs and opponents to the suit.

The outcome of the suit, mediated at the wealthy temple Fukuoji[*] , appeared to be largely a victory for the plaintiffs:
reference to kakae and mizunomi in the kumi membership lists was to be dropped, kumi would now be arranged
geographically, and kumi leaders were to be elected from within the kumi every other year. But it was a hollow victory,
because after four years, in 1773 and 1774 three separate but identical suits were submitted (by twenty plaintiffs, thirteen of
them new ones) about the problem of overlapping jurisdictions. This time, however, the plaintiffs' demands were on a smaller
scale: they wanted to be freed from their status as kakae and to be "recognized in the population roster as born in this village."
There was no critique of the kakae system as such, no insistence on abolishing it. This obviously was not a

― 175 ―

"class suit," for all kakae. Moreover, all petitions were also co-signed by the officials and the patrons in question. What was the
rationale behind this new drive, coopted, it seems, by all those with any power? Who among the kakae signed, and why not all
of them now that the drive enjoyed official sanction? (There is secondary evidence of an additional petition, now lost, by
another ten kakae, bringing the total kakae supporters to thirty, still just above half of the total kakae population of fifty-six.)

The petitioners made a distinct, on between kakae that were related by kinship and those that were not. What four years
earlier had appeared to be a united front as far as claims were concerned had broken down along kinship lines. Fictional
kinship was fine, the petitioners seemed to say, but it was not the real thing and should not be treated as such. Distinctions
were introduced, or made stronger, in order to separate the two and to produce distinction that was acceptable to the
authorities.

It annoyed the thirty kin branch houses that they would have the same status as the twenty-six non-kin branch houses, and
they wanted not simply to have their kakae status removed but to be acknowledged as hyakusho[*] , "peasants." (A similar
distinction between kin and non-kin kakae had been made in Shimo-Kaize.) The result, however, was not equality with the old
titled peasants, here called onbyakusho[*] , or "honorable peasants," but a further splitting of this status whereby the kin kakae
became second-class onbyakusho[*] , while the non-kin kakae became simply hyakusho[*] , peasants—but the term no longer
simply described an occupation: it now was a title designating their status within a finely graded hierarchy.

The population register of 1780 includes the following seven ranks of peasants (the number of peasant households in each
rank is given in parentheses):

1. "onbyakusho[*] [honorable, i.e., titled peasant] living in this village since ancient times" (18)

2. "onbyakusho[*] born in this village" (29)

3. "hyakusho[*] born in this village" (18)

4. "hoko[*] [in service] at X with Y (born as onbyakusho[*] in this village)" (5)

5. "X's kakae, born in this village" (3)

― 176 ―

6. "X's mizunomi ['landless' (tenant)]" (3)

7. "mizunomi born in this village (onbyakusho[*] since ancient times but now mizunomi)" (1)

This new status stratification was an ingenious device to accommodate some of the demands while preserving distinction and
the distinctions that constitute the symbolic capital—they had hardly any other—of the dominant fraction of the village. The
demands of the kakae, both kin and non-kin, were met: in ranks z and 3 they were now recorded as "born in the village" and
no longer as kakae but as peasants, hyakusho[*] . The special demand of the kin kakae was also met, for they were now
distinguished from the non-kin by the "honorable" prefix on . And former titled peasants (onbyakusho[*] ) were recognized as
such: in rank 4, for those who had to give up their title because they went into service outside the village while retaining some
land, and also in the lowest rank, because they had lost all their land (below the mizunomi of rank 6, who had some land).
There are, however, a few kakae left, in rank 5.

Later population rosters tell the continuing story of wranglings over status promotion and maneuvers against status pollution.
All non-kin hyakusho[*] of rank 3 became absorbed as onbyakusho[*] into rank 2, and all dependent mizunomi of rank 6 were
promoted to dependent kakae of rank 5, who later were gradually incorporated into rank 2. On the 1849 roster only the names
of peasants are listed, without any prefixes. Had equality been achieved? Not quite, because there are "sidescript" entries next
to the names. On this roster, twenty-one peasants are marked as osabyakusho[*] , "head peasants" (three of them mizunomi),
forty-eight as hyakusho[*] , "peasants" (among whom are three mizunomi), and two as kakae (one mizunomi). Finally, after
some two hundred and fifty years all but two in this peasant village had become "peasants." Had status been superseded by
class at the end of the Tokugawa period, as the historians mentioned at the beginning of this chapter argue? Not quite.
Mizunomi is a clear economic marker for landless peasants. Yet three mizunomi are among the elite twenty-one.

Similar struggles concerning the status of titled and nontitled peasants took place in the area around Kodaira at roughly the
same time. We shall look briefly at two additional cases that involved "mass" decisions, one involving promotion to, the other
demotion from, the status of titled peasant. A group promotion of kakae to titled peasants

― 177 ―

should not be interpreted too readily as a final step toward equality. Often, subtle or even explicit social distinctions were
maintained between the old elite and arrives.

In 1780 in Nagatoro (in today's Saku city, some fifteen kilometers east of Kodaira) thirty-five clients were set free by twenty-
five patrons, and they became full members of the neighborhood kumi (resulting in the creation of seven new kumi); but all of
them had to be given permission by their former patrons, who thus became their guarantors.[69] The reasons these guarantors
gave for setting their clients free included the following: "he has his own holding"; "he has a separate house but no land yet,
but he will cause no trouble"; "I shall give him land, and he will cause no trouble"; "he is my relative, and I shall give him 1.44
koku."

Having a clear title to some land (even one koku or less) was a basic condition for becoming a titled peasant in Nagatoro. It is
ironic, however, that these amounts of land were insufficient to maintain even a small peasant household. These peasants,
now nominally "full peasants," must have supplemented their income through nonagricultural pursuits. Economically, the
term "titled peasant" was an empty title; its significance was purely political and social.

The "independence" thus gained was conditional upon good behavior; the guarantorship certainly suggests as much.
Sometimes the difference between old and new titled peasants was codified and expressed in public signs. In Tanoguchi
village (also in Kita-Saku district) the village laws of 1681 stipulated that clients promoted to the status of titled peasant were
not allowed to wear haori coats, hakama ceremonial aprons, or ashida (high rain geta, or wooden clogs) or to use umbrellas,
and the pillars of their houses could not exceed eight shaku (2.40 meter); nor could they use tatami mats, and they were not
allowed to enter homes of "real" titled peasants, but had always to stay in the entrance way. (As we shall see in chapter 5,
prohibitions identical to these were issued against outcastes.) Such distinctions within the ranks of titled peasants could be
maintained for a long time. In one case, someone whose household had been promoted to titled peasant in the 1780s was
reminded in writing two generations later, in 1844, that he could not wear a kamishimo ceremonial over-

[69] Ichikawa Takeji, "Kakaebyakusho[*] ," 12-13.

― 178 ―

coat except when he was the main mourner at a funeral or the groom at a marriage (an allowance made because of relaxing
standards, it was stated). If these things were disregarded, he was warned, his house would return to its former kakae status.
[70]

The group demotion case, a rare occurrence, is from Kasuga-shinden (a few kilometers south of Kodaira), which, with only 145
koku, was a rather small village.[71] In 1767 the headman in one stroke demoted twenty-five titled peasants to kakae status.
The result was a suit (in the third month, as was so often the case) brought by the demoted to the headman. The headman,
however, single-handedly, and without any further endorsement by village officials, passed the suit on to the bakufu
authorities in Mikage-shinden—in an effort to doom it.

It should be noted in passing that this circumstance throws some light on the process of lodging suits. Dan Fenno Henderson
has argued that villagers' suits had to be approved by the village headman and could be initiated only after conciliation efforts
had failed.[72] Japanese scholars have disputed this point. According to Ishii Ryosuke[*] , no conciliation efforts were needed;
the headman had simply to provide an accompanying letter, which the headman failed to do in this case.

All four kumi heads (the only ones among the twenty-two patrons who had not been consulted, "because this would have led
to a dispute") immediately wrote to the same authorities, explaining the case and requesting an investigation. The headman
had put the twenty-five former titled peasants as kakae under their relatives (brothers or parents), who thus became their
patrons. His reason for this drastic step was that there were too many tiny peasants in the village with holdings of "0.02, 0.03,
0.05, 0.08, 0.2, 0.3, 1 and z koku, and too many irregularities, making it difficult to apportion expenses to each and every
peasant; it was much easier to rechannel this through patrons." The intendant investigated the case and returned the original
petition to the village, noting that he would attend to the matter if the request was forwarded properly, via all the village
officials. In the second month of the following year the kakae were reinstated as titled peasants.

[70] Ibid., 8-9.

[71] For a summary statement of the case and some of the documents, see ibid., 10-12. All full documents for this case can be
found in NAK-KS2 (1): 964-68 (nos. 490-93); for the village kokudaka, see ibid., 724.

[72] See Henderson, Conciliation 1:136; and Ishii Ryosuke's[*] critique in his long review of this work in Law in Japan 2 (1968):
216-17.

― 179 ―

Lineages, Goningumi, and Village Politics: Oashi,[*] , 1840S

Lineages were left untouched by the kumi organizations established by order of the overlords. As the suit by the Kodaira
kakae makes clear, lineages are not unalterable biological networks. The plaintiffs wanted to do away with the social construct
of patron and client, with the inscription of dependency onto relations between main and branch houses. As Naito[*]
suggested, the maintenance of hierarchical relationships between households along lineage lines may have been prompted by
the kabu structure of the title "titled peasant."

In the following case study, we shall take a close look at how lineage politics were structured by the institutions of titled
peasants and kumi membership. The location is Oashi[*] village (presently part of the town of Akashina in central Nagano
prefecture). A number of kumi rosters from that village dating between 1731 and 1872 examined by Ozaki Yukiya give one a
good sense of the composition of the kumi.[73] What is striking, but certainly not exceptional, is that in every case village
offices and kumi headships were passed to members of the same families.[74] That this was a privilege is clear from the fact
that when the number of kumi had to be reduced by one, and hence one family would lose its leadership position, it was
decided that the two former kumi leaders would hold the position in alternate years.[75] This tightly knit group had turned this
privilege into a monopoly by devising ways to prevent any member from losing his status and hence to prevent others from
joining.

[73] Ozaki Yukiya, "Chikuma-gun bakufuryo[*] ni okeru goningumicho[*] m goningumi no jittai: Chikuma-gun Oashi-mura[*] no
baai," Shinano , 16, no. 2 (1964): 45-60.

[74] Ozaki has documented this in detail for the period between the 1750s and the 1820s (ibid., table 7 [559]). Dan Henderson
presents an agreement from 1867 between the six titled peasants of one village as evidence for a collusion m rotate the
headmanship every five years among themselves and thus keep control of the land cadasters (see his Village "Contracts ," 164-
66): "If outsiders raise objections, the six as a group would come forward strongly, and try to settle the matter .... If besides our
six some outsiders become the headman or group chief [kumi head], some one of the six of us will carefully keep custody of
the survey records, and we will make sure to agree on all issues. Hereafter, besides agreeing within the six-house group, we
would definitely work together in harmony no matter what the problem is and help each other to the greatest extent" (164).

[75] Ozaki, "Chikuma-gun," 54.

― 180 ―

Scarcity and access had to be controlled in order to maintain the value of this political and social capital.

In 1848, however, five households wanted to split off from a nine-household kumi to form a separate kumi with their own
elected head because "the kumi was too big." Yet a nine-household kumi was not unusually large: the largest kumi counted
eleven households, and the smallest, three. The officials consulted the higher authorities, who gave permission for a new
kumi; however, "in consultation with the village officials" it was decided that the current head of the kumi would move with
the five petitioners as their new head, a strategy to maintain a monopoly on political power against the threat of a newcomer.
The five did not agree to this condition and so refused to sign the population register. They also rejected another peasant
whom the officials tried to impose upon them as head in a failed attempt at compromise. The result was a suit and a new
compromise: the village officials would be in charge of the new kumi for three years, and then they would decide on another
head.

Taroshichi[*] , the leader of the secession quintet, had been a troublemaker in the village before this incident. Nine years
earlier, in 1839, he and his son had been involved in a suit against two other peasants, also a father and a son, about collecting
firewood. The Buddhist priest and officials from a neighboring village, however, had worked out a settlement. In 1844 another
member of the secession group, Risaemon, had been involved in a quarrel at a wedding in his home. The quarrel arose after
Risaemon announced that he was going to change the name of his younger brother to Heijiro[*] , which he claimed was the
name of one of his forbears, and one of the guests made it clear that he found this ludicrous. The quarrel had simmered down,
but soon thereafter Risaemon and Taroshichi[*] had sued the guest for slandering them and causing not only trouble at the
wedding but also hardship by criticizing their family pedigree. They had proof, they said, that about two hundred years
before, in the 1650s, a Heijiro[*] had been recorded in the cadasters as a titled peasant. Thus, for at least ten years before the
secession incident Risaemon and Taroshichi[*] seemed to have been buddies prone to causing trouble.

The issue of the new kumi headship flared up again when, after the three-year interim, the five requested the promised
appointment of a kumi head and the village officials postponed it for another seven years!

― 181 ―

To the new suit the five filed, the officials answered that the five in question had not changed their behavior, so none of them
could be chosen as head because "in our village, heads have to come from households with a long pedigree (senrei nagatachi ni
kagiri , limited to [houses] with longstanding precedent)." The five argued that headships should have nothing to do with
pedigree (iegara ) or status (mibun ), that it was unreasonable to bring these things to bear on matters of village governance.
Risaemon had used the public occasion of the wedding to stake out his claim to a long pedigree. Given the political importance
of pedigree, undoubtedly his declaration was a political move and was perceived as such by the guest who ridiculed it.

Eight years later, in 1859, a new compromise was reached. The officials' requirement for headships was now hitogara , or
character, rather than pedigree. But, the officials added, since none of the five in question had the necessary character, they
had to wait until someone with the proper character emerged in the group who could be appointed as head. Therefore, each
year one of the five would function, not as hangashira (the local term for kumi head), but as hangashira-dai , a proxy head.

This is another example of an ingenious solution by "splitting." It provided face-saving compromises that allowed both parties
to claim victory. The village officials defended their turf: they stuck to principle by not capitulating to the demands of the five.
The latter, in turn, gained self-governance, which in practice put them on an equal footing with the other kumi heads. The real
solution would come about with time, which was on the side of the secessionists. It was inevitable that they would eventually
get what they wanted, perhaps in the next generation, when these fine distinctions had become, for everyone, a dead letter.
And this was indeed what happened. What was the nature of this "inevitability"? Why could five peasants hold out for so long?

This protracted conflict, of which we know only the peaks, points to social fault lines in the village that are invisible in the
texts of the suits. It turns out that Taroshichi[*] and Risaemon and the other three members of their kumi were related (see fig.
5). The fault lines demarcate divisions of lineage power. Over time, the five had come to constitute an economically powerful
lineage. The conflict over these several decades was one between old and new lineages, the former trying to block

― 182 ―

Fig. 5.
Lineage of Risaemon's Secessionist Kumi, Oashi[*] , 1725-1869

Source: Adapted from Ozaki, "Chikuma-gun," table 8 (p. 58).

Note: Names are of secessionists. The numbers are of kokudaka.

the translation of the latter's economic power into political clout by appealing to family pedigree, social standing, and personal
character.[76]

Pedigree was at issue in the quarrel over names (when Risaemon and Taroshichi[*] claimed a two-hundred-year-old ancestry)
and in the disagreement about kumi leadership. Since both disputes developed into suits, it is not surprising that the village
appears to have conducted a pedigree check. On the tribute list of 1725, under the ancestor Matsuemon (the holder of 1.1 koku
in fig. 5), the officials noted that the

[76] Writing these lines in October I992, I was struck by how the suppression of the economic in the name of "character"
during the debates preceding the U.S. presidential election was similar to the strategy employed by the village officials
discussed here.

― 183 ―

older records had been lost in a fire. Needless to say, there is no guarantee that this was an impartial investigation. Since the
name Heijiro[*] , which Risaemon intended to give his younger brother (Hei'emon), does not appear in subsequent village
records, it seems likely that his claim of titled ancestry going back to 1650 was not recognized.

In the 1700s, during the first two generations of the lineage as we know it, the household's holdings remained at the low
kokudaka level of 1.1. Around 1800, however, the holdings tripled, which allowed the household to create a branch house in
1815. This trend continued during the following decades, leading in 1848 to the attempt to consolidate the economic power of
the lineage (which now totaled 15.4 koku) by giving it a political dimension. The lineage continued to grow, and by 1869 it
counted nine households and a total of 26 koku. Ultimately, in 1872, the lineage split into two separate goningumi, headed by
Taroshichi's[*] two sons.

In order to acquire political power commensurate with their economic strength, lineages had to do more than simply create
ever more branch houses. Either they could consolidate their power within the village (and thereby strengthening intralineage
authority) by setting up a separate goningumi or they could acquire for some of their branch houses titled status when such a
share came on the market through the bankruptcy, abscondence, or venality of some shareholder. Whichever route they
decided to take, they needed foremost the approval of the local decision makers, and various factors determined the ease or
difficulty with which that was given. The careers of a few titled peasants from Gorobe-shinden illustrate the process.[77]

Gorobe-shinden, as the name indicates, was a village of newly developed fields. In the first two generations, since there was
little competition for land, social mobility was not very restricted. In 1695 three villagers from Gorobe-shinden wanted a
relative of theirs to move into the village as a tenant. They promised to take responsibility for him and establish him as a
kakae; eighteen years later this newcomer became a titled peasant with 1.6 koku. One couple came as servants (hokonin[*] ) in
the 1640s, bought some land, rented a house, became kakae, borrowed money to buy a house, then bought land in other
villages. In 1713 the household gained titled peasant status with 12 koku, two bond servants, and one horse. Another peasant
also started his career as a

[77] Oishi[*] , Kinsei sonraku , 125-27.

― 184 ―

servant, in 1673. In 1687 he was able to buy land, build a house, and take the place as titled peasant of one who had died.
Later, however, he sold the house and title, became a kakae again, and finally ended up landless (mizunomi).

Where land was scarce—a general problem toward the end of the seventeenth century—social mobility by means of upper-
status "shares" became restricted. Shares became available when a shareholder died without offspring or successor, left the
village, went bankrupt, or decided to sell his share. Symbolic capital, originally an expression of real capital, could be
converted back to it: in 1760 a share in Kasuga village (a few kilometers east of Gorobe-shinden) went for the outrageous price
of twenty ryo[*] ;[78] two years later in Gorobe-shinden itself someone paid only one ryo[*] .[79]

The market for the title of titled peasant was restricted in three ways. First, it was a village market, limited to the actual
residents. A peasant could be titled only in his own village (even if he had additional fields in other villages) because the status
entailed a position of village leadership. Second, the price for the title was set within the village, not on an intervillage "market
for titles." Third, cash alone was insufficient. Since the title was essentially a social one, its acquisition had to be sanctioned
socially. Beyond cash, symbolic capital consisting of social esteem and communal "harmony" was needed. There were specified
procedures for obtaining a title, as an incident of 1812, also from Gorobe-shinden, illustrates.[80]

That year someone's client bought the title of titled peasant from a third peasant. At the time of the yearly registration of the
population, he asked his former patron to erase his client status. The patron, however, insisted on a formal written request,
which met the objection that "things were never done that way before." "Au contraire, " the patron returned, "you have to go
through the village council." Inevitably, the intendant was presented with a suit.

[78] Ichikawa Takeji, "Kakaebyakusho[*] ," 10. Kobayashi Daiji lists prices for symbols of elevated social status in the second
half of the Tokugawa period (without giving place or date or citing sources): 5 ryo[*] for a kamishimo ceremonial outfit, 10 for
village headman's rank, 15 to use a surname, 20 for the rank of village group headman, and so on, all the way up to 100 ryo[*]
(see his Sabetsu kaimyo[*]no rekishi [Yuzankaku[*] , 1987], 193).

[79] Oishi[*] , Kinsei sonraku , 132.

[80] Ibid., 134-35.

― 185 ―

The officials, fearing trouble if the suit came to a judgment, decided on the following course. The plaintiff, they said, was a
diligent peasant and deserved to be a titled peasant but should have notified his patron before buying the title—or certainly
immediately after buying it—instead of waiting until the next population registration: he had to admit to wrongdoing. The
patron, on the other hand, should be honored to have cultivated such a qualified peasant, with a sizable holding, who had
contributed diligently to a government request for funds: such a peasant deserved to be set free.

Both were reprimanded, and both were satisfied. This is an example of a typical solution, with a balanced interplay of
apologies, shared blame and praise, and thus no clearcut winner or loser. The aftermath of this solution is also instructive.
Soon thereafter, three more clients of the same patron asked to become clients of the new titled peasant. The trio circulated
secret memos bad-mouthing their patron, but this tactic led nowhere: their status remained unchanged.

Cooperative Villages

The villages discussed in this chapter were mainly dogo[*] villages with tightly structured lineages. In contrast to the corporate
so[*] villages, predominant in the Kinai, they were characterized by strong hierarchical arrangements, their origins in large
extended families. Political equality always lay just beyond the horizon, always one more status distinction away. Both so[*]
and dogo[*] villages had their roots in pre-Tokugawa formations.

A third type, one that has not received much attention even in Japanese literature, goes by the convoluted name
shozokudanteki kyogyotai[*] , literally, "small, family-type occupational cooperative village.[81] Harada Seiji compared the
structures of twelve villages of Fukuyama domain (Bingo province, Hiroshima prefecture) as they appear in a document from
1647 and found what he calls patriarchal-type villages and cooperatives.[82] Traditional patriarchal villages were to be found
in

[81] An excellent bibliography of the few articles and books on this type of village can be found in Harada Seiji, "Kinsei zenki
sonraku no shoruikei: Fukuyama hanryo[*] o chushin[*] ni shite," Shigaku kenkyu[*] , no. 193 (1991): 26 n. 3.

[82] Harada rejects the term dogo[*] because most of the large landholdings had disappeared in the warfare at the end of the
sixteenth century, but his "patriarchal" villages look very much like villages other scholars refer to as being controlled by
dogo[*] .

― 186 ―
the hills and mountains and typically engaged almost exclusively in wet rice cultivation. Cooperative villages dominated by
small peasants were mostly in the coastal areas, with only dry fields, or on inland flat terrain, with a balanced economy of wet
and dry fields. The latter two regions are characterized by open terrain, making expansion possible and reducing competition
for land, which is in great contrast to conditions in the hills and mountains.

Harada's hypothesis linking in a general way ecological conditions with certain types of village organization is suggestive but
certainly far from conclusive, given his small sample. Nevertheless, his structural comparisons put in sharp relief different
village structures and can serve to conclude this part of our study. Harada divided the peasantry into the following categories:
(1) titled peasants without bond servants (genin): small, independent cultivators with holdings usually between 5 and 15 koku;
(2) tided peasants with genin, broken down further into peasants with 15 or more koku, that is, typical patriarchal families,
and peasants with less than 15 koku; (3) genin; and (4) mawaki , or kin branch houses.

Table 15 contrasts two such villages. In Futamori, with a kokudaka of 165 koku, the titled peasants were clearly divided
between those with genin (more than 20 koku) and those without; all genin belonged to the wealthier peasants. Harada
argues, however, that this top-heavy hierarchy (approximately 72 percent of the arable was owned by the five peasants with
holdings of 15 or more koku; 22 percent by those with 5 to 15 koku; the remaining 6 percent by peasants with below 5 koku) is
not a leftover from medieval times: in 160 t the respective percentages were approximately 26, 55, and 19. Pre-Tokugawa
villages would have been characterized by small holdings like those of Mizunomi village (466 koku). Typically, here only one
holder had more than 20 koku, while most titled peasants, both those without and those with genin, had 3 to 10 koku, as did a
large number of the genin. Of these twenty-three genin only eight were attached to the top three titled peasants; all the others
were genin of small peasants with holdings hardly larger than theirs. The smallest tided peasant (with between 1 and 3 koku)
had four genin, as did the titled peasant just above him.

Many small holders had genin who were taka holders; in some cases the combined kokudaka of the genin was almost twice
that of their host family (see tables 16 and 17). These "bond servants" obviously were

― 187 ―

Table 15.
Patriarchal and Cooperative Villages in Fukuyama
Domain, 1647

Titled Peasant

Total Without With

Koku le Genin Genin Genin Mawaki

Futamori Village (Patriarchal Type )

20-30 2 2

15-20 2 1

10-15 1 1

7-10 2 2

5-7 1 1

3-5 2 2

1-3 1 1

0-1 10 3

Total 22 2 3 9

Mizunomi Village (Cooperative Type )

30-40

20-30 1 1

15-20 2 2

10-15

7-10 6 3 3

5-7 13 10 2 1

3-5 38 27 1 1 6

1-3 57 15 1 9 13

0-1 65 13

Total 182 55 10 23 20

SOURCE : Adapted from Harada, "Shoruikei," foldout

table 9 (between pp. 14 and 15).

NOTE : Total number of ie (households) includes 7

muyashiki (those without a residence) for Futamori
and 66 for Mizunomi.

― 188 ―

Table 16.
Titled Peasants and Genin,
Uchitsuneishi Village, Fukuyama
Domain, 1647

Koku Titled Genin

Peasants

10-15 2 7

7-10 2 4

5-7 2 3

3-5 4 5

1-3 2 3

SOURCE : Harada, "Shoruikei," table

11 (p. 16).
Table 17
Holdings of Host House and Genin,
Kurome Village, Fukuyama Domain,
1647

Koku

Host House

Kichi'emon 11.964

Genin

Sakejuro[*] 6.028

Jinshiro[*] 5.823

Fujisaemon 3.045

Jinshichi 2.312

Soshiro[*] 1.728

Yoshichiro[*] 1.253

Total 20.189

SOURCE : Harada, "Shoruikei," table

13 (p. 17).

situated quite differently both socially and economically from those of Futamori, and they look more like the clients (kakae)
we have encountered elsewhere. Mizunomi was a typical "cooperative" village with large numbers of almost equal holdings
among mawaki as well, the result of the division of property through equal inheritance. This was a village with high status
mobility. In a relatively short time, these mawaki became fully titled peasants, and the genin seem to have had no difficulty
following the same route. Part of the reason may have been

― 189 ―

the availability of additional space: between 1619 and 1647 Futamori added only 6 koku to its arable, while Mizunomi added
55.

The reason these families organized themselves as a cooperative on a more or less equal basis was, according to Harada, to
meet the overlord's extraordinarily high extraction quota, which set a premium on rice, a crop that was not produced here, or
if it was, the yield was far less than in the hills and mountains. There is certainly room for debate and further research on an
argument that so categorically relates geographic location and type of agriculture with social organization and, further, links
the cooperative villages, through their type of agriculture, with extraction policies. What is clear, however, is that Japanese
rural society offered a number of organizational patterns and that status rigidity seems to have been linked with the
availability of space, a phenomenon we encountered in the case of Gorobe-shinden.

To conclude, then, status distinctions and status-generated distinction were part of village life throughout the Tokugawa
period. As the frequent appearance of virtually landless titled peasants in the records shows, status was not replaced as a
differentiating mechanism by economic class as such. The sources of status distinctions were intravillage economic factors and
extravillage political realities, which reinforced the former: the stature of traditional, established peasants, undoubtedly the
decision makers in the villages, was enhanced by their being assigned the communal responsibility for tribute and services
owed to the overlords. This official duty somehow made them the only ones who were real peasants.

Some overlord measures contributed either directly or indirectly, but not uniformly and universally, to limiting the number of
titled peasants. After the 1673 prohibition against partitioning holdings below a certain size, the villages seem to have resorted
to fixing the number of titled peasants while in reality putting very few, if any, constraints on actual practice. This necessarily
resulted in recording economically independent peasants as dependents, which solidified lineage formation, strengthened
status divisions, and limited political power to small segments of the village population.

Within these institutional constraints, resistance to status mobility depended on lineage expansion and the availability of
additional arable. In extended families, the hierarchical relations among members were unambiguously inscribed in their
economic and social positions, for it was clear who controlled the means of production and reproduction.

― 190 ―

Although this material reality was misrepresented symbolically through familial and paternalistic terms of kinship address
and on ritual occasions,[83] the symbolic had less autonomous determining power than it would have later. The head of the
family had a clear title to the land, owned the buildings and agricultural tools, and had the final word concerning which
members could set up their own families. With partitioning, control over these means was transferred to branch houses. The
preeminence of the main family came then to rely increasingly on symbolic means, status distinctions being one of the most
important. The availability of additional arable worked against the maintenance of rigid status hierarchies and promoted
status mobility.

When Japan reached an ecological barrier in the early eighteenth century, status again came to play a greater role in village
life (contrary to the opinions of many American scholars). It seems that status production, the generation of ever finer status
distinctions, was a powerful tool for offsetting the politically negative implications of economic decline. Economic
development affected eighteenth-century villages in two ways. Some peasants succeeded in accumulating more land, or
economic capital, which they sought to translate into political capital, but they encountered status barriers. More often,
however, a great economic leveling downward took place for most of the peasants, turning them into very small holders (and
an "objective" class), which is also contrary to prevailing interpretations. Either way, economic development brought pressure
to rearrange the structure of the political field in the villages.

This pressure could take a number of forms. Peasants could simply become nonplayers in the status game by removing
themselves from it and migrating to the urban centers. They could become part-time players by seeking employment
(hokonin[*] ) elsewhere for a period of time, as many did starting in the late seventeenth century, although this negatively
affected their position, standing, and status in the village.

[83] Smith astutely observes with regard m "ceremonial services" that "the participants themselves made no consistent
distinction between ceremonial and labor services, often calling both by the same name. It would seem—one cannot speak
with certainty in such cases—that the two services were thought of as being essentially the same. This is not because people
failed m observe that one had economic value and the other not, but because this difference though evident was insignificant
in view of the fact that the two had the same social character" (Agrarian Origins , 30). On kinship terminology, see ibid., 15 ff.

― 191 ―

They could also try to change the rules of the game, yet with the notable exception (in the present study) of the first of the two
drives in Kodaira to abolish the system of kakae altogether, this did not occur. Most of the demands were efforts, not to change
the rules of the game, but to change one's position within the field where the game was played, which ensured its longevity.

― 192 ―

4
Village Autonomy
In effect, the village had the security of the administrative state along with the freedom of the outlaw.
John Owen Haley

Village autonomy has two major components: financial and juridical. Japanese historians have given little consideration to the
former, analyzed in chapter 2, in their debate about whether or not (more often than the degree to which) Tokugawa villages
were autonomous. Their discussion has centered on the juridical aspect of village powers, judicial and prosecutorial, as found
in village regulations.[1] Between roughly 1900 and 1950 a number of village laws were made available in about half a dozen
compilations, mainly the work of legal scholars. Then, starting in the 1960s some historians began using these documents to
examine the relationship between village and state, questioning the legal scholars' assumption of village autonomy.

Hozumi Nobushige, one of the drafters of the Civil Code, and Miura Kaneyuki pioneered the study of the goningumi system
around 1900, no doubt inspired by the same concern that made the late Meiji (and Taisho) government promote "local self-
government." This research produced several volumes of goningumi regulations, edited by Hozumi Nobushige and Shigeto[*] ,
father and son, in 1921 and 1944, and Nomura Kanetaro[*] , also in 1944. To these works should be added Maeda Masaharu's
studies, especially his 1952 analysis and compilation of village codes , which were different from the goningumi laws because,

Epigraph: Haley, Authority without Power , 61.

[1] For an overview of this historiography, see Kadomae Hiroyuki, "Kinsei no mura to kinseishi kenkyu[*] ," Rekishi koron[*] ,
no. 95 (1983): 94-101; Uesugi Mitsuhiko, "Kinsei sonpo[*] no seikaku ni tsuite," Minshushi[*]kenkyu[*] , no. 7 (1969): 80-106, esp.
80-81; and idem, "Kinsei sonrakuron." See also NRT 3:401-5.
 ― 193 ―

according to Maeda, they developed mostly independently from the laws ordered by the extravillage authorities.[2]

This first phase of research has made available some six hundred village laws and two hundred village codes. In the
multivolume histories published by every prefecture for the last twenty years and in the innumerable local histories promoted
by the even more recent furusato bumu[*] , or boom of hometown-ism, hundreds of additional village regulations, as well as
statistical analyses of them by region, can be found.[3]

The first generation of scholars, with the exception of Maeda, were legal scholars trained in the German formalist tradition.
For them the mere existence of these village laws constituted proof of a tradition of self-government. But while such an
interpretation was of service to the Japanese state's attempt to revitalize local self-government in the interwar decades, it
conveniently overlooked the fact that goningumi laws had undeniably been an important instrument of central control over
local life during the Tokugawa period.

Maeda, on the other hand, subscribed to the autonomy thesis in perhaps less radical terms but for opposite reasons. Village
codes or rules (okite ), he argued, locating their origin and archetype in the famous codes of Nakano, Sugaura, Imabori, and
other corporate vil-

[2] Scholarly attention was drawn to village laws as soon as historical interest in the Tokugawa village as such was developed
with Fukuda Tokuzo's[*]Die gesellschaftliche und wirtschaftliche Entwickelung in Japan , Munchener volks-wirtschaftliche
Studien (Stuttgart: J.G. Cotta, 1900), an economic history of Japan published in German before it was published in Japanese.
These works on village laws are: Hozumi Nobushige, Goningumi seido (Yuhikaku[*] , 1902); idem, Goningumi seidoron
(Yuhikaku[*] , 1921); idem, comp., Goningumi hokishu[*] (Yuhikaku[*] , 1921); Hozumi Shigeto[*] , comp., Goningumi
hokishu[*]zokuhen , 2 vols. (Yuhikaku[*] , 1944); Nomura Kentaro[*] , Goningumicho[*]no kenkyu[*] (Yuhikaku[*] , 1944). The
compilation by Nobushige contains the documents he used for his first two books. Shigeto[*] , Nobushige's son, compiled two
more volumes, to which he added indexes for all 477 laws contained in both father's and son's compilations. Nomura's study
contains another 121 laws. Maeda Masaharu studied 216 village codes m his Nihon kinsei sonpo[*]no kenkyu[*] (Yuhikaku[*] ,
1952) and "Ho[*] to sonraku kyodotai[*] : Edo jidai ni okeru sonpo[*] o chushin[*] ni," in Hoken[*]shakai to kyodotai[*] , ed.
Shimizu Morimitsu and Aida Yuji[*] (Sobunsha[*] , 1959), 169-214.

[3] Kanzaki Naomi, for instance, recently analyzed 143 village laws of the old Musashi province (today's Gunma prefecture) in
"Musashino-kuni no sonpo[*] ," Tama no ayumi , no. 65 (1991). Examples of similar studies are idem, "Kozuke[*] kuni no
sonpo[*] ," Gunma bunka , no. 225 (1991): 39-55; Yokota, "Kinsei sonraku"; and Uesugi, "Kinsei sonpo[*] ."

― 194 ―

lages from Omi[*] province, some of which dated back to the mid fifteenth century,[4] constituted an independent tradition.
The overlord-generated goningumi regulations (the goningumicho[*] zensho, "preamble to the goningumi roster"), called here
village laws , adopted many items from these customary codes (especially from the end of the seventeenth century), which in
turn took on many formal and substantial features of these laws over time.[5] Maeda's rather tenuously argued thesis of village
autonomy and its postwar critique by scholars such as Kodama Kota[*] ,[6] mostly in the form of polemical counterassertions,
made clear the need to examine closely how these two kinds of village regulations were related to each other and to
extravillage authority. Two scholars addressed these questions in the 1960s: Uesugi Mitsuhiko, who compared laws and codes
for content and form, and Oide[*] Yukiko, who focused on the juridical status of the codes.[7]

Village Laws and Village Codes

Using fifteen categories, Uesugi compared the content of all of Maeda's village codes with a good number of Hozumi's village
laws and concluded that, given an overwhelming similarity, the village codes functioned to supplement overlord law as
expressed in the village laws of the goningumi rosters.[8] He pointed out that no village code contradicted any overlord
stipulation (to which we shall return), but he also warned, following Maeda, that neither these codes nor the laws covered all
aspects of village life; the more extensive laws reflected village life

[4] Ten of the twelve pre-Tokugawa village codes Maeda looked at are from Omi[*] province (Nihon kinsei sonpo[*] , "Sonposhu-
hen[*] ," 1-15); for a translation of some of the Imabori codes, see Tonomura, Community and Commerce , 198-207 passim.

[5] Maeda, Nihon kinsei sonpo[*] , 7-9, 18.

[6] For some critiques, see Uesugi, "Kinsei sonrakuron," 176; and idem, "Kinsei sonpo[*] ," 81 n. 6.

[7] Uesugi, "Kinsei sonrakuron"; Nanba Nobuo, "Hyakusho[*] ikki no hoishiki[*] ," in Seikatsu, bunha, shiso[*] , ed. Aoki Michio et
al., Ikki, 4 (Tokyo Daigaku, 1981), 45-88; Oide[*] Yukiko, "Kinsei sonpo[*] to ryoshuken[*] ," (Nagoya Daigaku) Hosei[*]ronshu[*] ,
nos. 18 (1961): 1-31 and 19 (1962): 73-128; Mizumoto, "Kogi[*] no saiban," 283-316; Mizubayashi Takeshi, "Kinsei no ho[*] to
saiban," in Chusei[*]no ho[*]to kenryoku , Chuseishi[*] koza[*] , 4 (Gakuseisha, 1985), 144-71.

[8] Uesugi, "Kinsei sonpo[*] ," 86-87. The village codes were more detailed in certain areas, for example, rules to implement
attendance at village council meetings, tribute collection, sharing of commons (use of mountains and water), regulations of
festivals, and penal matters.

― 195 ―
only as it related to overlord interests.[9] After the 1720s the codes' form and language conform closely to the overlord laws,
and a good number of them stress the need to obey the laws.[10] According to Uesugi, if there was resistance against overlord
power as Maeda argues, there are no signs of it in the texts of the village codes.[11] Over time, village codes came to look more
like the lords' laws in form and content, growing lounger in response to lordly expectations, especially in sumptuary matters.
Oide[*] presents a more detailed picture of the relationship between village codes and laws, using as her basic data Maeda's
and Hozumi's compilations. She first addressed the question of origins and saw an increasing trend of lordly influence in the
codification of village codes, especially in the second half of the Tokugawa period.[12]

The overlord's interference took various forms. It was most pronounced when the codes were a direct response to an order
from above. A striking example of such lordly initiative is the drafting in 1827 of a lengthy village code for the Kanto villages
that were grouped into leagues, kumiai . Regional village leagues, linking dozens of villages, developed in the last decades of
the eighteenth century.[13] Sometimes a third party outside the village (usually a headman from a neighboring village) was
made to participate in the process. This occurred frequently in Mino province in the late Tokugawa period on the occasion of
recalls of headmen.[14] More often, villages proceeded to draft codes on their own with regard to a specific overlordly
regulation. Other villages presented their codes to the lord on their own initiative. Thus, the "officialization" of village codes
was not simply the result of explicit coercive power from above. It was just as often sought from below by local officials
adopting a submissive posture, wanting to ingratiate themselves with the authorities by delivering voluntary proof that they
were "within the law."

[9] Ibid., 88; Maeda, Nihon kinsei sonpo[*] , 59.

[10] Uesugi, "Kinsei sonpo[*] ," 88-90.

[11] According to Uesugi (ibid., 88), Maeda argues that village codes reflected villagers' resistance ("Ho[*] to sonraku kyodotai[*]
," 211-13), although Maeda speaks only of areas of autonomy, especially in matters of justice.

[12] Oide[*] , "Kinsei sonpo[*] ," 11-19.

[13] For the text of the 1827 code, see Maeda, Nihon kinsei sonpo[*] , no. 128 (159-66). On these village leagues, see Anne
Walthall, "Village Networks: Sodai[*] and the Sale of Edo Nightsoil," Monumenta Nipponica 43, no. 3 (1988): 279-303.

[14] Oide[*] , "Kinsei sonpo[*] ," 112-23.

― 196 ―

What was the legal status of these village codes? Officially, the bakufu acknowledged Shikitari yaburi , the breaking of custom,
as a legal ground for suits, but in several lists of "civil" court cases compiled during the Tokugawa period, Oide[*] could find
only a single case based on that claim and one other case that explicitly argued a breach of a village code .[15] Obviously,
infractions of village codes were almost never brought before the overlords. In other court cases, overlords acknowledged
village codes if these supplemented their own laws without contradiction. The greatest claims for autonomy, however, have to
do with village policing and penal powers, which received detailed attention in many codes.[16] What was the bakufu's attitude
toward village justice?

In a rare explicit statement on the matter, a response to a query of 1829 on whether a village could impose fines for gambling
beyond those stipulated by bakufu law, the superintendent of finances (kanjobugyo[*] ), who also functioned as the overseer of
litigation, revealed that the bakufu did not forbid village justice but silently acknowledged it as a secondary and separate form
of justice , not as the lowest echelon of its own judicial machinery. In practice, village justice was allowed in matters the bakufu
did not or could not handle. Local administrators, however, could deal with village disputes kokoro jidai , "as they saw fit."[17]

From this we can conclude a general tendency on the part of the bakufu as well as the locals to keep things within the village,
[18] often through the mechanism of conciliation and apologies. This tendency created room for potential conflict, because
certain serious crimes had to be reported upward, for example, manslaughter, arson, even theft and gambling. Obviously not
all thefts were reported, although bakufu, domain, and village laws all stipulated that they be reported.[19] When

[15] Ibid., 24. This is case no. 58 (dated 1781) from the 160 cases collected in what was left after the Tokyo earthquake (1923) of
the Saikyodome (Record of judgments), cases that came before the bakufu's supreme court, the Tribunal, or Hyojosho[*] . The
other two collections perused were the Mokuhi (Private articles) and the Kujiroku (Record of suits).

[16] Maeda, Nihon kinsei sonpo[*] , 93-149; Uesugi, "Kinsei sonpo[*] ," 93-94.

[17] Oide[*] , "Kinsei sonpo[*] ," 75-76.

[18] Maeda, Nihon kinsei sonpo[*] , 31, 116.

[19] For the bakufu's general law, see article 56 of the Kujikata osadamegaki of 1742, in TKKk 3:189; for an English translation,
see Hall, "Japanese Feudal Laws III," 755. For reference to domain laws and village laws, see Oide[*] , "Kinsei sonpo[*] ," 78-79.

― 197 ―

gambling became a serious problem in the second half of the period, the bakufu explicitly allowed local punishment: in 1723
the bakufu proclaimed that it was no longer necessary to report gambling in bakufu and daimyo domains of the Kanto area,
and the law was extended to the whole country in 1788.[20]
That villages had certain penal powers did not mean that they were free to apply any kind of punishment. Death penalties
were strictly forbidden, although Mizumoto has found village codes in Echigo that listed the death penalty among possible
punishments for about fifty years in the second half of the eighteenth century, an extremely rare case according to Mizumoto.
[21] Maeda reports one village law in 1621 that prescribed the death penalty for gambling "after obtaining the opinion of the

overlord."[22] And we shall soon see that the village codes of Hozu stipulated that (village) samurai could cut down peasants
who offended them.

Banishment, or tsuiho[*] , was allowed when its application was not "private" or arbitrary: permission had to be secured in
advance from intendants; sometimes, however, reporting it after the fact made it legal.[23] There was one way of banishing
people that was not labeled tsuiho[*] : a person could be legally taken "off the [population] roster" (chogai[*] ) in response to a
request by village officials or relatives, which resulted in loss of residence and legal status for the person in question.
Banishment might also consist in expelling a person to a location just beyond the village border and letting him or her live (or
die) there in a makeshift shelter. Shinzo[*] in chapter x was a victim of both removal from the roster and expulsion beyond the
village. Perhaps the most well known form of village punishment is ostracism (murahachibu ), whereby a person was cut off
from social interaction and most mutual help. This particular form was generated by the village itself. It is found only in
village codes, not in village laws.

Throughout the period, village codes relied on overlord authority—"borrowed" its power, so to speak—by stipulating that any
infringement of the codes would be reported to the lord. Oide[*] sees a gradual

[20] Oide[*] , "Kinsei sonpo[*] ," 82.

[21] Mizumoto, "Kogi[*] no saiban," 295.

[22] Maeda, Nihon kinsei sonpo[*] : "Kenkyu[*] ," 103; "Furoku," 16.

[23] Oide[*] , "Kinsei sonpo[*] ," 85-87. See also article 26 of the goningumi rules of 1662 from Shimo-Sakurai, Kita-Saku district,
Shinano, in appendix 3.

― 198 ―

merging of the two kinds of justice, lordly and village justice. Punishments became lighter; physical punishments (banishment
or mutilation) tended to become rarer from the middle of the period onward and were replaced by fines; increasingly, crimes
were reported beyond the village. With the rural crisis of the nineteenth century, the village leadership was no longer able to
handle village affairs and came to rely on the overlords to resolve conflicts, thereby increasing considerably the time and
energy intendants spent adjudicating disputes.

Class and Status Codifications

I mentioned earlier that the bakufu's main status codification related to the separation of warrior rulers from commoner
subjects. It did not legislate status differences among peasants aside from setting the village officials (murakata , "village
persons") apart from the rest of the peasantry, from amongst whom those responsible for the tribute were singled out as titled
peasants. This legislation, however, is not to be found in the bakufu's best-known (but secret) code of 1742, the Kujikata
osadamegaki , or in the famous rural administrative regulations for intendants from the Kan'ei period (the 1640s), but in the
goningumi laws. That is where the overlords' expectations were made clear to the mass of commoners.

Many of the stipulations concerning tribute payment, open information about each peasant's share of the tribute, and the
village budget (which was discussed earlier and for which I referred to bakufu legislation addressed mainly to intendants)
were incorporated into village laws very early on. This can be seen, for instance, in the expansion of successive versions of
goningumi regulations for the same village. Appendix z offers the translation of one of the earliest village laws in bakufu
territories (twenty-one articles), from Shimo-Sakurai, in Kita-Saku, dated 1640. The second extant version, from ten years later,
consists of forty-seven articles, twenty of them from the bakufu's nationwide directions for villages, issued twelve months
earlier, in 1649/2 (see appendix 4).[24] By 1662 this village law, now comprising fifty-six articles, had acquired the form it
would have for the next two hundred years (see appendix 3).

[24] See Hozumi Shigeto[*] , Goningumi hokishu[*]zokuhen , 1:28-36; cf. TKKz 5:159-64 (no. 2789, arts. 28-47). On the controversy
concerning the historical status of this famous edict, see appendix 4.

― 199 ―

Before turning to these village laws, however, we shall examine perhaps the most famous of the shogunal Keian edicts, the one
of 1649, which was issued for all villages in the realm. Status is hardly an issue in the edict. Only article 16 deals with it by
prescribing cotton as the only fabric to be used by peasants (see appendix 4), a typical sumptuary regulation that under the
guise of ethics (frugality) pretends to regulate economics (consumption) but in fact targets politics by codifying status.[25]
Economic production and well-being per se are given far more emphasis. A basic concern is the maintenance and
reproduction of the means of production, that is, the peasants: they should be fed more when labor is more demanding (art.
10), keep in good health (art. 22), maintain collective productivity through mutual help (art. 29), and so on. Much advice is
given about ways to increase productivity and wealth. Among the items included are how to select grains (art. 6); injunctions
to sharpen hoes and sickles once a year (art. 7); how to build toilets to collect manure (art. 8); the necessity of supervising the
labor force, planning around the weather (arts. 11, 12), and disposing of surplus children (art. 18); encouragement to make a
profit (arts. 5, 21) and raise one's standard of living by being business-minded (shoshin[*] , akinagokoro ) (art. 17).These were
all things peasants did of their own accord, one would assume, but here they become the object of bakufu "national"
legislation.
Economic well-being (shinsho[*] ), its improvement, and its social reward in the form of respect (art. 30), as well as the scorn
that can be a product of its absence, are major themes of the edict (arts. 2, 17, 30, 35). Economic well-being is seen as the basis
of morality, because poverty leads to crime, a process described as an alarming cascade of worsening conditions: "poverty may
cause illness, warp the mind, cause one to steal, violate the law, wind up in prison, receive the death penalty, be crucified," to
which is added an ominous "etc." (art. 32). This is an interesting "economization" of virtue in the same alarmist terms the
Confucian scholar Yamazaki Ansai (1618-82) used to characterize the inevitable fall from one evil thought into the
abomination of abominations, the killing of the lord.[26]

[25] During the seventeenth century in Kishu[*] domain, village group headmen were allowed to wear pongee; headmen and
elders could use silk only in collars; and peasant dress was limited to cotton (see Hirayama Kozo[*] , Kishu-han[*]nosonpo[*]no
kenkyu[*] [Yoshikawa kobunkan[*] , 1972], 185).

[26] Ooms, Tokugawa Ideology , 220, 223, 247.

― 200 ―

In some sixteenth-century writings virtue was quantified and externalized through its identification with calculation and
foresight in the management of people by lords.[27] In Ansai's neo-Confucianism, virtue was politicized and internalized in
terms of the values of respect and silent loyalty.[28] In these village regulations, it was given an economic and social slant.
Article 31 looks very much like a transposition into an economic key of the famous passage in the Great Learning that
describes the expansion of the ruler's virtue throughout the realm. Here, however, the source is not the ruler but any subject
whose creation of wealth will be imitated within the village, in neighboring villages, in the province, and in neighboring
provinces. Another crucial virtue, filial piety, is transformed in a similar way: "No matter how devoted one may be to one's
parents, it is difficult to be filial when one is in misery (fuben )" (art. 32). This was one way that Confucianism reached the
masses. Buddhism and even folk belief in curses are marshaled in the Kita-Saku version of this edict to enjoin peasants to take
great care of their horses and bulls (art. 14).

Social status and standing are unambiguously linked to economic well-being: people will start to respect you, be nice to you,
and even promote your official status from a lower seat (in shrine associations and councils) to an upper seat if you improve
your material conditions (art. 30). We know from chapter 3, however, that village practice was altogether a different matter:
status was used to hold off competitors for political power who were agitating on the basis of their economic achievements.

Status issues were far more prominent in village laws and village codes than in the Keian edicts. The prohibition on wearing
anything but cotton appeared in Shimo-Sakurai's village law (art. 10 [see appendix 2]) in 1640, a year after it was proscribed by
the bakufu in its Kan'ei edict. In the next extant edition of that law (1662 [see appendix 3]), a sartorial status distinction is
made between headmen and peasants (art. 10), and it is stressed that one should not exceed the bounds of one's status on
ceremonial, that is public, occasions (art. 12). Moreover, the first article presents a hierarchized list of people, from house
owners to servants, and so on, down to outcastes and nonhumans. Elsewhere too

[27] Ibid., 24-25.

[28] Ibid., 219-20, 247-48.

― 201 ―

(arts. 2, 9, 22) the religious types are lumped together with the outcastes and criminal elements.

In numerous other village laws , similar concerns with status are expressed beginning around the mid seventeenth century.
Rules stress the forms of greeting and behavior that peasants (even former samurai) should use in the presence of samurai.[29]
Small peasants were to observe a specific etiquette in their dealings with the village leaders.[30] Interestingly, the latter
stipulation dates from the end of the century, when the small peasants, who had come into their own economically, were
agitating for higher status within the village.

Status maintenance, however, is threatened not only by disrespect. Self-assertive behavior can be an expression of economic
security: many elements of material culture—specifically, as Weber notes, modes of consumption—function as status
signifiers. Hence, regulations in this area reveal where threats to status distinctions developed. "In the building of houses, at
weddings, funerals, regular village council meetings, behavior and dress should be according to previous regulations and one
should not make light of status in anything" (1684, 1697). Permission was needed to build new-style houses (1698). Umbrellas
and geta might keep one dry when it rained, but peasants could not use them; instead they had to make do with straw hats
and, presumably, straw sandals (1828, 1838). "Only taka-holding peasants entered on the land surveys can build structures
with gates, walled fences, and eaves; branch houses of kin that have such features may keep them, but from now on this will
not be allowed" (1767).[31] Similar status-related prohibitions can be found also in village codes . To cite one rather late
example from 1832, cotton was prescribed for all seasons, and silk was forbidden, as were haori coats; umbrellas were only
for officials; geta and leather-soled sandals were forbidden; and sake consumption was to be limited to one or two sho[*] , that
is, two to three liters.[32]

These regulations, whether generated from within the village as codes or without as laws, were normative attempts to protect
social and

[29] Hozumi Shigeto[*] , Goningumi hokishu[*]zokuhen , 1:33 (1650), 68 (1679).

[30] Ibid., 1:235-36 (1698).

[31] Ibid., 1:129 (1684), 222 (1697), 236 (1698); 2:1159 (1828), 1193 (1838); 1:407 (1767).
[32] Maeda, Nihon kinsei sonpo[*] , no. 137 (177-78).

― 202 ―

economic positions via status signs whose symbolic power rested on their being relatively rare. Such regulations were,
perhaps, less necessary in the early period, when extended families prevailed and hierarchy was so obviously tied to
gradations of economic dependency that proper behavior was instilled through a socialization that could dispense with
written codes. However, the obvious leveling effects of the commercialization of the peasant economy necessitated
codification that generated differences.

When it was dissociated from the status holders' shrinking economic base (landless titled peasants), status ceased to function
as an obvious sign of class value (except as a déclassé one). The draining of the economic, however, revealed (and created) the
political, now the shrunken signified of status. Hence the struggle between, on the one hand, status holders (old titled
peasants) revalidating their now purely symbolic capital by reference to ancestry (when this status value was certified in land
survey records) and, on the other, challengers (new economically autonomous peasants) appealing to the same logic that had
created status as an expression of economic autonomy, which they now possessed, and had given political voice, which they
were now being denied.

Since village codes were drafted by the village elite, it is not surprising that from the beginning some of them quite explicitly
reveal a clearly defensive "class" effort to preserve privilege, status, and power, as well as their economic base. Such strategic
class-based linkages, however, are hard to demonstrate if one limits oneself to the textual study of codes. Links can best be
established by focusing one's research on a village or region whose socioeconomic structure can be identified. Very few
historians have engaged in this kind of research, but there are at least two such studies on which we can rely: Igeta Ryoji's[*] ,
for the village of Hozu in Tanba (near Kyoto), and Oide[*] Yukiko's, for the Mino region (Gifu).[33] They reveal a village
structure different from the three studied so far (the corporate village, the village dominated by dogo[*] , and the cooperative
village).

[33] Igeta Ryoji[*] , "Hokenteki[*] sonraku kyodotai[*] to mura okite: Tamba-kuni Hozu-mura gomyo[*] shudan[*] no sonraku
shihai," Doshisha[*]hogaku[*] (1960-62): 58 (1960): 52-78, 61 (1960): 80-109, 62 (1960): 27-56, 65 (1961): 23-45, 70 (1962): 66-98, 75
(1962): 87-108; idem, Kinsei sonraku no mibun kozo[*] (Kokusho kankokai[*] , 1984); Oide[*] , "Kinsei sonpo[*] ."

― 203 ―

Strategies for Elite Reproduction: Hozu Village, 1500S—1900

The Hozu locale (in Kameyama domain, Tanba province, which today is part of the township of Kameoka, some fifteen
kilometers, across the mountains, west of Kyoto) consisted administratively, at least from the 1580s, of two villages, a northern
(Kita-) and a southern (Minami-) Hozu, which together constituted a large community. In 1749 it numbered 376 households
with 2,068 koku. Kita-Hozu's 125 households formed one kumi, while Minami-Hozu's 251 were divided into three regional
kumi. In addition, there was a fifth kumi, a gomyo[*] ("five names [shoots]"), that was not regional, but comprised a number of
households in both villages. The gomyo[*] households (24 in 1616; 53 in 1669; 61 in 1801; 83 in 1872)[34] each bore the family
name of one of the five village founders, and they called themselves samurai throughout the Tokugawa period.[35] Officially, in
the eyes of the overlord in Kameoka (less than two kilometers away), they were titled peasants (otonabyakusho[*] ). In addition
to and separate from these five kumi there was a community of outcastes concentrated mainly (in 1872, 69 of the 74
households) in Minami-Hozu.[36]

In Tanba there were many groups like the gomyo[*] , samurai of the soil who maintained their family identity well into the
Tokugawa period.[37] What was unusual about the gomyo[*] was that, at some point in the sixteenth century or perhaps later,
these families had consolidated into an organization that maintained itself until the Meiji period

[34] Igeta, "Hokenteki[*] sonraku," 58:59-60. Hayashi Motoi has a different reading for gomyo[*] , namely, gobyo[*] (see Hayashi
Motoi, Kyoho[*]to Kansei , Kokumin no rekishi, 16 [Bun'eido[*] , 1971], 395). I am following Igeta, who did the original research,
and he confirmed the reading gomyo[*] .

[35] Igeta, "Hokenteki[*] sonraku," 58:55-56.

[36] For data on the outcaste population of Hozu, see Igeta Ryoji[*] , "Kinsei koki[*] no buraku sabetsu seisaku,"
Doshisha[*]hogaku[*] 110 (1969): 28-49 and 111 (1969): 53-84, esp. 54 and 66 for the population data; idem, "Mikaiho[*]
burakumin no iriaiken," Minshoho[*]zasshi 78, suppl. no. 1 (I979): 201-16; and idem, "Mikaiho[*] buraku to iriaiken: Kyoto-fu[*]
Kameoka-shi Hozu-mura no baai," Doshisha[*]hogaku[*] 136 (1975): 101-44.

[37] For instance, neighboring Umaji was ruled by two such families called ryomyo[*] ("two names") (see Igeta, Kinsei sonraku ,
esp. chap. 2; see also Igeta Ryoji[*] , Ho[*]o miru Kurio no me: rekishi to gendai [Kyoto: Horitsu[*] bunkasha, 1987], 89-101).

― 204 ―
Plate 3.
Hozu Village. From the bank of Hozu River, facing north, midway
between Hozu and the castle town of Kameoka. Photograph by author.

when they formally incorporated. The gomyo[*] was clearly an institution set up to defend class interests against other classes
of villagers, such as the bond servants (genin) employed by these households and the small peasants, who were kept small for
over two hundred years. The threat against the reproduction of this local upper class came from social change that took three
successive forms. First, as landed samurai these families lost institutional support from the overlords, who did not recognize
village samurai. Second, economic growth enriched the small peasants and led to the emancipation of many bond servants.
Finally, the gomyo's[*] own reproduction, if not checked, would weaken their economic base if with each generation holdings
were divided among new branch families. This class succeeded in meeting these challenges until the nineteenth century.
Because these families ruled the two Hozu villages, their strategies inform the village codes, which reveal where normative
defenses had to be erected.

The 1636 "village" code, eleven short, one-line articles, was explicitly called "Rules for the Genin Peasants of Minami-Hozu"
(the ones

― 205 ―

for Kita-Hozu are not extant).[38] All the articles are prohibitions to prevent status pollution: genin peasants cannot wear
swords or leather-soled sandals, use umbrellas, or go by samurai names; children should not address their parents as toto
(dad) and kaka (mom); a genin peasant cannot refer to himself or herself as midomo or to others as ore ; anyone rude in the
company of samurai shall be cut down.

The last three rules need some explanation. Toto and kaka are children's terms, obviously devoid of reverential connotations.
One can only guess why they constituted a problem in the eyes of these village samurai. Perhaps these samurai sensed that, as
Bourdieu has remarked, "concessions of politeness ... always contain political concessions"[39] and thus felt that they needed to
dictate intrafamily behavior—the personal is political. Midomo and ore are status- and rank-marking personal pronouns, the
former used by samurai to refer to themselves, the latter used for inferiors.[40] Thus, genin peasants were not to speak or
behave as equals to "samurai," a privilege the gomyo[*] , who perceived themselves as samurai, preserved for themselves. This
contrasts sharply with articles one finds in numerous village laws expressing the overlord's point of view that former samurai,
which legally the gomyo[*] were, are to be treated as peasants.[41]

The article affirming the samurai's right to cut down peasants encourages the swift action that had become outlawed by
Ieyasu's rule of 1603 that "the killing of peasants for no reason (muza to ) should be stopped, but if [a peasant] is guilty, he
should be caught, arrested, and punished in accordance with a verdict by shogunal officers (bugyojo[*] )."[42] The samurai of
Hozu village, one presumes, were not only going by a pre-Tokugawa practice: they were codifying it explicitly in 1636, an
unusual example of a code contradicting lordly law.

[38] Igeta, "Hokenteki[*] sonraku," 58:58.

[39] Bourdieu, Logic of Practice , 69.

[40] Nowadays, ore is used only as first-person pronoun; in Tokugawa times, however, it was also used as the third-person
pronoun for inferiors. For midomo , see the entry in Kojien[*] (Iwanami shoten, 1955); for ore , see the entry in Daijiten , 13 vols.
(Heibonsha, 1934).

[41] Hozumi Shigeto[*] , Goningumi hokishu[*]zokuhen , 1:33 (1650), 68 (1679).

[42] The rule outlawing the killing of peasants "for no reason" was the last article of a seven-article decree issued by Tokugawa
Ieyasu (the only one he issued about local governance) (see Kodama Kota[*] and Oishi[*] Shinzaburo[*] , eds., Kinsei
nosei[*]shiryoshu[*] , vol. 1, Edo bakufu horei[*]1 [Yoshikawa kobunkan[*] , 1966], 1).

― 206 ―
Slow economic growth made genin potential buyers of land. In order to control the transfer of land out of gomyo[*] hands,
"village" rules were issued in 1691 (and again in 1731 and 1746) prohibiting the sale of paddies and fields to genin; they could
acquire only homesteads with garden plots of dry fields. Similar economic restrictions were instituted against the outcastes
elsewhere. Other sartorial regulations in Hozu, such as the prohibition against wearing leather-soled sandals, are also
identical to those imposed on outcastes. In Hozu, therefore, which also had an outcaste community, small peasants, even taka
holders, were treated like outcastes in a number of respects.[43]

There were no restrictions on land sales among peasants. A samurai unable to find a samurai buyer and wanting to sell to a
peasant had first to report the price to the samurai council, whose directions for the sale had to be followed; that is, the council
set the price. In 1759 a new regulation forbade others than samurai to fish with nets in the Hozu River. The river was also the
source of additional wealth through the rafting of logs, but the village samurai had a monopoly on this business too.[44]
Although by the mid Tokugawa period they lost to the overlord their control over the use of the mountain, they still had a
monopoly on the use of the commercially very valuable standing trees.[45] These "samurai" clearly constituted an economic
class, both objectively, since they were the important landowners and entrepreneurs of sorts, and subjectively, because they
were very conscious of the interests that bound them together, which they defended through status and commercial
legislation that denied the small peasants economic opportunities.

In the eighteenth century, class differentiation occurred slowly within both the dominant and the dominated class in the two
Hozu villages: the greatest holdings—between 1596 and 1636 two samurai had increased their holdings from 56 and 65 koku
to 75 and 104 koku, respectively—leveled off to 50 koku, and peasants came to hold land. Yet the two groups remained distinct
economically: the gomyo[*] holdings clustered in the 5-20-koku bracket. In the 1720s and 1730s a number of peasants became
very small owners (see table 18 for Kita-Hozu),

[43] Igeta also makes this point in his "Kuchi-Tanba chiho[*] no hisabetsu shuraku[*] ," in Zenkindai Kyoto no burakushi , ed.
Buraku mondai kenkyujo, Kyoto no buraku mondai, 1 (Kyoto: Buraku mondai kenkyujo, 1987), 122.

[44] Igeta, "Hokenteki[*] sonraku," 58:60-61, 65:44, 24-25.

[45] Igeta, "Mikaiho[*] buraku," 104-5.

― 207 ―

Table 18.
Kita-Hozu Holdings, 1723-1872

1723 1740 1763 1820 1872

Koku G P G P G P G P G P

50+ 1 1

30-50 6 5 1 2

20-30 2 1 2 3 3

10-20 2 2 4 4 6 2 5

5-10 3 2 3 2 4 1 1 3 5

1-5 6 9 5 9 2 10 9 2 7

-1 3 33 5 32 1 31 1 17 4 40

0 1 33 3 35 3 42 2 69 4 23 [5]

Total 24 79 26 79 19 87 19 86 23 75 [5]

Total G + P 103 105 106 105 103

SOURCE : Igeta, Kinsei sonraku , tables III-17, III-37, and III-44 (pp. 191, 308, and 335).

NOTE : G = gomyo[*] ; P = non-gomyo[*] small peasants (fudai and genin houses not
included); numbers in square brackets refer m outcaste households; Igeta gives no
earlier figures for outcastes.

but they clustered in the category of less than 1 koku. In the following decades, however, many became landless again,
although the overall number of peasant owners remained at a higher level than before the 1720s. In striking contrast to
Kodaira village (analyzed in chapter 3), where titled peasants became economically indistinguishable from non-titled
peasants, the gomyo[*] in Hozu succeeded extraordinarily well in maintaining economic supremacy. Socially too the two
groups remained distinct. There was no intermarriage between the village samurai and the rest of the village population,
although there were numerous intermarriages between peasant owners and genin and fudai.[46]

Gomyo[*] strategically regulated landownership, so that the non-gomyo[*] peasant population (genin and small peasants)
consisted only of very tiny holders, who thus had to become servants and tenants to the gomyo[*] as a group. The restrictions
on land sales were only one part of this overall strategy. In 1748 rules forbade villagers to become tenants

[46] Igeta, "Hokenteki[*] sonraku," 58:65, 61:92, 101

― 208 ―

or seek employment elsewhere. The gomyo[*] thus kept wages for servants low and rents for tenant farmers high. In 1781-82
all this was challenged by demands that the wages of women servants and of rafters employed in the logging business be
increased, that the prohibition against service in other locales be lifted, that land rents be lowered, and, aiming at the heart of
the system, that treating peasants as hereditary vassals be stopped.[47] We shall return to this crisis shortly.

In order to prevent the downsizing of their holdings, the village samurai started practicing single inheritance by one son
around 1700. The other siblings, to whom the prohibition against outside employment obviously did not apply (the "village"
rules were for the non-gomyo[*] ), left the village to seek employment as servants or engage in trade in Kyoto or Osaka and
eventually return with their own assets.[48] This is interesting because the survival of older patterns of authority and power is
usually considered characteristic of "backward" areas; in the Kinai region these patterns are usually said to have disappeared
rather early. In the present case, however (which was not exceptional in southern Tanba), the proximity to Kyoto and Osaka
was an important factor in making the reproduction of these relations possible.

The small peasants responded to this class organization of the village samurai or landlords by forming their own, as a
religious (nenbutsu ) sodality. The first documented trace of its existence dates from 1733, but it may have been formed earlier.
Although it was divided geographically into several kumi, this organization spanned the two villages and functioned as an
official institution to defend the small peasants in their confrontations, physical or legal, with the gomyo[*] . The village
authorities, that is, the gomyo[*] , acknowledged the peasants' organization as a semipublic institution, since they
corresponded with it and allowed its representatives to testify in suits involving peasants. Other villages in the area, such as
Umaji, had similar organizations, which, although structured "horizontally" as fellowships, nevertheless had their own
hierarchy.[49]

[47] Igeta, "Kinsei koki[*] seisaku," 111:71. The "label" fudai , as opposed to the "reality"—a contrast perhaps best understood
again as tatemae versus honne—had disappeared from records after the 1750s; these were mainly newcomers from outside
who had been living with gomyo[*] households since the 1720s (idem, "Kuchi-Tanba," 120).

[48] Igeta, "Hokenteki[*] sonraku," 62:35-36.

[49] Ibid., 61:105-9.

― 209 ―

In 1781/6 an incident among a few villagers soon developed into a confrontation between the two classes.[50] One day a few
young peasants walked through the village wearing leather-soled straw sandals. A couple of gomyo[*] members from Kita-
Hozu berated the young men for breaking village regulations by wearing rain gear and beat them up. The angry victims
rallied the support of the nenbutsu sodality and went to the house of one of the assailants, where they were denied an
audience. Then the peasants decided to boycott the gomyo[*] by refusing to work for them in the fields and in a business of
fruit products owned by one of the gomyo[*] .

The gomyo[*] ordered an end to the boycott because it interfered with their livelihood, but the peasants refused because the
whole affair, they argued, was the result of the other side's unreasonable behavior. The assailants admitted to having beaten
up the young men but thought their behavior had been justified because the latter had broken the rule forbidding peasants to
wear leather-soled straw sandals. Such incidents obviously were symptoms of other economic and social tensions and signaled
the start of the various demands mentioned earlier. In the 1780s Hozu and the Tanba region, like much of the rest of the
country, witnessed peasant disturbances.[51]

Although they had maintained their economic dominance (see table 19), by 1800 the gomyo[*] were under considerable
pressure from the peasants, who for over a decade had been challenging their authority in the various ways described. At the
same time, the financially strapped domain lord came to rely on the gomyo[*] for "loans." The gomyo[*] exploited this situation
to enhance their members' status within the village in the following way.

The domain had a number of schemes (five of which were implemented within the domain) to borrow money from the
peasants and from other sources of revenue in Kyoto and Osaka.[52] Two of those schemes are relevant for the present
discussion. The first one consisted in obtaining individual loans from wealthy peasants. Thus, in 1764 and 1800 the daimyo
borrowed a total of 45 kanme (copper currency equal to 1,000 mon ) from three wealthy Hozu villagers (with combined hold-

[50] Ibid., 70:67-68.

[51] Anne Walthall discusses the unrest in Hozu following this "sandal incident" (Social Protest , 109-11).

[52] Igeta, "Hokenteki[*] sonraku," 75:89-97.

― 210 ―

Table 19.
Gomyo[*] Holdings, Minami- and Kita-
Hozu, 1801

No. of
Koku Households Koku

100+ 1

50-100 9

30-50 7

(17) (928.5)

20-30 9

10-20 9

5-10 7

1-5 11

-1 8 (436.1)

Total 61 1,364.6

SOURCE : Igeta, Kinsei sonraku , table III-36

(p. 308).

NOTE : The total number of households

was 376: 61 (16%) gomyo[*] and 315 (84%)
peasants. The total kokudaka was 2,068:
1,364 (66%) gomyo[*] , with an average of
22.3 koku each; and 703 (34%) peasants,
with an average of 2.2 koku each. It is
unlikely that the kokudaka includes plots
owned by outsiders. Also, most of the fields
listed under peasants were held under
tenancy, so that most of the peasants had
less than 1 koku, which amounted to some
vegetable plots around their homesteads.

ings of 227 koku). In addition, loans were obtained from whole villages in the form of advance tribute payments. As always,
repayment was a problem, redemption was postponed, interest rates were lowered, and so on.

The bulk of these loans, private and communal, were shouldered by the large landowners. Unlike the other villages in the
area, which were receiving interest payments with return of the principal postponed, "Hozu village" generously forfeited its
principal (and thus also the interest payments). But the gomyo[*] samurai-landowners wanted something in return from the
lord for the real capital they had provided him: symbolic capital. They requested official recognition of their samurai pedigree
as goshi[*] (landed samurai), which of course they had always considered themselves to be, this no doubt in order to shore up
their position within the village. A hundred years earlier decrees had explicitly forbidden the wearing of swords by peasants
and adopting the title goshi.[53]

[53] Igeta, "Kuchi-Tanba," 124.

― 211 ―

The lord, however, aware of the frictions and tensions between this group and the common peasants, was wary of further
accentuating class differences by making them official; he therefore proposed allowing thirty non-gomyo[*] village notables,
like the leadership of the nenbutsu confraternity, to wear haori (ceremonial coats, a status marker). The gomyo[*] objected
strenuously and got their way. For the lord, there were only titled and common peasants; ancestry did not play a part, since
membership in these groups normally shifted over rime. Moreover, the gomyo[*] would not be genuine landed samurai, but
cash samurai who had bought their status. The gomyo[*] , however, wanted the peasants to be officially their vassals (kerai).

Another point of friction with the lord developed when he "discovered" that all gomyo[*] households formed one single
separate kumi, while all other peasants were grouped in their own kumi. (Since Hozu was only a fifteen-minute walk from the
castle, this "discovery" two hundred years into the Tokugawa period is not to be taken too literally.)[54] The gomyo[*] were
supposed to form the leadership of each kumi; otherwise some peasant households would acquire a semihereditary position
as kumi heads, resulting in a new status next to that of the titled peasants. From the lord's point of view, village government
consisted of village officials, kumi heads, and peasants, and the status system contained only rifled and small peasants. The
lord, however, was in no position to follow through on the principles he had announced in the negotiations with the gomyo[*] ,
and he wound up recognizing for thirty-nine members of the gomyo[*] a venerable samurai pedigree going back over three
hundred years to Hosokawa Takakuni (1484-1531). Similarly, in neighboring Umaji, under a bannerman, thirty-nine of its
"titled peasants" became landed samurai in 1807, and two years later the remaining twenty-two were also promoted, for
reasons not unlike those in Hozu.[55]

Within a year of their promotion as officially certified landed samurai, the gomyo[*] had to deal with a consequence of a
scheme they had resorted to in order to control their peasant tenants, who had been agitating for lower rents. For some time,
possibly since the agitation in 1781-82, the

[54] This is not to suggest that Hozu was governed by the same house throughout the entire Tokugawa period. Hozu had six
different overlords during the seventeenth century. In 1702 the Aoyama daimyo and finally, in 1748, a Matsudaira house came
to rule the fifty-thousand-koku domain of Kameyama (from the castle town by that name, now Kameoka).

[55] Igeta, "Kuchi-Tanba," 124.

― 212 ―

gomyo[*] had been renting land to the outcastes as a countermove against some small peasants who had gone on a rent strike
in order to press their demand for lower rents. These renters had refused to continue renting some low-quality fields. The
gomyo[*] had simply replaced them with outcaste tenants. In Umaji, which had had no outcaste community, the ryomyo[*]
("two names"), a group like the gomyo[*] but consisting of two lineages, imported thirty-three outcaste households in 1808 for
similar reasons. These were pure peasants, who did not engage in the removal of dead cattle, a job that was taken care of in
the area by five other outcaste communities, one of which was Hozu's.[56]

This heightened the tension between the outcastes, who were now also strikebreakers, and the peasants, who had seen their
boycott come to naught. The tension erupted into violence when, in 1802, peasants repeatedly destroyed the grass and
firewood that the outcastes had gleaned from the commons, which the outcastes needed for their fields.[57] The disputants
were reconciled, but violence erupted some six times in the summer of 1808, forcing the outcastes to take legal action first
with the Hozu leadership, then with the domain. In 1811 this led to a settlement that favored the outcastes by increasing their
portion of the commons, and this was followed by an additional increase in 1815. But the peasants had already scored against
the outcastes in 1809: "in order to avoid trouble in this time of tension," the outcastes were not to deliver their rice tribute on
the same day that the peasants delivered theirs.

In this conflict with the peasants, the outcastes found allies in the village leadership, that is, the gomyo[*] , and the overlord. It
is clear, however, that the gomyo[*] played the outcastes off against the peasants in order to keep rents high and that the lord
supported this policy because the solidity of his tribute base, namely, the gomyo's[*] economic well-being, depended on it.

This outcastes' increased access to mountain land was strictly limited to foraging for agricultural use: none of the collected
material could be sold, and no (commercially valuable) standing trees could be cut down. Moreover, based on the Meiji data
(see table 20), it appears that the
[56] Ibid., 97.

[57] The following information is culled from Igeta, "Kinsei koki[*] seisaku," 111:55, 61-65, 71-72, 77, 81; and idem, "Mikaiho[*]
burakumin," 209-12. See also his "Kuchi-Tanba."

― 213 ―

Table 20.
Minami- and Kita-Hozu Holdings, 1872

Gomyo[*] Peasants Outcastes

Koku M K Total M K Total M K Total Total

80+ 1 1 1

50-80 3 3 3

30-50 6 2 8 8

20-30 12 3 15 15

10-20 16 5 21 21

5-10 7 3 10 1 5 6 16

1-5 6 2 8 16 7 23 31

-1 4 4 8 66 40 106 2 2 116

0 5 4 9 110 23 133 67 S 72 214

Total 60 23 83 193 75 268 69 5 74 425

SOURCE : Igeta, Kinsei sonraku , table III-44 (p. 335).

NOTE : M = Minami-Hozu; K = Kita-Hozu.

share of the area of mountain land available per household was heavily slanted in favor of the peasants, in a ratio of zoo to 28,
or roughly 4 to 1, a ratio that remained unchanged throughout the nineteenth century. In 1909, as a result of activism by the
burakumin descendants of the outcastes, this ratio was changed, to 4 to 3. But it was only in 1961 that equality was achieved,
at least in terms of land. Igeta suggests, however, that in terms of quality of mountain land, discrimination continued to
persist.[58] Although Igeta does not mention this, it may very well be that some of the disproportion in the common land
available for use to

[58] "Mikaiho[*] buraku," 108. A famous case for the clear right of access to commons by burakumin was settled in the Niboku
court (Nagano) as late as 1973, when it was decided that, contrary to earlier arguments, burakumin were indeed full members
of rural communities, with equal rights, and did not constitute separate, nonagricultural communities. Hozu outcastes were
involved in litigation for over a century and a half (see Igeta Ryoji[*] , "Meiji koki[*] no kenri toso[*] no ichi jirei: Mikaiho[*]
burakumin no byodo[*] iriai yokyu[*] ," in Nihon kindai kokka no ho[*] kozo[*] , ed. Nihon Kindai Hoseishi[*] Kenkyukai[*]
[Bokutakusha, 1983], 482-514. On the Niboku case, see Igeta, "Mikaiho[*] buraku," 101-6; and Aoki Takahisa, "Niboku iriaiken
jiken to hanketsu no igi," Buraku mondai kenkyu[*] 40 [1973]: 34-70). Frank Upham kindly alerted me to the existence of these
modern cases.

― 214 ―
peasants and to outcastes is to be explained by the fact that shares may have been calculated by holding size rather than per
household.

Through a number of devices and strategies, a clearly identifiable class of samurai-landowners succeeded in securing its
survival as a class for some three centuries. In part these village samurai resisted overlord authority by maintaining a posture
of samurai vis-à-vis the peasants whom they continued to rule as vassals. On the other hand, they also relied on the overlord,
who acknowledged their dominant position as titled peasants, and they succeeded in forcing him to officialize their position
further by granting them the title of landed samurai, which they had coveted for some two hundred years. Village regulations
were an important means to preserve this class's economic preeminence, and status was an effective weapon to maintain class
difference.

Village legislation was put into the service of similar overt class interests in Mino province, as Oide[*] Yukiko has
demonstrated. There the peasants were divided between kashirabyakusho[*] ("head peasants," i.e., titled peasants) and
wakibyakusho[*] ("side," or regular, perhaps marginal, peasants). The former held a monopoly on village power and relied on a
variety of customs that set them apart from the others: they had their own protective deities (ujigami) and wedding
ceremonies, and they did not marry across class lines. Such an upper-class peasantry, with its marks of distinction, was found
throughout the province, whatever the jurisdictional character of the overlords—daimyo domains, Tokugawa houseland, or
intendant territories.[59] The familiar scenario of economic growth threatening the preeminent position of the dominant
fraction of the peasant class also played itself out here. The village codes of this entire region were almost exclusively
concerned, not with cultivation, but with maintaining the relationship of domination between the two classes.

Who were these titled peasants? They defined themselves either as families who were registered on Hideyoshi's survey; or as
landed samurai "since the ancient past" privileged to have surnames, wear swords, and have their own protective deities; or
quite straightforwardly as large property owners. Descent and ownership, however, were not sufficient to defend them
against the economic threats that would eventually erode their distinction as a class. Hence the reliance on sumptuary vil-

[59] Oide[*] , "Kinsei sonpo[*] ," 111-14. The next several paragraphs are based on ibid., 113-22.

― 215 ―

lage codes controlling consumption, specifically its most visible form: houses. Almost universally there were detailed building
codes prohibiting regular peasants from having gates at their residences or entrance-ways, roofs with eaves, or rooms with
ceilings.

To these were often added other limitations, for example, on public display of wealth: stone tombstones were prohibited, as
were cremation huts and the use of palanquins. None of this legislation regulating status came from the overlords, although
they generally acquiesced in it. As in Hozu village and the Tanba region, this system did not meet any serious challenge until
the first decades of the nineteenth century (1800-1840), although related incidents occurred sporadically from the 1630s.

Increasingly, with the rising number of infringements, class and status conflicts were less likely to find a solution in the
divided villages. The overlords distanced themselves as much as they could from these problems and insisted on in situ
conciliation. For this purpose they relied more and more on the mediation of the leaders from surrounding villages. They also
began to inquire into the content of the village codes, and they even had these same leaders from other villages preside over
the drafting of new codes for problem villages. Needless to say, such village regulations are hard to reconcile with the standard
image of Tokugawa villages as consensual and autonomous communities. Even from the beginning these regulations were not
based on communal consensus, and they lost their autonomous character as well when they were redrafted by nonvillage
members.

In the (very) long run, the struggle to maintain status regardless of economic reality was lost. Ultimately, regular peasants
bought their privileges from the lord: six of them offered to pay three hundred ryo[*] to be reclassified as titled peasants; in
Ogaki[*] domain (100,000 koku) the lord even decreed that all peasants be promoted to that status in 1867.

Igeta sees the Hozu pattern of status manipulation as distinct from the more familiar dogo[*] and corporate villages; and it
should be added that they were also different from the cooperative villages of Fukuyama. Without the pyramidal, paternalistic
structure of villages tying individual households to individual magnate landlords, a whole "incorporated" class of large
landholders kept another class of very small landholders dependent as a group; many tenants had multiple "landlords," with
whom they interacted as a group. This, Igeta suggests, was typical of the mountainous parts of the larger Kinai region
(including South

― 216 ―

Tanba, Yamato's Yoshino district, Omi[*] , Kii, Settsu, Wakasa, and Mino).[60] However, historians have traditionally contrasted
villages in Kinai to dogo[*] villages, which were typical of the central regions, the Kanto and the northeast, as corporate
villages having more fraternity-like miyaza organizations in which the hierarchy of large and small landholders was not
directly related to patterns of domination. Igeta surmizes that this may have been typical only of the plains around the capital.
Elsewhere, in the mountainous parts, old village elites were able to neutralize the pressure for change prompted by economic
growth: the mechanism of status legislation was directed at a small peasantry, to whom the elite were not necessarily linked in
a one-to-one landlord-tenant relationship.

Ostracism (Murahachibu)

Murahachihu, perhaps the best-known village-specific penal measure, was absent from lordly legislation. There were multiple
ways of getting rid of undesirable community members—banishment, disinheritance, stripping of "civic status," and so on—all
of which had to be sanctioned by supravillage authorities in the form of prior requests or at least post factum reporting.
Ostracism, however, was purely a village affair. It was the result of a decision made "by the community" to sever all social
intercourse with a member and cut the household off from almost all mutual assistance. According to one popular
interpretation of the term (which literally means "eight parts [out of ten]"), those ostracized could count on community
assistance only for two "parts" of community life, namely, fire and funerals, but not for the other eight: coming-of-age
ceremonies, weddings, memorial services, births, sickness, floods, travel, building and repairs (e.g., roof thatching).[61] In
addition, ostracized members would not he greeted and were not allowed to participate in meetings or festivals. In other
words, it meant social and economic ruination.

In the following case we shall look at the process of ostracism and glimpse the tensions that crisscrossed village life. This case
is unusual and of special interest because the victim initiated a suit with the bakufu

[60] Igeta, Ho[*] o miru , 95.

[61] Akutsu Muneji, "Mura gitei ni tsuite no ichi kosatsu[*] : toku ni seisai sadame no shojirei," Gunma-kenshi kenkyu[*] 25
(1986): 99.

― 217 ―

Plate 4.
Rethatching a Roof, Mera Village, Minami-Izu, Shizuoka Prefecture,
1964. Community assistance was needed for the work but was denied when the
household was ostracized. Photograph by Haga Hideo, courtesy of the Haga
Library, Tokyo.

authorities and the case was settled in court.[62] This incident took place in Sasaemon village in Musashi province, a medium-
sized village of 1,000 koku with some seventy households (now part of the town of Sugito in the eastern part of Saitama
prefecture), located along the highway to Nikko[*] in a large bakufu territory of over 110,000 koku, which an intendant
administered from Edo. The incident started at the end of 1827/2 and was resolved with surprising speed—in just two months'
time—at the intendant's office in Edo.

The plaintiff was one Chojiro[*] , who because of ill health was allowed to be represented by his son Kichitaro[*] . There were
nineteen defendants, including the headman, kumi heads, and titled and nontitled peasants. Chojiro[*] was probably a well-
established peasant; he even had a branch house, which was headed by Takejiro[*] , a kumi head in another village. Chojiro[*]
belonged to the Terabori kumi but was closely related both

[62] Fuse Yaheiji, "Shiryo[*] : Mura hachibu no sosho[*] ," Nihon hogaku[*] 23, no. 3 (1957): 96- 109.

― 218 ―

geographically and socially to two other kumi. The defendants were spread over Chojiro's[*] own kumi and a fourth one, the
Tosho kumi, a sign that the community was probably divided.

At noon on 1827/2/23 three village members—Kyujiro[*] , Shigeshichi, and Seijiro[*] —visited Chojiro[*] at his home and
insisted that he immediately repay them loans totaling more than five ryo[*] . Kyujiro[*] and Shigeshichi lived in Chojiro's[*]
kumi, while Seijiro[*] lived in the Tosho kumi. Kyujiro[*] was "homeless" (mushuku) or not registered anywhere. He had been
punished with medium banishment for some crime, which meant that his home and land had been confiscated and that he
was not allowed to travel to or reside in any of a number of provinces, including the one where Sasaemon village was located.
He was an illegal resident, an offense punishable by maximum banishment.[63] He lived with and worked for Shigeshichi's
father. Seijiro[*] , who was registered in a village of another district, was married to Sen, daughter of the Tosho kumi's head.
He and his wife had gone to live with his in-laws after causing some trouble with his own parents in his native village.

When he was confronted by the trio, Chojiro[*] expressed great surprise and did not recall anything about the loans. A quarrel
ensued, and the three alleged creditors decided to help themselves to some bales of rice and other things. A fight broke out,
and they started destroying tools and farming implements. One of Chojiro's[*] sons ran off to the branch house to alert
Takejiro[*] . Soon after Takejiro[*] arrived on the scene, about a dozen peasants from Chojiro's[*] kumi appeared, led by the
kumi head. Now negotiations started. Chojiro[*] was ready to pay off two of the three claimants, leaving the "homeless" out of
the bargain. The headman, however, was adamant that Kyujiro[*] be part of the settlement negotiated by the whole gang of
villagers. Thus Chojiro[*] was forced to pay over three ryo[*] , rather than the claimed full debt of five ryo[*] , a debt he never
acknowledged. It was Takejiro[*] , from the branch house, who put up the money.

That was not the end, though. Two days later a gang of twelve young men from Chojiro's[*] kumi appeared at his doorway and
shouted that he had been ostracized. When Chojiro[*] asked why, they did not

[63] Light banishment was accompanied by confiscation of land; medium banishment by confiscation of land and homestead;
maximum banishment by confiscation of land, homestead, and household effects (see Tsukada Takashi, "Kinsei no keibatsu,"
NNS 5:97).

― 219 ―

answer, but turned and left, hurling insults at him as they went out. Then the Tosho kumi head, Seijiro's[*] father-in-law, and
three other members of that kumi came, and they also declared him ostracized. Chojiro[*] , now outlawed by two of the four
village kumi, including his own, looked for mediation to the headman, one of whose main tasks was to settle intramural
disputes before they became lawsuits. Once again, the headman stubbornly refused to intervene, which left Chojiro[*] only one
recourse, namely, to lodge a suit with the intendant, even though his chances of receiving any response other than an order to
settle the matter internally were slim.

The suit was drawn up according to the required format and presented a clever argument, which is why Fuse Yaheiji, the
scholar who discovered it, thinks that Chojiro[*] may have relied on the legal expertise of a suit inn (kujiyado ) in Edo. The suit
document acknowledged that matters such as these were usually settled out of court in order to avoid lawsuits but stated that
the present issue was not a frivolous one, because the intendant's immediate interests were at stake. The main argument was
that ostracism would prevent Chojiro[*] from producing tribute or from performing horse corvée on the Nikko[*] road (which
was of particular importance to the bakufu, since it led to Ieyasu's shrine). The result was a quick summons for the defendants
to appear by 10:00 A.M. on 3/28 at the intendant's office in Edo. They were to bring a written reply to the accusations (which
would be read to the plaintiff) for a confrontation with the plaintiff; otherwise they would be found guilty and fined
accordingly.

From the settlement document drafted a month later, we can reconstruct what happened at the hearing. First of all, the illegal
"mura lien," Kyujiro[*] , was not included. There were a total of sixteen defendants with an unrelated village headman as their
spokesman. Another village headman functioned as mediator between the defendants and the sole plaintiff, Chojiro's[*] son
Kichitaro[*] . Kichitaro[*] maintained that the money dispute and the ostracism were related. The defendants not only denied
such a link but also denied that any declaration of ostracism had ever taken place! Typically, the mediator forced both parties
to acknowledge that they each were in the wrong. The mediator considered the money matter to have been resolved before
the trial, since the debt had already been paid off, although not to the complete satisfaction of the three claimants. Therefore,
the plaintiff withdrew his accusation of a linkage. The defendants admitted to the declarations of ostracism and agreed to

― 220 ―

write an apology and reinstate the plaintiff as a fully privileged community member. As was customary at such settlements,
both parties agreed that the suit had been settled and that no new suits would follow.

The affair had split the village into two camps, illustrating, perhaps, Oide's[*] thesis that while there were many instances of
intravillage settlements made by extravillage authorities in the early period, toward the end of the Tokugawa period village
divisions often ran so deep that recourse to outside authority became increasingly necessary.[64] That village ostracism was
open to abuse is obvious, but it was not until the year after the case just discussed that the bakufu began to regulate it,
forbidding young people to initiate it.[65] An indication, perhaps, of increased village strife, other cases must have drawn the
bakufu's attention at the time. Ostracism, like banishment, was a solution that created more problems, producing bankrupt
peasants, "homeless" types, and vagrants, which had been a serious concern of the authorities since the eighteenth century.

As far as the details of the above case are concerned, there are obscure areas that raise unanswerable questions. Chojiro[*]
denied liability, and yet his branch house was willing to pay off the alleged debt. We do not know the reason for this turnabout
or the nature of the pressure Chojiro[*] put on Takejiro[*] to forward the money. There are also overtones of gang behavior and
extortion, perhaps fueled by the marginal members of the village, who seem to have received protection. Yet it is also unlikely
that the three so-called creditors would have wholly fabricated their claims. They and their young supporters may have
created an atmosphere of opportunity for Chojiro's[*] enemies within the village to settle old scores; the headman does not
appear to figure among his friends.

Indeed, the headman deliberately allowed things to escalate, first by refusing to allow the illegal resident Kyujiro[*] to be
excluded from the settlement negotiated at Chojiro's[*] doorstep and then by letting the ostracism stand. He did this even
though he apparently lacked the support of half the village and he might get into trouble during the course

[64] See, e.g., Yamamoto Yukitoshi, "Kinsei shoki no ronsho to saikyo: Aizu-han o chushin[*] ni," Kinsei no shihai taisei to shakai
kozo[*] , ed. Kitajima Masamoto (Yoshikawa kobunkan[*] , 1983), 79-127.

[65] Fuse, "Murahachibu," 107-9.

― 221 ―

of the interrogations if the matter of Kyujiro's[*] status were to surface. He may have been responsible for the fact that no
mention was made regarding Kyujiro[*] that might alert the intendant, which makes one wonder what he held over Chojiro[*]
that kept the latter from playing that card: could it have been the threat to force payment of the remaining two ryo[*] ?

Ostracism was different from banishment (tsuiho[*] ), whereby one was expelled from the village; however, the two were
similar in that ostracism was in effect intravillage banishment. Extravillage banishment could be triggered by a vote whereby
someone was declared, without any material proof, to be the perpetrator of some crime. One case will illustrate this point.[66]

In Iwaya village (Tanba province), a certain Shohyoe[*] reported the theft of some of his belongings to the village officials, who
decided to take a vote in the village on who the thief might be. The ballot pointed to Rokuzaemon, who lived across the street
from the victim; Rokuzaemon, however, had absconded before he could be punished. He was sentenced in absentia to
banishment of five ri (20 km) beyond the village, with the proviso that if he returned home and was reported, the father of the
fugitive would not resist his whole family's banishment. Three years later the son returned secretly. When he was discovered
and reported, a powerful village elder interceded with the village council on the man's behalf and succeeded in having the
entire family's banishment suspended. The son, however, had to leave again.

After her son's second expulsion, the mother started slandering Shohyoe[*] , who, although a victim of a theft, had been the
source of the family's suffering for a crime that had not even been proven. Again the family was reported to the village
officials; an "investigation" was launched, and it was decided that the woman had become deranged and was thus not
accountable for what she had been saying. A year and a half later (1717/5), however, Shohyoe[*] and his wife, finding life
unbearable, moved to the other part of the village, which since 1664 had been divided between two jurisdictions. This example
introduces another form of village justice, one based on the results of the ballot box.

[66] Taoka Koitsu[*] , "Murahachibu to tsuiho[*] ni tsuite," Chihoshi[*] kenkyu[*] 12 (1954): 4-6.

― 222 ―

Fighting Crime by Popular Vote (Irefuda)

Tokugawa villages, so the generalization goes, were ruled by consensus, not by the democratic principle of majority rule. In his
study of Japanese law, published in 1991, Carl Steenstrup is quite categorical about what he calls "the Japanese custom" in this
regard: "Decision by majority was only known among monks in monasteries. It is an Indo-European custom, the basis of
'democracy'; and in Japan only became temple practice, because of the import of the monks' rules (vinaya ) from India. The
Japanese custom is to discuss, until agreement is reached. And if no agreement, no decision."[67] Kodama Kota[*] , however,
who wrote the entry "irefuda" for the first volume of the Kokushi daijiten (1979), mentions that election of certain officers by
vote was practiced in various sectors of society: for certain positions at the imperial court, in Kato[*] Kiyomasa's domain (1562-
1611), in the Pure Land sect, and often for the election of village officials but also for the allocation of newly developed paddy,
the assignment of village corvée in connection with the alternate attendance system or public works, and the setting of prices
for goods.[68]

Consensus in Tokugawa villages was not a universal, communal consensus. Many historians of peasant societies have
romanticized it by taking at face value a decision making that presents itself as consensual agreement. Others, pointing to its
representational function, understand consensus to be a ploy to consolidate elite power by making dissension within that elite
seemingly less real (because not publicly voiced).[69]

[67] Carl Steenstrup, A History of Law in Japan until 1868 (Leiden: E.J. Brill, 1991), 132; the emphasis is in the original, and
Steenstrup cites no sources.

[68] KDJ 1:828. Anne Walthall mentions an occasion in a village when six women signed ballots used for voting upon village
officials ("The Life Cycle of Farm Women in Tokugawa Japan," in Recreating Japanese Women , 1600-1945, ed. Gall Lee
Bernstein [Berkeley: University of California Press, 1991], 68-69).

[69] James Scott's Moral Economy of the Peasant (New Haven: Yale University Press, 1976) is representative of this romantic
view of village politics in general, which Irokawa Daiichi shares in regard to the Tokugawa peasantry (see his "Survival
Struggle of the Japanese Community," Authority and the Individual in Japan: Citizen Protest in Historical Perspective , ed. Victor
Koschmann [Tokyo: University of Tokyo Press, 1978], 257-58). Samuel Pop. kin, in The Rational Peasant (Berkeley: University of
California Press, 1979), is specifically skeptical of "consensus decision making" (58-59). Carol Gluck, who emphasizes the
practical local exploitation of ideology in her Japan's Modern Myths: Ideology m the Late Meiji Period (Princeton: Princeton
University Press, 1985), shares Popkin's perspective in this respect but winds up refracting ideology through the play of very
localized interests to the point that its specific power evaporates; yet trees can have a forest effect.

― 223 ―

Whatever the interpretation of "consensual" decision making, one important area of life was decided on by vote in many
villages, the fight against crime. Upon the recurrence of some crime such as arson or theft, a vote could be taken on the
identity of the putative offender. The person who received the majority of votes (together with his kin and neighborhood
presumably) would then be incriminated, as would anyone who did not participate in the vote.[70] A number of village codes
refer to the procedure as irefuda (also read nyusatsu[*] ), or "putting a tag [ballot] in [a box]."[71] Recently Ochiai Nobutaka has
looked into this practice in northeastern Japan more systematically.[72]

Villages were not allowed to punish crimes of arson and theft, but they were empowered to search for the criminals and
apprehend them. With regard to arson, Ochiai reports the following case from Higashi-Kami-Isobe village (Kozuke[*] province,
in present-day Gunma prefecture). After several instances of arson over the span of a year, the following decision was made at
the end of 1777 by six headmen of the fief, six kumi heads, and nine peasant representatives. A vote would be taken
concerning the possible perpetrator and the "winner" would be thrown in jail; his daily ration of five go[*] (0.9 liter) of rice was
to be provided at village expense. If the fires continued, however, a second vote was to be taken and the new "winner" jailed. If
the real arsonist were caught, he was to be handed over to the authorities, and the suspect already in custody would be freed
by the decision of the village officials.

A certain hapless Seigoro[*] was thus voted into jail, though not without the protest of his family and kin. One of the six
headmen consulted with his colleagues and the lord of the fief and then filed a petition with the fief's commissioner of finance
for an investigation into the damage suffered by the village, since this had hindered the performance of their highway portage
corvée. A number of officials from other villages and

[70] For a case in Imabori that probably dates from 1639, see Tonomura, Community and Commerce , 206 (no. 290).

[71] Maeda, Nihon kinsei sonpo[*] , 45-46.

[72] Ochiai Nobutaka, "Kinsei sonraku ni okeru kaji, nusumi no kendanken to shinpan no kino[*] ," Rekishi hyoron[*] , no. 442
(1987): 63-84.

― 224 ―

Buddhist priests were involved as go-betweens; the case went to Edo and was settled out of court a year later in 1778/3.
Nevertheless, the lord's authority upheld the decision already made by the village.

Crimes like arson and theft, especially when they recurred, in small lineage-based communities might easily lead to feuds
between groups of families. To prevent such intravillage vendettas, it was foremost necessary to remove any possible
justification for retaliation against putative criminals. This logic may explain the recourse to balloting, which provided at least
the semblance of a restoration of order, an effect that may also have been obtained by the mere talk of resorting to a ballot, or
by postponing the counting of ballots. Moreover, the victim of the ballot was most likely to be someone with few allies in the
village, who would thus divert and diffuse more serious interlineage tensions.

While in principle arson and theft had to be reported and the suspect punished by the overlords, and villages certainly in
principle had no power to impose the death penalty, unlike some corporate villages in pre-Tokugawa times, historians are
finding more and more exceptions. As already mentioned, Mizumoto reported the only cases he knew: some villages that had
the death penalty in their codes for a few decades in the late eighteenth century. Ochiai reports several codes stipulating the
stoning of arsonists, followed by the burning of the arsonist's house and the expulsion of his or her family or, if the arsonist
fled, the stoning to death of a family member.[73]

The village of Higashi-Kami-Isobe also had to deal with thieves. In 1786, a group of 132 peasants in Higashi-Kami-Isobe asked
its six village headmen and seven kumi heads to consider allowing searches of homes for stolen goods. This option was
weighed but postponed until a new theft took place, when a request for a search would be submitted to the village officials. If
the culprit were found, he would be fined or put to death; if he were not caught, a vote would be taken and the highest vote
getter fined. Here again we have the assumption of the right to impose the death penalty. This clause also appears in a village
code of 1723 in Mino province. Following the theft of rice a half-year later, a house search was approved by the village
officials; the suspected thief was caught and fined 15 kanmon , but he was not reported to the authorities. Such searches could
be made by the victim himself, the vil-

[73] Ibid., 67.

― 225 ―

lage officials, or the whole village and were sometimes conducted over several villages, depending on the local custom or the
nature of the case.

As mentioned by Kodama, voting was practiced also for the election of village officials. In such cases, however, ballots were
signed! It is not difficult to imagine what this meant in terms of village power alliances, but sodanfuda[*] ("consultation
balloting") or collusion was forbidden. The same prohibition held for crime voting, although it seems that in that case the
ballots were not signed; scratching the ballot was sufficient according to some village regulations. Sometimes swearing oaths
before the gods and drinking holy water (a practice also followed at the beginning of ikki, or uprisings) were required prior to
balloting, the peasants swearing that they would not be swayed by "favoritism or prejudice." The setting at a Shinto shrine, the
taking of an oath, and drinking of holy water were by no means common, but point to a religious origin. And indeed,
anonymous voting has ancient roots and was widely practiced in antiquity and medieval Japan, as the medieval historian Seta
Katsuya has demonstrated.[74]

Rakushogisho[*] ("dropped written oaths [before the gods]"), as they were then called, were also used to anonymously identify
a criminal. Some of these "oracles" were not votes but single anonymous accusations dropped in front of shrines. Since these
were viewed as signs from the gods having no link to the profane world, if one was picked up, the finder was obliged to
implement it.

Anonymous accusations were frequently posted on walls and were a means to openly denounce corruption at temples or to
reveal plots. Nevertheless, posted denunciations were outlawed in antiquity and also by the pre-Tokugawa bakufu authorities,
[75] who preferred to pursue criminals through their own courts, where, incidentally, they also relied on oracles to solve
unclear cases. The Tokugawa village practice of irefuda (which Seta does not mention) undoubtedly has its origins in these
rakushogisho[*] . In Tokugawa times, however, the "election" of criminals (in contrast to that of officials) was not only initiated
by the village but also ratified and even ordered by the overlords.

On 1696/12/27 three bales of tribute rice were stolen from the village storehouse in Fuse village, Shimosa[*] province (Chiba
prefecture). Two days later 131 peasants and the village headman and kumi leaders
[74] Seta Katsuya, "Shinpan to kendan," NNS 5:58-86, esp. 65 ff.

[75] Ibid., 59, 61-62.

― 226 ―

decided to resort to a vote; the winner would be banished. Guards were stationed throughout the village to prevent collusion,
and oaths were required. After the ballots were counted, two peasants were banished, and three others, who received only
one or two votes, were condemned to house arrest. Two of the latter fled to a temple, the Buddhist priest interceded on their
behalf, and their sentence was suspended. The fields of the two who were banished became village property, and their houses,
horses, tools, firewood, and other belongings were distributed among the family members. All this for the theft of three bales
of rice. But this was no ordinary rice: it was the lord's tribute rice. The results were reported to the lord, who may very well
have put pressure on the village to find a culprit.

There were numerous variations on the ways penalties were calibrated to the ranked outcome of the ballot or the number of
votes received. In descending order, those who received the highest number of votes might be fined five kanmon plus
ambulatory exposure, and those who received the second and third highest number might be fined three and one kanmon,
respectively. Or those with fewer than five votes might be declared not guilty, and so on. This system had the additional effect
of putting on notice those community members whom "public opinion" judged to be of somewhat questionable character.

In the early period, punishments for thefts were very harsh, banishment without mutilation, which rendered a person a hinin,
or registered beggar, being among the lightest. But even in 1711 there were villages where the criminal would be expelled
after his ears and nose were first cut off, marking him forever, in a most visible way, as a criminal. It seems to have been the
custom, at least in the early period and in some locales, not to physically mutilate female offenders but to strip them and
parade them through the village, subjecting them to what one could call gaze mutilation.[76] Of course another possibility was
ostracism; or making convicted thieves wear red caps, or ring bells at weddings and funerals (very public occasions) until the
next thief was caught (which could be weeks, months, or years), or treat the village to three sho[*] (5.4 liters) of sake every year
on a particular date; or even assigning them field guard duty until the next thief was arrested. These measures ingeniously
mobilized time to protract indefinitely the effect of public exposure (yet another form of punishment, usually limited to three
days).

[76] Tsukada Takashi, "Kinsei no keibatsu," 96.

― 227 ―

Plate 5.
Exposure of a Thief. This form of village punishment usually lasted
three days. Stolen bamboo shoots lie next to culprit. Woodblock print, c. 1745,
from Miyatake Gaikotsu, Shikei ruisan , Miyatake Gaikotsu chosakushu[*] , 4
(1985). Reprinted with permission from Kawade shobo[*] shinsha.

― 228 ―
If theft was a crime, so was not reporting it, perhaps according to the same logic that informed crime voting: it neutralized
reasons for private intravillage vendettas by moving them into the public sphere. In some villages, if a theft came to light that
had gone unreported, the victim received the same treatment as the thief: banishment. And many village codes specified
rewards for reporting thieves to the village authorities, which leads one to believe that although nonreporting was a serious
problem for the officials, it was looked upon positively by the common peasant. And there was often opposition to the practice
of irefuda. Otherwise why would some village rules stipulate that those who argued against it be treated as if they were guilty
of the crime itself?

That crime voting often only created new problems was well understood by the peasants, hence their maneuvering to
postpone the vote and perhaps increase their efforts to find a criminal. To accommodate such concerns, some villages decided
not to open the ballots after the first vote, but to resort to a second vote after a new occurrence and then combine the results
of both ballots. It was often the village officials who favored the voting procedure and the common peasants who resisted it.
The latter often saw in it a weapon officials wielded as a means of control over the ordinary peasants. After all, the voting
occurred in the presence of the officials, who were immune to the result because the victim was never an official.[77]

In 1769 in Tomikura village (Shinano) there was a rash of thefts, and the village officials conducted investigations in several
neighboring villages. Then a rumor was started, most likely by these other villagers, that the thief was in Tomikura itself. Thus
a decision was made by "all the village officials" and "all" the peasants to proceed with a vote, apprehend the largest vote
getter as the thief, and confiscate his homestead and fields. In order to thwart possible opposition, it was also decided that if
the designated criminal did not abide by the rules (one assumes by fleeing or suing), then the village officials would take their
suit against the seventy-six peasants of the village for obstructing village justice to the shogunal authorities.

This was hardly the much-heralded self-determining practice of autonomous villages: the officials invoked the specter of a suit
to enforce their will. And there are other instances of the village leadership's in-

[77] Ochiai, "Kaji, nusumi no kendanken," 79.

― 229 ―

voking the mobilization of shogunal authority (including the possibility of torture) to confront the rest of the peasants with
crime voting. While the village leadership allied itself thus with the overlords, the peasants often fled to temples for asylum or
sought to stop suits brought against them by the leadership and the overlords.

Just as Oide[*] saw the increasing convergence of village laws and lordly laws throughout the second half of the Tokugawa
period, Ochiai has documented a similar trend in the practice of village justice. One example of this trend is yet another case
from Higashi-Kami-Isobe village. In the aforementioned arson case of 1771 extramural authorities became involved mainly
because of the protests leading to a suit by relatives of the jailed Seigoro[*] . In the far less serious case of theft in 1786, the
matter was resolved intramurally, and the voting was postponed as a last resort, the village exercising its right to conduct
house searches instead. When the culprit was found in this instance, he was fined, but he was not reported to the authorities.
In a new village regulation from 1838, however, the only solution set forth for theft was the most extreme one: if the village
authorities heard rumors (fubun[*] ) about thefts from fields, they would immediately proceed to a vote, the result would be
reported to the bakufu, and the "guilty" would be banished from the village.

The reform of the village leagues instituted in the 1820s in the Kanto area strengthened further this convergence of village and
lordly justice. In such village leagues, after the 1820s, voting no longer took place in a single village, but was conducted
throughout the village league as a whole. Thus, when a theft occurred in one village, a vote would be taken in all the villages
and the winner would be reported to the bakufu.[78]

It seems that at this point the village had abandoned any claims for self-governance in the area of penal jurisdiction.
Confronted with increased social problems, the village leadership increasingly functioned in reality, if not in principle, as an
executive branch of the overlords. This may suggest an answer to a question Anne Walthall raised in her article on village
leagues.[79] Walthall represents the leadership of these village leagues as supporting "peasant interests," although some
peasants must have been served better than others, since historians have

[78] Ibid., 81.

[79] Walthall, "Village Networks," 286.

― 230 ―

stressed fractional divisions among the peasantry. She then raises the question why, unlike in earlier cases of village protests,
no league leadership was ever punished by the bakufu. If the above practice in penal matters is any indication, the answer
may well be that the bakufu had no reason to doubt that the village leadership was doing its bidding.

Another aspect of penal jurisdiction that should be mentioned is the categories of people who were not subject to village
justice, even if they were residents. Registered blind people, registered beggars, and out-castes were status marginals who
were not attached juridically to villages, districts, domains, or townships. They were subject to regional (e.g., the Kanto, with
its center in Edo, or the Kinai around Kyoto) or national jurisdictions of their status organization. Intrastatus civil and criminal
matters of these groups were handled by Danzaemon in Edo, for beggars and outcastes of most of eastern Japan, and by
Kengyo[*] in Kyoto, for all the registered blind.[80]

Justuce by Ordeal

Pinning down a criminal by voting after taking an oath before the gods gave the outcome the character of a divine verdict—
one whose numinous power suffered serious entropy later in the period, as we have seen. Genuine ordeals, however, were
still used on a number of occa-

[80] Henderson, Conciliation , 1:91; Tsukada, "Kinsei no keibatsu," 102-4. Danzaemon and Kengyo[*] , originally names of
persons, became hereditary title names attached to the respective leadership positions. Although Kengyo[*] meted out the
death penalty as late as 1696, his powers seem to have gradually weakened in the second half of the Tokugawa period. In the
Goshioki saikyo cho[*] , a list of 974 cases of punishments of prisoners (divided into 231 categories) between 1657 and 1699,
Tsukada Takashi has found two clear instances in which the Kengyo[*] decided in favor of the death penalty (Tsukada, "Kinsei
no keibatsu," 103-4). The first instance, in Edo in 1683, involved an accomplice in an affair another man was having with the
wife of a blind man. The blind husband had killed the accomplice and wounded the adulterer. Although the law permitted a
husband to kill his wife and her lover if he caught them in the act (Hiramatsu, Kinsei keiji soshoho[*] , 581), the blind man was
nevertheless entrusted to his zato[*]nakama guild (the guild for the registered blind), perhaps because he had killed the
accomplice and not the adulterer. The guild asked Kengyo[*] in Kyoto to have him wrapped up in a mat (sumaki ) and
drowned. He was handed over to Kengyo[*] , who had him executed that way. (This punishment, absent from the bakufu's
quite elaborate penal code, seems to have been used often by yakuza as a private death sentence [KDJ 8:141].) The second case,
in 1696, concerned a blind arsonist from Echigo (Niigata) who was sentenced to be burned at the stake in Edo.

― 231 ―

sions in the first half of the seventeenth century, precisely when overlords were trying to monopolize the exercise of violence,
wresting away penal powers for serious crimes from local power holders, whose practice they declared a "private" and illegal
justice. Since ordeals were still practiced well into the Tokugawa period, one might assume that they were local "illegal"
affairs. Yamamoto Yukitoshi, however, has presented a number of incidents from Aizu domain that clearly show that this was
not the case.[81] The overlords often permitted trial by ordeal, sometimes ordered such trials, and even adjusted their own
verdicts according to the results of subsequent ordeals. It should be noted that all of Yamamoto's cases concern conflicts
between villages and that many of these intervillage disputes revolved around new borders drawn as a result of the murauke
system. In some cases, common forage land was divided among five villages each belonging to a different overlord.

The first case concerns a dispute of 1619 between two villages, Tsunazawa and Matsuo, about the use of a mountain area
commons. The conflict had escalated into an armed confrontation that resulted in one death. The domain took several
depositions, made arraignments, and dispatched investigators, but the facts could not be established. This was in the middle of
the winter. In midsummer, with tensions at a peak, the two villages were ordered to resort to a fire ordeal in the presence of a
domain official and villagers from the area.

As it turned out, the lord, unable to determine the common border on the mountain, had arbitrarily decided on one. The
villages, however, had refused to accept his decision, and it was they who had requested the ordeal, which took place at the
Shinto shrine of a neighboring village. No one from Tsunazawa dared volunteer to hold the red-hot iron, so that it was up to
the headman to step forward after making a short farewell speech, as if he did not expect to survive the ordeal, as the saying
goes. His opponent was a burly fellow who had no qualms about volunteering. Both donned ceremonial dress, received the
kumanogoohoin[*] (a talismanic document from the Kumano shrine used for oaths), with which to grasp the iron, and
approached the fire (see pl. 6). The domain official lifted the red-hot iron out of the fire (with tongs, one presumes) and handed
it over to the contestants. The headman grabbed it three times and then put it aside. The burly representative of

[81] Yamamoto, "Kinsei shoki no ronsho," 107-21. The next several paragraphs of this section are based on ibid.

― 232 ―
Plate 6.
Kumanogoohoin[*] . A talisman dated 1538/10/27 from the Kumano
shrine, used for oaths. The image on the front (top ) represents five Sanskrit letters
formed by the shapes of seventy-five birds, messengers from the Kumano
gongen , centered around a pearl bearing the stamp of the Cattle King (goo[*] ), a
reincarnation of the life-giving (ubusuna ) gods. The oath was written on the
back. From Miyaji Naokazu, Kumano sanzan no shiteki kenkyu[*] , Miyaji Naokazu
ronshu[*] , 3 (Soyosha[*] , 1985). Reprinted with permission from Miyaji Harukuni,
Tokyo.

― 233 ―

Matsuo, on the other hand, dropped it in the fire, having been burned by the talisman, which had instantly burst into flames.
He lost, and what happened next explains the headman volunteer's initial reluctance and his swan song before submitting to
the ordeal: the loser's hands and feet were cut off, and he was buried, his tomb serving to mark the new border between the
two villages. These are the details Yamamoto found in the Aizu gazetteer account written two hundred years after the event, in
the early nineteenth century. The colorful details aside, the lord, his own solution having been rejected, granted the petition
for an ordeal.

The most recent ordeal discovered by Yamamoto dates from 1653 and took place in Echigo (Niigata) to settle a mountain-use
dispute between two groups of villages, numbering four and three villages, respectively. A suit had been filed, but the dispute
continued. A couple of months after the suit was filed, the two village group headmen decided to seek the advice of the gods.
An agreement was drafted on the back of a kumanogoohoin[*] and was signed with the names and blood seals of the ten
representatives who would take the test. The talisman was burned, its ashes mixed with dirt from each border claimed by the
two parties, and the mixture then swallowed by the ten representatives. The agreement was that whoever got sick after seven
days and seven nights would lose the match. This was probably unlikely to happen, hence the Solomonic proviso that if no one
became sick the new border would be midway between those claimed by the two parties. During a tense week of observation
the contestants showed no symptoms. The new border was fixed as agreed upon and ratified by the domain authorities. Yet
thirteen days later the results of the ordeal were reconfigured by the two village group headmen, who declared one side the
loser. The shogunal authorities reversed the original decision, and a new map was drawn up according to the final results of
the ordeal.

Settling otherwise unresolvable village disputes by ordeal was quite common prior to the Tokugawa period. Yet enthusiasm
for decisions by ordeal may have been on the wane. In a village in Omi[*] , for instance, rules drawn up in 1606 and 1607
rewarding volunteers for the red-hot iron ordeal handsomely with a land grant of twenty koku, tax-exempt in perpetuity,
point to the need of incentives to recruit volunteers. Moreover, given the formidable presence of overlord authority after 1600,
we see how gradually this way of local dispute resolution lost ground. Nevertheless, to perceive ordeals as irrational survivals
from

― 234 ―

medieval times that were fundamentally incompatible with Tokugawa practice, would be too simple.
Admittedly, the outcome of an ordeal has more in common with a court verdict—both produce a winner and loser—than with
the prevailing Tokugawa solution of punishing both parties in violent quarrels (kenka ryoseibai[*] ), a form of martial justice,
or with its peacetime equivalent of conciliation, in which both sides usually shared the blame. In this sense, ordeals are of the
same order as medieval court practice. On the other hand, ordeals also signal abdication of worldly authorities in juridical
decisions about right or wrong; they restore order by diffusing tensions through the mechanism of scapegoating. In the above
two examples, the bakufu each time reversed its decision; the merits of the cases did not matter as long as peace was restored
in one way or another, a stance typically taken by the courts throughout the Tokugawa period. At the village level also, a
similar practical logic prevailed when balloting produced a scapegoat.

Loopholes and Legal Maneuvering

Thefts had to be reported to village authorities, but the very codification of this expectation and the penalties for failing to
follow it point to peasant reluctance. Village authorities in turn, had to report thefts to the overlords, as virtually all village
laws required, but often this did not happen. The death penalty was supposed to be a prerogative of overlords, yet historians
continue to discover village codes, and not only from the beginning of the period, that recorded it as a dreaded possibility.

Laws are not always observed, and it is not uncommon to judge the significance of laws by how often they are used. It is
argued that since Tokugawa edicts were issued again and again, they were ineffective and thus insignificant. Other legal
historians who have studied the codes, however, hold the opposite view, assuming that the norms of the codes reflect practice.
Rather than refute these views by argument, it is perhaps best to counter them by looking at actual practice. Thus the question
is not whether villagers circumvented some laws—they did, hence the need to reiterate them—but how. Evidence of such
circumvention will provide some access to the level of villagers' law consciousness. At least it will reveal some points of
resistance to lordly law,

― 235 ―

which is relevant in relation to the question of village "autonomy." We shall proceed again by looking at some cases.

The murauke system of subcontracting the village for purposes of tribute and internal governance created a setting in which
under certain circumstances the village, which was set up to do the overlord's bidding, could instead rally around its own
interests. When villages did so openly, the issue was likely to receive a hearing from the overlords. When something else was
envisioned, however, secrecy was important. Such cases most likely left few documentary trails. The first traceable examples
of such secrecy are from corporate villages that from pre-Tokugawa times had a high degree of independent governance.

In 1591, while the land survey was in progress in Imabori (Omi[*] province), the village members signed a double pledge: (1)
that they would pay the full tribute, and if any household absconded, the five neighboring households would shoulder the
arrears; and (2) that if their tax petition was not accepted, "we the villagers, will flee and we are all of one mind to do so; if
anyone acts against these decisions, we will dissociate from him or her as pledged."[82] Another document, from Nakano
corporate village in the same area, dated 1638, is a combination of a pledge and a village rule, recording a measure taken by
the villagers against a possible reassessment adding new fields to the cadasters. It stipulates that none of the 107 signatories
would reveal hidden paddies, not even to their wives or children.[83]

By Tokugawa times, villages had to officially acknowledge in writing that they had received overlord directives, a requirement
that seems not to have existed in medieval times.[84] How could this acknowledgment and practical consequence be
circumvented, since ukeru ("to receive") implied that ignorance of the law was no valid excuse?

The ninth stipulation of an overlord directive, dated 1658/10/9, to

[82] Tonomura, Community and Commerce , 167, 205. For a similar measure, taken in the context of a tax appeal, see ibid., 203
(no. 468); and Ishio Yoshihisa, Nihon kinseiho[*]no kenkyu[*] (Bokutakusha, 1975), 12.

[83] Tonomura, Community and Commerce , 180; Yokota, "Kinsei sonraku," 144, 155. Henderson reports a case from 1832 in
which six villages in Mino province jointly agreed to protect their grassland against conversion into fields, resisting all such
applications no matter where the proposed reclamation might be located (Henderson, Village "Contracts ," 179-82).

[84] Yokota, "Kinsei sonraku," 155; the example that follows is from 155-57. See also Tonomura, Community and Commerce ,
185-86.

― 236 ―

the same Nakano corporate village ordered the abrogation of village credit unions (tanomoshiko[*] ). The directive was
acknowledged as received by the village kumi's signatures as well as by a separate letter of receipt by the headman, dated
11/10. A corroborating letter of receipt by the kumi, dated 11/16, was addressed to the headman. However, two weeks later, on
12/2, the kumi addressed another document to the headman in which they stated that since compliance with the overlord's
order would cause financial hardship (even for paying tribute), they had decided to continue the operation of the credit union
secretly. If, however, this leaked out and the village leaders were summoned by the lord, the whole village would go to the
lord, explain that it was not their fault, and apologize.

The reason the lord gave for his abolition decree was that there was too much eating and drinking at credit union meetings. In
the document of the kumi, vowing the secret continual, on of the credit union, it was stipulated that henceforward self-
restraint would be exercised in this matter. Thus, they selectively decided to interpret the law's intent by complying with the
lord's> reason and otherwise disregarding the order. It is interesting not only that the kumi resisted the lord's> order but that
the kumi did not think that it would be possible to negotiate openly with the lord over a point that on the surface seems quite
reasonable. The sense must have been strong that one could not force authority to acknowledge mistakes. For this reason, the
shogunal courts never issued a verdict of innocence in criminal cases. So that this outcome would be assured, before the court
hearing proper a "preliminary investigation" with or without torture produced a written admission of guilt from the suspect.
[85]

In 1659/10 the prohibition was reissued, but that was the last time the overlord legislated on the matter; he seems to have
tacitly accepted the credit union's existence after that. Yokota, the scholar who researched this case, based this conclusion on
the fact that in 1677, when

[85] In contemporary Japan, with only one lawyer per 9,300 people, in contrast to one lawyer per 360 people in the United
States, police investigations continue to play a far greater role in the process of convicting suspects than prosecution lawyers
do. Prosecution proceeds only after the police have gathered enough evidence to ensure a conviction, so that the conviction
rate is more than 98 percent (Leslie Helm, "A Long Haul for Japan's Plaintiffs," Los Angeles Times , January 14, 1991, A12;
Teresa Watanabe, "Japan Casts Envious Look at U.S. Crisis Management," ibid., May 7, 1995, A10-A11).

― 237 ―

a dispute between two groups of villages from two different districts broke out concerning credit union funds, one party
considered bringing a suit before the authorities.

Vicarious responsibility was a cornerstone of overlord law. That this caused resentment, jealousies, and complicated human
relations at the village level is not difficult to imagine. It is surprising, however, to find a village code that flatly states that
everyone will be held responsible for his own portion of the tribute and that no help will be forthcoming from the rest of the
village. This is what one historian found in the 1759 edition of the code of Ichijoji[*] village (Yamashiro province).[86] The
relatives, the kumi, and the village were supposed to take over fields of households of orphaned minors, as clearly spelled out
in standard village laws (see appendix 3, art. 40). In Mochizuki (Kita-Saku, Shinano) the intendant got involved in such a case
in the late 1680s.[87] A certain Sotaro[*] died, leaving a son too young to work the fields, but relatives were reluctant to step in.
The village officials, unable to find a solution, queried the intendant, who ordered the obvious: the son should be raised by
some relative, and the fields taken care of collectively by all the relatives. But that was precisely the problem. Upon receiving
the order, the relatives got together to decide whether to comply. Two of the half-dozen relatives refused to affix their seal to
the order and were summoned to the intendant's office for a good scolding. Even so, one of the two recalcitrants held out and
refused to affix his seal. There are no further data to allow one to pinpoint the problem or know what happened to the one
relative who held out. He may have had good reasons for refusing to be part of the solution, and the intendant may have felt
that five relatives out of six were sufficient to take care of the fields. It is interesting, however, to find peasants pondering
whether to follow an intendant's order, some of them categorically refusing to comply.

This sort of resistance is of a different order than the uprisings (ikki) on which American historians have so often focused;
nevertheless, it illustrates the disregard with which peasants could treat lordly authority. In large-scale confrontations,
according to admittedly romanticized

[86] Katada Seiji, "Kinsei sonpo[*] no henshitsu ni tsuite: Yamashiro-kuni Atagogun Ichijoji[*] mura no baai,"
Chihoshi[*]kenkyu[*] , no. 34 (1958): 33.

[87] Ozaki Yukiya, "Shinshu[*] no nomin[*] to buraku," Mochizuki no burakushi , no. 4 (1978): 41-42.

― 238 ―

and heroicized accounts, the peasants spoke their mind quite bluntly.[88] Such uprisings were condemned by the rulers as
illegal conspiracies (toto[*] ). When, during the ikki of 1754 in Kurume domain, the daimyo's officials tried to tell the peasants
that the prohibition of "conspiracies" was the foremost law of the realm (tenka dai ichi no gohatto ), the peasants who heard
this, the record tells us, "burst out laughing, asked why the officials appeared today, beaming with authority and full of
contempt for the people, reading to us orders high up from horseback, to us who are waiting here holding authority in high
esteem and observing the laws. We know that the prohibition of conspiracies is the foremost law of the land. But the lord
indulges in debauchery...."[89]

Banishment was a punishment that sometimes needed to be concealed. Banishment itself was not completely forbidden, but
its arbitrary application was (for a village law, see appendix 3, art. 26). However, since banishment resulted in the
disappearance of a name from the population register, the effect of its application was hard to hide. From headmen's notes or
diaries we learn how this could be done.[90]

In 1757 a village guard caught a woman stealing rice from a field. The village decided to follow precedent and sent her into
exile. According to the principle of vicarious responsibility (enza), her husband deserved the same punishment, but since he
was absent (being in service elsewhere) he was forgiven. The house, however, was put up as collateral for the yearly tribute,
and its effects were given to the guard. The woman's banishment was handled as follows. She was kept on the population
rosters until she had settled somewhere else. Then she was handed a certificate of leave, and her brothers had to sign a sworn
agreement never to allow her in the village again. Thus, on record it appeared as if she had moved. Obviously, such tricks
worked only as

[88] Herbert Bix (Peasant Protest ) cites sarcastic statements peasants allegedly addressed to samurai, such as the following:
"Since they [inspection officials] are so skilled in local affairs, we wish to learn from them how to grow rice" (79); "Since you
officials think that 'peasants' are particularly useful beings, we wanted m let the samurai do the hard work of peasants and see
for themselves how profitable we are" (94); "We don't need your kind to look after US" (179).

[89] Nanba, "Hyakusho[*] ikki no hoishiki[*] ," 76. This is an important article on the people's consciousness of the law related
to the ikki , which is beyond the scope of the present study.

[90] The following examples are from Mizumoto, "Kogi[*] no saiban," 306-8.
 ― 239 ―

long as nobody in the village reported them—which would have been done at the risk of all sorts of informal reprisals.

A diary from 1826 reveals the following strategy to hide a banishment by making it appear as if the person in question had
absconded. A culprit had been caught by guards from a different village, and his village council had decided to kick him out
but instead reported him as missing. As was usual in such a case, the domain ordered a search for the fugitive, which the
village feigned to conduct, and then reported the culprit as still missing. After a second search (usually lasting thirty days) was
conducted, he was again declared missing and was then officially declared nagatazune (literally, "long search"), that is,
homeless (mushuku, or nonregistered). As it turned out, the man returned to the village five years later gravely ill, died soon
thereafter, and was quietly buried at the local temple.

As we saw in Ken's case in chapter 1, local officials could collude with district magistrates or intendants, who tried as far as
possible to avoid becoming involved in dispensing justice. If "private consultation" (naibun nite ) among themselves led to the
conclusion that "village justice" might reestablish order, then both parties would benefit from adopting it even if the solution
might bend official policy. Why would overlords so readily relinquish their authority? What was the logic of such practice?

Law as Tatemae

Mizumoto has found a document from 1751 that may shed light on these questions.[91] It is a memo (kojo[*]no oboe ) from a
village group headman to an intendant's assistant from Takada domain (Echigo province, Niigata prefecture). It seems to be an
exchange of opinion between two local intermediary officials from the village and the district intendant office.

On 1751/7/17 a peasant had been seen returning from the fields with stolen rice. An investigation had followed, and a
confession that included four other thefts as well had made the situation serious enough to report to the village group
headman. Thus it had become a matter for the domain's court (gosaho[*] ). The village group headman, however, had his own
ideas about how to proceed. Thinking it might be better to get

[91] Ibid., 309-10.

― 240 ―

rid of this habitual troublemaker altogether, he wanted to change the village headman's report into a request for banishment.
Hence, the village group headman's request for the private opinion (gonai-i ) of the intendant's assistant on this course of
action.

The document implies that although in principle only the court should deal with thefts, the village group headman suggested a
solution that would spare the court the trouble of dealing with the particulars of the case, leaving the actual legal process up to
the village. Mizumoto finds further support for this interpretation by quoting Hiramatsu Yoshiro[*] , the great authority on the
Tokugawa penal system:

Private and public law were mixed together in the following sense. The principle governing punishment and
pleading by defendants in civil lawsuits, which not only aims at solving the disputes between litigants but at the
same time is satisfied with establishing "shogunal authority" (goiko[*] ), permeated not only criminal trials but also
the execution of the penalty. And this is sometimes the case also for minor misdemeanors and especially for
criminal cases involving private interests.[92]

Officially (tatemae), the shogun and daimyo proclaimed a monopoly on juridical matters, but their primary concern certainly
was not social justice nor even the maintenance of overall public order as such. They were preoccupied foremost with
safeguarding their authority. Therefore, they silently tolerated "private justice," as long as it was not publicized and did not
openly challenge that authority. This two-sided stance affected the status of village codes and village justice.

The Lawless Village

According to John Haley, the Tokugawa village was "an institutional structure that in allowing evasion of official legal controls
also promoted external deference and internal cohesion"; it had "the security of the administrative state along with the
freedom of the outlaw,"[93] The application of the outlaw metaphor to Tokugawa villages may at first

[92] The quote is from Hiramatsu, Kinsei keiji soshoho[*] , 837.

[93] Haley, Authority without Power , 61. Haley further develops the thesis that the Tokugawa village pattern of justice and the
peculiar relationship between communities, on the one hand, and the authorities and the law, on the other, discussed as "law
as tatemae ," still govern life in modern Japan (187-190). Mizumoto also discusses the tatemae aspect of Tokugawa law in his
"Kogi[*] no saiban," 308-12.

― 241 ―

seem problematic. Haley uses it only to convey a strong sense of the weakness of centrally controlled mechanisms of law
enforcement in the village. The metaphor is appropriate in another way as well, however. Although all villages had laws and
codes, village life as it is revealed by the various cases studied often appears to have taken place in an outlaw world, or at least
in a world in which the world of Kurosawa's Yojimbo[*] was a possibility.

The general propensity on the part of both extra- and intravillage authorities to keep things within the village, and the
structural setup that made this possible, is perhaps the main reason for the impression of the "lawless village." This mutual
avoidance had two results. On the one hand, the semblance (tatemae) of internal harmony had to be maintained so that
overlords would feel content to leave villages alone. On the other hand, this left the exercise of power unchecked. The
gomyo[*] in Hozu and similar groups elsewhere had a free hand in securing their lasting domination by putting themselves
"outside the law" that they wrote for fellow villagers and imposed on them.

The main concern of local wielders of power was that conflicts might develop to the point of alarming the overlords. As a
preventive gesture to keep the overlords out, the local power holders often deferentially submitted to the lords the codes they
had written to demonstrate that they were "within the law" of the lord. In a prophylactic move to keep the villagers contained,
the village authorities discouraged them from initiating lawsuits by painting those who did as troublemakers in village
regulations (compare the negative tone regarding lawsuits in appendix 3, arts. 16 and 54, with the earlier, neutral stance in
appendix 2, art. 19). Obviously, individuals seeking access to overlord authority did not use it only as the only means for
contesting local powers through suits and petitions. Local authorities invoked the image of overlord intervention as a last
resort, "borrowing" extramural power to avoid actually activating it, to keep (themselves) on top of things. Aside from actual
collusion (in Ken's case), the local authorities sometimes threatened lawsuits or presented the specter of torture in the lord's
court.

If one keeps in mind that villages were informal power fields where the competition was often between lineages, one may
perhaps start to perceive in a new light some of the practices discussed in this chapter—ostracism, crime voting, and the
measures taken against collusion in casting ballots. As Chojiro's[*] lawsuit illustrates, ostracism was not always initiated by a
"unanimous village," but often by intravillage

― 242 ―

factions, to the point that the bakufu limited its use in the nineteenth century. In villagers' minds, solutions to unresolved
crimes often tended to be structured along lines of splits in the community. Crime voting may have been a solution to prevent
such situations from being used for settling old scores. The posting of guards throughout the village to prevent "consultation
balloting" suggests that such a scenario is not far-fetched. Similarly, unreported crimes might lead to retaliations or vendettas
against one's enemies as the putative perpetrators; hence the pressure to make the nonreporting of crimes a crime.

From this perspective, villages no longer resemble autonomous, harmonious, egalitarian communities. Rather, like social
formations anywhere, they constitute spaces where groups of people, households, or lineages vie for positions hierarchized by
class, status, and law using the weapons of class, status, and law.

― 243 ―

5
Status and State Racism From Kawata to Eta
The animality of man, in man and against man (whence the systematic "bestialization" of racialized individuals and
human groups) is the particular method that theoretical racism adopts in thinking about human historicity.
Etienne Balibar

In the preceding chapters, an argument is made for the pervasive functional importance of social and political status in village
life throughout the Tokugawa period. High social status, buttressed by bakufu measures as well as village codes, was the single
most significant source of local authority and power. Moreover, with the development of a commercial economy in the
eighteenth century, political status was not displaced by economic status, even though titles became available for purchase on
a restricted village market. Instead, officialized status (originally in part a by-product of the overlords' assignment for tribute
responsibilities to local magnates) took on an independent, symbolic value, having the potential to politically offset the loss of
material assets. Status, however, retained important links to a material base. Economically successful peasants frequently
tried, sometimes effectively, to use recently acquired wealth to secure access to, or even displace, the old village "status
aristocracy." Ultimately, economic status, whether past or present, was always at issue in claims of political and social status.

The existence of a "nonstatus" category, however, forced certain people to exist apart from and outside of the status system
described above: the outcastes, as we have referred to them thus far, or the so-called eta . These people did not refer to
themselves by that derogatory term meaning "plentiful dirt," or polluted. Instead, they called themselves kawata , literally,
"leather workers," after their principal occupa-

Epigraph: Etienne Balibar, "Paradoxes of Universality," in Anatomy of Racism , ed. David Theo Goldberg (Minneapolis:
University of Minnesota Press, 1990), 283.

― 244 ―

tional profession, a practice I shall follow here. The history of discrimination against the kawata as a function of status in
Tokugawa society directs attention to the early modern relationship between sociopolitical status, legislative practice, and a
form of racism—a term that may cause surprise but is not used lightly. Like other social groups, for example, the samurai,
merchants, and peasants, the kawata were legally restricted to their birth-ascribed status. Registered beggars, or hinin
("nonhumans"), also constituted a "nonstatus" group, but one that was not hereditary: a person could become a hinin, be
sentenced to that status, or be allowed to leave it, but once a kawata, one died a kawata.

In his work on systems of social stratification, the anthropologist Gerald Berreman argues that all status systems assign people
a "differential intrinsic worth" according to their position in the social hierarchy.[1] In the case of the kawata this intrinsic
worth was negative to the extreme, for they were marked as existentially and unredeemably polluted, defiled, impure. In
Japan, where notions of purity and pollution have historically functioned as powerful categories of ritual classification, the
kawata suffered the worst consequences of the social application of these values. Even though the status system, with specific
regard to the kawata, was legally abolished in 1871, their descendants are the only persons in Japanese society today who
continue to suffer discrimination because of their ancestors' status during the Tokugawa period.

Today the kawata's descendants are referred to officially as burakumin , or "hamlet people," after dropping the Meiji era's
qualifying prefix tokubetsu , "special." Estimated as numbering as many as 3 million people, the burakumin are Japan's largest
"Japanese minority."[2] There are still some six thousand segregated communities throughout Japan, where most of them live.
Many burakumin try to "pass" into Japanese society; some succeed, to the point that their descendants are

[1] Berreman, "Social Inequality," 14.

[2] I have adopted the (perhaps) high figures provided by the Buraku Liberation Movement. For a discussion of these figures,
see Nihon shomin seikatsu shiryo[*]shusei[*] , vol. 14, Buraku , ed. Harada Tomohiko (San'ichi shobo[*] , 1971), 1 (hereafter NSSS
14). See also George De Vos and Hiroshi Wagatsuma, Japan's Invisible Race: Caste in Culture and Personality (Berkeley:
University of California Press, 1966), 114-18; and Ninomiya Shigeaki, "An Inquiry concerning the Origin, Development, and
Present Situation of the Eta in Relation to the History of Social Classes in Japan," Transactions of the Asiatic Society of Japan . 2d
ser., 10 (1933): 113.

― 245 ―

not aware of their own burakumin background. The 1871 Emancipation Decree did not make private or even public
discrimination illegal (although article 14 of the postwar constitution prohibits discrimination by the state). The Emancipation
Decree simply repealed the legally sanctioned status distinctions in the realm of occupations and social status: "The
appellation eta/hinin shall be abolished; henceforward they shall be treated as common people in social status and
occupation." Frank Upham writes that legal emancipation "meant little. The government was at best indifferent to significant
change.... In fact, government policy directly contributed to continued discrimination by registering Burakumin as 'new
commoners.'"[3] Their situation is not dissimilar to that of blacks after the abolition of slavery in the United States, which did
not eliminate racial inequality and discrimination.

In attempting to account for this discrimination against the kawata/burakumin, we need to address the role of Japanese
cultural beliefs about purity and pollution. Undoubtedly, pollution was the idiom of discrimination that turned kawata into
eta. Moreover, insofar as they constituted a status group, it may be that they not only shared assigned universal characteristics
of status differentiation but revealed in its purest form the function of status as socially sanctioned disrespect for a particular
group of people. How the ascriptive logic of status systems works is thus a second question that needs to be considered. Closely
related to this question is a third one, that of racism. There are several reasons why this term can be considered appropriate in
a discussion of the kawata/burakumin.

Berreman makes a good pedagogical case for using analogical terms—caste and race —that by definition are alike only in
certain (but sufficient) respects.[4] To make the notion of caste and its effect on people's lives understandable to Americans,
Berreman encourages speaking of a caste system as a form of racism, because Americans are familiar with that reality. (He
would encourage the reverse in attempting to explain racism to Hindus.) He writes that burakumin "comprise a 'race' in the

[3] Frank K. Upham, Law and Social Change in Postwar Japan (Cambridge: Harvard University Press, 1987), 80. For the text of
this brief decree, see Burakushi yogo[*]jiten , ed. Kobayashi Shigeru et al. (Kashiwa shobo, 1985), S.V. "kaihorei[*] " (56).

[4] Gerald Berreman, Caste and Other Inequities: Essays on Inequality (Meerut, India: Folklore Institute, 1979), 190.

― 246 ―

sociological sense of Western racism, but an 'invisible' (i.e., not genetic or phenotypic) one."[5]

The kawata/burakumin are more than a sociologically "invisible" race. In the course of his research on caste and race,
Berreman found that "all systems of birth-ascribed stratification seem to include a claim that the social distinctions are
reflected in biological (i.e. "racial") differences."[6] This observation, while not true for all status groups in Tokugawa Japan,
certainly applies to the kawata. Although the kawata are indistinguishable from other Japanese, explicitly racial theories,
which differentiate , and racist theories, which discriminate , have developed since the Tokugawa period to explain the origin
of the kawata in order to justify discriminatory practices against them.

The absence of physical properties to distinguish them from "majority" Japanese (hence "invisible" differences) and the
construction of a racial theory (hence a different "race") have combined to produce a particular kind of intra-race racism, one
based exclusively on descent. Genealogies, however, are known only to those who literally know the record. Without that
knowledge, discrimination against burakumin would be impossible. This is where the state plays a crucial role. As mentioned
above, the 1871 Emancipation Decree did not outlaw discrimination per se. Indeed, all former kawata were entered on the
population registers as "new citizens" (shinshimin ), so that state records perpetuate the kawata's separate identity, making it
possible today for anyone sufficiently interested to consult the old population registers to discriminate against burakumin.[7]
Without active complicity by majority Japanese, discrimination would disappear; some, however, still care enough about
racialized status to check these records.

Revulsion and defilement, the cultural idiom and social disposition through which discrimination against the kawata is
commonly expressed, are often correlated to the high valuation of purity in the Shinto tradi-

[5] Ibid., 188. This invisibility can have peculiar effects; thus, unaware of their own descent, some Japanese may share
prejudices against burakumin until the day when their own burakumin background is revealed and they are shocked to
discover that they have been discriminating against their own (Sam Jameson, "Japan's 'Untouchables' Suffer Invisible Stain,"
Los Angeles Times , January 2, 1993, A24).

[6] Berreman, "Social Inequality," 14.

[7] The ineffectiveness of recent restrictions of access to these records is discussed below.

― 247 ―

tion. Yet, the first laws to identify categories of Japanese associated with specific occupations as senmin , literally, "despised,
mean (iyashii ) people," were laws adopted in toto from China in 645, during the Taika Reform. Scholars often trace the
development of discriminatory practice in Japan to that point.[8] However, the composition of these "despised" social groups,
their position in society, and the nature of prejudice against them has changed considerably over time. Legal discrimination
decreased greatly during the early Heian period (ninth century) and became insignificant in medieval times. For example, a
law of 789 declared children from mixed marriages between "good" (ryomin[*] ) and "low" people (senmin) to be ryomin[*] ,
whereas previously they had been marked as senmin. Moreover, whole subclasses of senmin were simply set free when
government offices to which they had been attached as dependents were dismantled and slavery was abolished in 907.[9]
Although many nonagricultural occupations were socially devalued in medieval times, the first reference to some of them as
"plentifully polluted" (eta written with the characters still used today) appears only in the 1450s.[10]

It is often said that the Buddhist reverence for life, as well as Shinto notions of pollution associated with death and blood,
played an important role in targeting butchers and tanners as particularly defiled, and indeed a few late texts confirm this.
Thus one finds a clear reference in a Buddhist monk's diary of 1446 to "those who carve up dead cattle and horses for
consumption" as "the lowest kind of people,"[11] although the source of abomination seems to be more the eating of meat than
the butchering. Nowhere does one find reference to hereditary pollution dissociated from occupational succession.

Medieval concepts of despised occupations and human defilement were social and occupational (hence functional in a sense),
loosely defined and applied, and formulated neither by law nor as a hereditary condition. In 1871 a law had to be enacted to
abolish the official status "eta-hinin," which points to a legal dimension for discrimination during the Tokugawa period that
was missing in medieval times. In Toku-

[8] For one such example in English, see Ninomiya, "Inquiry."

[9] Teraki Nobuaki, Kinsei buraku no seiritsu to tenkai (Osaka: Kaiho[*] shuppansha, 1986), 22.

[10] Ibid., 26.

[11] Watanabe Hiroshi, Mikaiho[*]buraku no keisei to tenkai (Yoshikawa kobunkan[*] , 1977), 70.

― 248 ―

gawa Japan lie the origins of the state's role in the ongoing discrimination against the kawata/burakumin. Some scholars
estimate that more than one-third of today's approximately six thousand buraku communities originated in the eighteenth
century,[12] when we begin to find discriminatory and segregationist laws, racist in their effect, being enacted throughout
Japan against the kawata.

Today one can see how cultural notions of pollution and racism conspire with bureaucratic practices to perpetuate
discrimination against burakumin. But how can the introduction of the new status for kawata in the Tokugawa period, and the
new laws, practices, and ideologies that developed along with it, be explained? What role did bakufu legislation play in this?

I shall start by discussing a number of episodic events, legal cases, and incidents involving kawata. The need for such an
approach stems from the paucity of microhistorical narratives of the burakumin's Tokugawa past in English. Emiko Ohnuki-
Tierney's symbolic functional analysis posits the kawata ahistorically, as a necessary prop for Japanese culture to construct a
counterimage of itself.[13] Historical and cultural distance easily robs past agents of their subjectivity even for well-intentioned
scholars, who nevertheless tend to assume that ideological representations were internalized by those who suffered the most
from them. Thus Ian Neary, in his book about the modern liberation struggle of burakumin (to which he is obviously
sympathetic), writes quite confidently: "The eta:hinin regarded themselves as impure, not quite human, not deserving equal
treatment."[14] He conflates what Bourdieu calls an objective, socially assigned position and a subjective relation to that
position on the part of those who occupy it.[15] The kawata did not

[12] Teraki, Kinsei buraku , 111.

[13] Emiko Ohnuki-Tierney, The Monkey as Mirror: Symbolic Transformations in Japanese History and Ritual (Princeton:
Princeton University Press, 1987).

[14] This sentence is preceded by the following imperious passage: "All that existed within the eta:hinin groups was simply a
deviant reflection of the majority culture of the locality. With neither leadership nor culture there was no sense of identity.
Indeed the development of a set of religious beliefs which supported the regulations systematically imposed from the
eighteenth century, effectively prevented the emergence of any kind of self-esteem which would provide a basis from which to
challenge superior bodies" (Ian Neary, Political Protest and Social Control in Prewar Japan: The Origins of Buraku Liberation
[Manchester: Manchester University Press, 1989], 26).

[15] Bourdieu, Logic of Practice , 27.

― 249 ―
even call themselves eta. When, for instance, a bakufu intendant in Harima province inquired at a certain village whether
there were any eta, the answer was that there were some kawata, yes, but no eta.[16] To my knowledge, the political situation
of the kawata during the Tokugawa period has mostly been ignored by Western scholars.[17] The events to be discussed will
expand our understanding of the position, treatment, and self-perception of the kawata first through an analysis of social and
political practice. They will supplement the macrocultural or microindividual and ideological explanations that are often
invoked in discussions of the kawata/burakumin.

Episodes

"In Service of the Shogun," 1743

The date is 1743/11, the place Kami-Hosoya village in Yokomi-gun of Musashino province (Saitama prefecture).[18] A few weeks
before, a local criminal had been sent to Edo to be executed. Now his head was being returned for a three-day public exposure.
To manage the event, the headman called upon the chief of the kawata hamlet attached to Shimo-Wana village.

In many ways the kawata were not under the rule of the regular village headman. Hence, instead of contacting his colleague in
Shimo-Wana, Kami-Hosoya's headman communicated directly with the head (etagashira ) of the kawata/hinin community
associated with Shimo-Wana village. Moreover, the headman had no authority over the head

[16] Wakita Osamu, Kawaramakimono no sekai (Tokyo Daigaku shuppankai, 1991), 5. Hatanaka Toshiyuki reports that in Kinai
the eta called themselves kawata ("'Kawata' mibun to wa nani ka," MK, 307). In Hozu village, discussed in chapter 4, the
kawata, referred to officially as ego[*] , literally, "polluted district (the e of eta plus go[*] )," referred to themselves with a
homonym, substituting the e of Edo ("inlet," "bay"), perhaps because their community was the "capital" of the five kawata
communities in south Tanba (see Igeta, "Mikaiho[*] buraku," 110; and idem, "Kuchi-Tanba," 95, 111).

[17] One notable exception is Frank Upham, who shows an acute awareness of the various political aspects concerning the
kawata as they have been dealt with by Japanese scholars (Law and Social Change , 79-80). Most of the studies of the
burakumin deal with the modern liberation movements. De Vos and Wagatsuma (Japan's Invisible Race ) provide no more than
a general background for the Tokugawa period.

[18] Tsukada Takashi, "Kinsei no keibatsu," 119-23.

― 250 ―

of the kawata and could not give him direct orders; rather the Kami-Hosoya headman invited him to take charge of the
situation.

Why Shimo-Wana and not some other kawata community? The headman had no choice in the matter. Kami-Hosoya was part
of a specific geographical area that included a number of kawata hamlets operating under their own leadership in Shimo-
Wana. These territorial divisions went by various names, the most common being kusaba or dannaba . The kawata performed
prescribed duties for the area: skinning and disposing of dead cattle and horses and reworking the hides into leather goods
(footgear, drum skins, armor, bow strings, etc.), catching fugitives and lawbreakers, guarding prisoners and executing
criminals, patrolling the villages, and policing festivals and markets. Such duties or rights were divided among the
kawata/hinin as shares (dannaba kabu ). The Shimo-Wana dannaba comprised twenty-five villages and was one of five such
dannaba in Yokomi district.[19]

After the job was done, the two kawata heads in charge handed the implements used (gokogisama[*]godogu[*] , "the honorable
tools of the honorable public authority [the shogun]") back to the headman, who drafted a receipt properly addressed to
"Mister (dono ) Jin'emon and Mister Genzaemon from Shimo-Wana village." During their three days on duty, all twelve kawata
and two hinin were served meals in a room at the headman's, as were a number of kawata heads from other hamlets who had
come to take in the scene.

This was a rare occasion for the kawata/hinin to function with pride and authority at an extraordinary event as full-fledged
officials. Four of the higher-ranked kawata wore long swords, the eight others short ones. Swords, a professional attitude,
polite addresses, and being entertained at the home of the village headman, apparently without much concern about
"pollution," were all marks of public respect. Their authority was manifested in other ways as well.

At noon on 11/15 an intendant's representative arrived to prepare for the display of the head he had brought from Edo. That
evening he conferred with Jin'emon and two other kawata leaders about, among

[19] Tsukada Takashi, Kinsei Nihon mibunsei no kenkyu[*] (Kobe[*] : Hyogo[*] buraku mondai kenkyujo[*] , 1987), 18-20. In the
Kanto area, shares were time allotments of certain days to certain households; in the Kinai, the shares were territorial
segments that were given to households for their operations (Igeta, "Kuchi-Tanba," 102).

― 251 ―

other things, how much to charge onlookers for viewing the spectacle, since it was "the custom in the countryside to charge
eight mon for a crucifixion and nine for a beheading." According to the kawata's reconstruction of the event, Jin'emon replied,
"We do not take money as executioners; besides, if you charge fees in the countryside, people will not come near but crowd the
roads and look from afar and this would be of no service to the shogun." As it turned out, a fee was charged, and forty-six
people came from eleven villages and had a jolly, drunken time.

In Edo at least, executions were not grand public spectacles, but took place in the prison courtyard. Perhaps the authorities
were aware of the dysfunctional crowd behavior executions provoked, as described by Michel Foucault in Discipline and
Punish . However, other aspects of penal practice, such as parading the criminal or exposing the corpse or head, were meant to
have a salutary effect. The kawata understood this deterrent effect, for they wanted people to "come near" and see. More
people would show up if no fee were charged; they protested that if executions were staged and money paid to view them,
dignity and honor would be lost—the shogun's as well as their own.

In their meeting with the intendant's emissary the kawata leaders also argued about protocol. The official announced that it
was a "national law" that ten guards be used on an occasion like this. The kawata heads countered that in Edo twelve kawata
and a great number of hinin were used for an ambulatory exposure, but if there were to be no ambulation, then two hinin in
addition to twelve kawata would suffice; in other words, a total of fourteen guards, and not ten, were needed. Moreover, they
said, they had no orders from Danzaemon in Edo to lower the numbers. The official nevertheless ordered them to limit the
number of guards to ten. The kawata leaders agreed, with the proviso that they would check with Danzaemon on what to do
on such occasions in the future.

Satisfied that all preparations had been made, the official left the next day at noon, having handed over the display item. The
three-day event began the following morning with fourteen rather than ten guards on active duty. The official did not get the
last word after all. Clearly, the kawata had followed their own professional judgment and the procedures as proscribed by
their (semi-)national leader Danzaemon, who received a written report from them a week later.

Almost a hundred years after this event, a crucifixion took place in

― 252 ―

the same village. Again local kawata were mobilized to stand guard for the three days the corpse was to be exposed. The
crucifixion proper, however, was performed by a team of two of Danzaemon's underlings, whom he had dispatched with
fifteen helpers from Edo.

Shimo-Wana's dannaba was located within the territory of one bakufu intendant. This was not always the case, for in practice,
if not in principle, dannaba often cut across a number of jurisdictions and even domains. Therefore, the kawata's social and
political topography was different from that of the peasants. They lived in their own hamlets under a separate local and
regional jurisdiction in fairly close contact with their leader Danzaemon in Edo. They were often required to, or simply did,
cross into social and geographical areas generally forbidden to them while performing their duties. This was especially true
during the late Tokugawa period, as the following example illustrates.

Catching Thieves, Arresting Vagrants, 1848

In 1848/3 three prisoners escaped from the bakufu intendant's rural office (jinya ) in Mikage-shinden (in the center of the Kita-
Saku plain, Shinano).[20] They headed south but were caught some fourteen kilometers away, at the village of Kutsusawa, just
as they were about to disappear into the mountains. They were escorted back to prison by a team of kawata and a lower
official from the Mikage-shinden office. The official had difficulties enforcing his authority over the kawata.

First of all, the kawata disobeyed his orders concerning the size of the escort. Because many men from different hamlets had
joined the party, their number was well beyond that prescribed for such an occasion. Moreover, when the noisy troupe
arrived at the Nakasendo[*] way station of Shionada at midday, they protested the official's decision to feed them the same
riceballs (nigirimeshi ) the prisoners were being fed: while performing official business, they said, they deserved a full meal
served on a platter. The group became angry, forced their way into two inns, and proceeded to eat and drink their fill.

Afterwards an investigation was launched by the kawata head of Mikage-shinden. Twelve kawata were summoned, but none
of them had been part of the escort from Kutsusawa; they had joined later. All in all, seventeen kawata from nine villages
(representing a bakufu fief

[20] Saito[*] Yoichi[*] , Gorobe-shinden , 251-60.

― 253 ―

and three domains—Tanoguchi, Iwamurata, and Komoro—within a radius of some fifteen kilometers) were said to have
feasted at the inns. Ultimately, the headmen from the parent villages of the hamlets of the kawata in question wrote a joint
request for pardon. One kawata from Gorobe-shinden was finally jailed.

Strictly speaking, the Mikage-shinden kawata head's jurisdiction was limited to kawata in that bakufu territory. However, in
an area like Kita-Saku, where the fertile plateau had been carved up by half a dozen overlords, he often found it necessary to
coordinate activities, such as the catching of thieves, in regions beyond his jurisdiction. The foregoing investigation concerned
two issues: a false initial report of the identity of all participants in the event; and the forced entry into the inns under the
rubric of official business. Almost everywhere in eighteenth-century Japan laws were issued forbidding kawata from entering
the homes of ordinary commoners "even if it rains." In Kita-Saku such laws suddenly appeared in the domains of Komoro,
Iwamurata, Okudono, and Tanoguchi, all in 1738.[21] The second issue thus involved the breaking of laws ordering segregation
between kawata and commoners.

In 1855 four vagrants (mushuku) wearing long swords were reported at an inn at Kodai. Dispatched from Mikage-shinden to
apprehend them were two kawata armed with sticks—which was unusual, since they customarily wore one sword when on
official business, although some regulations seem to have prohibited swords for kawata (see art. 3 of the regulations for
outcastes in Okudono domain, Kita-Saku, in appendix 5). The vagrants resisted arrest, assaulted the two kawata, and escaped,
leaving each kawata with deep cuts (some eighteen centimeters long) across the face and forearm. When peasant uprisings
escalated to armed confrontations, kawata were mobilized along with samurai retainers to quell the peasants. Such incidents
involving kawata police have been documented for Fukuyama domain (1787), Shinano's districts of Aida (1869) and Kita-Saku
(1782), and a number of other places.[22]
[21] Ozaki Yukiya, Shinshu[*]hisabetsu buraku no shiteki kenkyu[*] (Kashiwa shobo[*] , 1982), 17.

[22] On Fukuyama and Aida, see Bix, Peasant Protest , 124 and 202; on Kita-Saku, Banba Masatomo, "Mibunsei no teppai e:
Shinshu[*] ni okeru burakushi sobyo[*] (4)," Shinano 16, no. 12 (1964): 36. For another example from 1755, see Teraki, Kinsei
buraku , 63. John B. Cornell mentions kawata being called upon in 1871 to defend the castle of Takasaki (Gunma prefecture)
and the mobilization of more than five hundred kawata to garrison a fortress during an uprising of some seventy thousand
peasants in northern Kii ("From Caste Patron to Entrepreneur and Political Ideologue: Transformation in Nineteenth and
Twentieth Century Outcaste Leadership Elites," in Modern Japanese Leadership: Transition and Change , ed. Bernard Silberman
and H. Harootunian [Tucson: University of Arizona Press, 1966], 63).

― 254 ―

When kawata were used against rioting peasants, they were not looked upon as officials to whom respect was due. For
example, when three kawata hamlets were mobilized in the Fukuyama riot of 1787, the peasants apparently voiced their
anger and indignation by shouting things like "If you dare insult us by turning the likes of abominable eta on us, we'll kill them
and feed them to the dogs!"[23]

At other times kawata joined in rebellions, apparently without objections from the other peasants.[24] Such was the case with
Oshio[*] Heihachiro's[*] uprising in Osaka in 1836. It sometimes happened that kawata communities were split between those
who were mobilized to suppress the unrest and those who actively participated in it. In 1823 four kawata were among the
twenty-six people punished as ringleaders of a large rebellion in northern Kii.[25]

When peasants rioted, kawata were mobilized to suppress them, but when kawata rioted, a rare occurrence, peasants, not
other kawata, were mobilized to bring them under control. The authorities deployed commoners against commoners to keep
the peace, playing off the emotionally loaded status antagonisms between them. The next incident involves villagers who
mobilized such hostile sentiments when they discredited a fellow peasant by defaming his pedigree as polluted by low-status
elements.

[23] Teraki, Kinsei buraku , 64.

[24] Hatanaka ("'Kawata' mibun," 327) lists seven riots between 1748 and 1866 in which kawata participated.

[25] Teraki, Kinsei buraku , 64. Some historians have raised the question whether kawata were actually ringleaders in all
instances where they were punished as such. They may have been the victims of lower officials' eagerness to demonstrate
results of their mopping-up operations (see Kobayashi Shigeru's remarks in Kobayashi Shigeru et al., "Zadankai: Kinsei
hisabetsu buraku ni kansuru horei[*] o megutte," in Kinsei hisabetsu buraku kankei horeishu[*] : Tenryo[*]o chushin[*]to shite ,
ed. Kobayashi Shigeru [Akashi shoten, 1981], 498). In the Chihara riot of 1782, described and analyzed by Anne Walthall (Social
Protest , 21-23, 151-55, esp. 154), the kawata community of Minami-Oji[*] was singled out for special punishment (see
Hatanaka, "'Kawata' mibun," 321-24).

― 255 ―

The Bonboku Incident, 1777-1780

In Kita-Saku in 1777/3 a certain Yojiro[*] was excluded from a young men's dance at the village festival in Ozawa[*] .[26] The
village officials had concurred with this expulsion, because Yojiro[*] was someone "with a bad pedigree." This claim did not go
unchallenged by Yojiro[*] and his father Chuemon[*] , and led to a lawsuit two years later. At issue was whether the officials
had slurred Chuemon's[*] ancestry, and if so, on what grounds.

The intendant in charge tried to take an easy way out. Avoiding the central issue, he cited Chuemon's[*] violent behavior
during the present quarrel and sentenced him to be handcuffed for a time. This triggered a number of personal appeals, not to
the intendant, but directly to the domain's investigators in Edo. In 1779/9 one of Chuemon's[*] relatives went to Edo.
Chuemon[*] himself was joined in an appeal to Edo by two of his sons in the third and sixth months of 1780. Domain
authorities were thus forced to address the substance of the dispute and come forth with a decision regarding Chuemon's[*]
status, which is why we know the arguments on both sides.

Chuemon's[*] rebuttal consisted of several points. First, one of his ancestors had been entered as a titled peasant on the land
survey of 1629, one hundred and fifty years earlier. Second, another ancestor had served as a local guide to land surveyors in
the 1650s and 1660s. And third, he had commended land to the village shrine. This, in his eyes, entitled him to treatment as a
regular community member.

Of these three points, the village officials accepted as valid only the first one, even though they acknowledged Chuemon[*] as a
tribute-paying peasant. Yet ancestry and tax status notwithstanding, the village officials believed that their discriminatory
stance was justified. Their reasons were also threefold: Chuemon's[*] family had not married local people; he acted as if he
were on equal terms with the blind; and he did not relate to kawata the way the other peasants did. Through circumstantial
evidence and innuendo it was argued that the social behavior of this otherwise legally certified peasant justified
discriminatory treatment. The absence of local marriage alliances suggested that Chuemon[*] had family ties with "others."
The blind associates referred to here were

[26] Ozaki, Shinshu[*]hisabetsu buraku , 101-2, 259-60.

― 256 ―
zato[*] , not just people with impaired vision (mojin[*] ) but members of an official organization for storytellers, musicians
(biwahoshi[*] ), masseurs, and so on, all of whom were blind individuals who earned an independent living in their
occupations.[27] As such, zato[*] were a subcategory of senmin, lower people, just as the kawata were. Thus the statement that
Chuemon[*] was on a level with them suggested that he too was of a lower status.

The same suggestion was made with regard to Chuemon's[*] association with the kawata. He was even accused of not paying
the "sheaf" (ichiwa ), which every peasant household was required to pay the kawata for the services they performed for the
village. Moreover, it was claimed that one of Chuemon's[*] ancestors, a certain Saemon, had made his living as a bonboku , a
local name for entertainers who performed the lion dance (shishimai ).

Entertainers were yet another category of senmin; they were eventually put under the jurisdiction of the kawata chief
Danzaemon in Edo during the eighteenth century. Like other holders of low occupations, they were usually housed in separate
quarters. For example, in the land register of Azumi district in Shinano circa 1650, yamabushi yashiki (compounds for
mountain ascetics), doshin[*]yashiki (Buddhist priests without temples), and a yashiki for shishimai performers are listed. In
1687, in a village of the same district, officials had to intervene in a dispute between kawata, Ebisu (Ainu), and shishimai
dancers regarding group rank, a dispute that was settled in favor of the kawata.[28] Throughout the seventeenth century the
authorities had often settled disputes among various classes or occupations of "lower people," sometimes by consigning them,
as they did the hinin and entertainers, to the jurisdiction of the kawata.[29]

In the case of Chuemon[*] the local temple mediated a conciliation. The domain meted out punishments to both parties,
differentially but graded according to status, for having caused so much trouble. Chuemon[*] was

[27] The penal code distinguishes between mojin[*] and biwahoshi[*] (see TKKk, 4:234). Internally, the zato[*] were organized
hierarchically into dozens of ranks at a national level under the jurisdiction of Kengyo[*] in Kyoto (see KDJ 6:406, s.v.
"Kengyo[*] ").

[28] Banba Masatomo, "'Buraku' no seiritsu ni tsuite: Shinshu[*] ni okeru burakushi sobyo[*] (2)," Shinano 12, no. 9 (1960): 49,
59.

[29] See Tsukada Takashi, "Kasomin[*] no sekai: 'Mibunteki shuen[*] ' no shiten kara," in MK, 225-67.

― 257 ―

sentenced to domiciliary confinement (oshikome ) for ten days and was manacled for between fifteen and thirty days. The
village officials received a lighter sentence: domiciliary confinement (enryo ) for between ten and thirty days.[30] The
authorities did not settle the question of Chuemon's[*] status.

As a very peculiar genealogical reckoning, genealogy was invoked as the ultimate proof of Chuemon's[*] low status. One
entertainer ancestor erased his long and venerable peasant lineage. "Genealogical descent," like occupation, language, or skin
color (here as in other times and places), becomes the arbitrary, socially defined basis for a particular form of discrimination—
one of Tokugawa Japan's unique ("early modern") contributions to Japanese cultural and social practice that is not yet part of
the past even today. It did not exist as such in the early Tokugawa period, nor was it a legacy from the Middle Ages; rather, it
was a product of eighteenth-century Tokugawa society and culture. Another important point is that the verdict could have
been different. Aside from local public opinion, manipulating "genealogy" and association, official legal "world-making" had
also to do with whether one was a kawata or not.

The Kidnapping of a Hinin Woman, 1781

Kawata exercised policing functions in numerous villages far from the locale of their residence. They lived together, often only
a couple of families, especially in Shinano, which lacked the large kawata communities that were common to the Kinai—not
in, but next to, villages,

[30] There were various degrees of domiciliary confinement. For commoners only there was tojime (door closure), which
entailed nailing a bar across the door to prevent contact with the outside world; having only some effect in the close quarters
of city neighborhoods, it was replaced by reprimands and fines in 1740. Oshikome (shutting up) was not limited to commoners
and forbade social intercourse for twenty to one hundred days, but the doors were not locked. For warriors and priests, there
were three other kinds of confinement. In heimon (gate closure), the most serious penalty of the three, all gates and windows
were closed, though they were not nailed shut, and contact with the outside world was forbidden for a period of fifty to one
hundred days. In hissoku (forced closure), the gates were closed for thirty to fifty days, but the prisoner was allowed to sneak
discretely through the wicket in the main gate or through a side entrance at night. In enryo (restraint), the lightest form of
domiciliary confinement for elites, the gates were closed, but the wicket or side door was left unlatched (hikiyosete oku ),
allowing similar goings and comings at night (see TKKk, 4:231-32, and the relevant entries in KDJ).

― 258 ―

outside the gates, at the borders, or near the entrances or exits of way stations.[31] Although they were segregated from the
other villagers, they nevertheless were considered part of the village in some ways, especially when it came to confrontations
with other villages, as the following example, also from Okudono domain, shows.

This incident, in 1781, involved Mon, the daughter of Kanesuke, a hinin from Kami-Kaize village (Minami-Saku district).
Kanesuke had lived there for several decades without having been properly registered: he was an illegal resident. He made a
living working for two kawata who had shares in the dannaba where Kami-Kaize was located. One of these, Magoichi, was
from the kawata hamlet of Tanoguchi; the other, Tahee, was from Kaetsu (both villages were located seven kilometers north of
Kami-Kaize).[32] Hinin often worked for kawata, but in this area there were no hinin communities. Kanesuke, not registered
and without the support of a network of peers, had a client-patron relation with Magoichi and Tahee.

Mon had been involved with a certain Heizo[*] , a registered peasant from Shiga village (seventeen kilometers to the north, in
bakufu territory), for over a year when, on 1781/5/26, she disappeared. Suspicion fell immediately upon Heizo[*] , who had
visited her that day. Kanesuke alerted his patron Magoichi, who notified Yata, the prison guard at Tanoguchi and probably the
head of Tanoguchi domain's kawata.

Heizo[*] was found in Oiwake station on the Nakasendo[*] , close to Karuizawa (thirty kilometers from Kami-Kaize and
eighteen kilometers from Shiga). Yata tied him up and took him into custody in his own house. After learning from Heizo[*]
that he had taken Mon along and entrusted her to an inn in Oiwake, Yata handed him over to Magoichi on the third day of the
following month (the intercalary fifth month). The next thing to do was to get Mon back to her father, Kanesuke. Until then
they would keep Heizo[*] with the kawata community in Tanoguchi. The matter became complicated, however.

On the seventh, the village officials of Kami-Kaize became involved and pressed more details from Heizo[*] . Mon and Heizo[*]
had known each other for one year and had plans to get married. They had eloped (or he

[31] Ozaki, Shishu[*]hisabetsu buraku , 30, 32.

[32] Ozaki Yukiya, "Shinano kuni Saku-gun ni okeru buraku no shiteki kosatsu[*] ," Buraku mondai kenkyu[*] 18 (1964): 93-98; a
short version of this incident can be found in Ozaki, Shinshu[*]hisabetsu buraku , 261-62.

― 259 ―

had taken her), but since he could not marry her soon, he had thought of putting her into service for four or five years at an
inn at Oiwake. Why did the village officials become involved in a matter of an illegal resident?

The kawata had originally hoped to solve this case by themselves, without recourse to the authorities, in part, no doubt
because of Mon's and her father's nonregistered status. Their plan was to quietly arrange an exchange of Heizo[*] for Mon
without much further ado. They had expected Heizo's[*] village of Shiga to find Mon and bring her home. Shiga village,
however, had no intention of looking for Mon, but assumed, nevertheless, that Heizo[*] would be allowed to return without a
problem. On the fifth day of the month, they had even refused to negotiate the matter with the kawata. That was when Yata
and Magoichi handed the affair over to the village officials of Kami-Kaize. Heizo[*] was transferred there on the seventh, and
negotiations with Shiga started.

As before, the Shiga officials showed no interest in helping search for Mon. In addition, they were lukewarm about Heizo's[*]
return, because they found it "difficult to accept that Heizo[*] had been imprisoned by kawata from Tanoguchi." The
assumption may have been that it was the kawata's job to catch criminals, but they had no business locking them up, for we
know that the representatives from Kami-Kaize objected to this way of putting it. They argued that what they had done to
Heizo[*] was something altogether different from throwing him in jail. Shiga village complained also that "the apprehension
and jailing by a kawata would cause Heizo[*] to lose his status."

Legally speaking, this was nonsense. The mere fact of being caught for a misdemeanor or felony did not automatically result in
loss of status. When commoners were sentenced to become hinin, this was the result of a specific verdict. The allegation here,
however, is that close contact with a kawata would lower Heizo's[*] status (the same argument as in the Bonboku Incident), or
at least cause the loss of his good name, although the term used was mibun , "status." After learning either that Heizo[*] had
intended to marry a hinin (which toward the end of the eighteenth century would result in loss of commoner status) or that he
had kidnapped a woman who had then disappeared, his own villagers thought that perhaps Heizo[*] had become a liability
and therefore already considered him not to be one of them.

In the negotiations on the eleventh between the two headmen, Shiga indeed refused to take Heizo[*] back and insisted that
Kami-Kaize help

― 260 ―

with the search for Mon. Kami-Kaize had lost its leverage for forcing Mon's return now that Shiga, for the most spurious and
insulting reasons, did not want Heizo[*] back. The infuriated officials of Kami-kaize decided to lodge a formal suit, risking even
more trouble because of Kanesuke's illegal situation. They even threatened to carry their case to Edo if the local intendant's
office did not agree that Heizo[*] should be handed over only after Mon had been returned.

Shiga's insult by way of status pollution pushed the confrontation to new heights by exploiting to the extreme the shared
popular notions about kawata. Initially the kawata had aimed for an informal solution, without drawing official attention. As
the village authorities became involved, Kami-Kaize identified increasingly with the kawata and hinin that served the village.
Local pride took over, and Kami-Kaize declared itself to stand by them all the way.

The respective intendants of each village, notified of the lawsuit, were still eager to settle out of court. They chose new
intermediaries, who succeeded in having both sides agree to Heizo's[*] return to Shiga without any reference to the problem of
Mon's return to Kami-Kaize. The reason given was that neither the innkeeper to whom Mon had been entrusted nor Mon
herself could be found. It is likely that the officials from Shiga had known this all along and had already abandoned any hope
of getting Heizo[*] back, since they could not offer Mon in exchange. Shiga's professed reasons had driven the Kami-Kaize
officials to initiate the suit, but after they found out that Mon's whereabouts were not known they could not comply with the
mediators' insistence that Heizo[*] be allowed to return to his village. This, however, was what the intendants ultimately
ordered.
The intendants judged the case closed after Heizo[*] had returned to the village where he was registered. They do not seem to
have further investigated Mon's disappearance or even her alleged service contract. After all, she was a woman, daughter of a
hinin, and unregistered, far less valuable than a registered peasant and his return to his village. While the village authorities
of Kami-Kaize rallied around one of their subjects, the "law" as represented by the intendants shared the popular
discriminatory views. We are left in the dark about Mon's fate.

Anyone who knows anything about Oiwake at that time, however, would probably share Ozaki Yukiya's supposition that Mon
was sold into prostitution there and was whisked away as soon as the affair came

― 261 ―

to light. Oiwake, with approximately two hundred prostitutes, was one of the largest pleasure stops on the Nakasendo[*] .
Many of these women came from the ranks of kawata and hinin. Therefore, it is likely that the intendants did not think Mon
was worth the trouble it would take to find her.

This incident illustrates how the notion of status pollution or illegal residency could be played up or ignored, how values could
be strategically modified in a power game in which pollution was not the issue but a weapon. A village (Kami-Kaize) could
decide, at some risk, (1) to treat an illegal hinin as one of its own, (2) to back efforts by kawata to right a wrong, and (3) to
threaten with a lawsuit even if it meant going to Edo. Or a village (Shiga) could write off one of its own registered peasants
under the pretext that his status had been polluted through contact with kawata. The outcome suggests that the authorities
beyond the village shared prejudicial views against kawata and hinin, favoring the peasant.

The Clog Thongs Riot, 1843

The law sometimes was slanted significantly in favor of the peasants. Most famous in this respect is the outcome of a riot of
peasants against kawata known as the Clog Thongs Riot.[33] The production of footgear was one of the common occupations of
kawata.[34] On 1843/7/22 Tatsugoro[*] , from the kawata hamlet of Nagase village in Iruma district (Musashino province), went
to sell thongs (hanao ) for wooden clogs (geta) at the market of Ogoseimaichi village. At the end of the day, Tatsugoro[*] still had
eighteen pairs of thongs, which he tried to sell to a certain Hinoya Kiee, who did not show much interest in them. A number of
people gathered to watch the two men haggle over the price. They soon joined in, hurling insults at Tatsugoro[*] , and
eventually forced him to sell his thongs at a price far below what he considered fair.

[33] The four documents concerning this incident, two of which are court records, are available in NSSS 14:569-96. I rely
heavily on the introductory summary of the incident (569-70) rather than on the brief sketch, inaccurate in some detail, given
in Teraki, Kinsei buraku , 64-65.

[34] Watanabe village in Settsu province employed four hundred people making various kinds of footgear, which they were
allowed to sell on fixed dates at nine designated locations in Osaka (Teraki, Kinsei buraku , 51). On the economics of the leather
trade in nearby Minami-Oji[*] , see Hatanaka, "'Kawata' mibun," 308-21.

― 262 ―

Angry, he protested vociferously, but the hinin market guard forced him to return home.

That evening Tatsugoro[*] talked things over with the members of his hamlet, and a fairly large number of them set out for
Hinoya Kiee's house in Ogoseimaichi, yelling and shouting. The next day they forced their way into his compound. At that
point someone from the neighboring village of Ueno, acting as mediator, quieted the crowd down, listened to both sides, and
put the blame squarely on Kiee. The people from Ogoseimaichi, in accord with neighboring villages, filed a suit with the Kanto
Bureau of Investigations. The kawata, in turn, filed one with Danzaemon in Edo. Meanwhile, following the incident,
Ogoseimaichi decided to bar kawata from their market in the future. The devastating effect this would have on the livelihood
of the kawata may help explain the massive response that followed.

Twenty-nine peasants from Ogoseimaichi forced their way into the Nagase kawata community on 8/5 with the intention of
capturing Tatsugoro[*] , allegedly with the approval of the Kanto investigating officer. The kawata, however, were prepared.
With the help of some "500 kawata" from the area armed with bamboo spears, they instead rounded up the 29 intruders and
locked them in a shed. Two days later, on the seventh, an officer from the Kanto bureau arrived and succeeded in getting the
captives released on a promise that both sides would get a fair hearing. Then an investigating officer arrived from Edo, but
with quite a different agenda. The two officials conferred on the tenth, and the higher-up from Edo decided on a policy of
forceful suppression.

The next day he mobilized over 300 peasants from the surrounding villages, 34 of them armed with rifles, and ordered a
roundup that yielded 252 kawata from twenty-three villages in three days. On the fifteenth, yet another officer arrived from
Edo to mobilize 830 peasants and 101 riflemen as a show of force and to prevent any retaliation. One week later, on the
twenty-fourth, 97 of the kawata arrested were accompanied by over 700 peasant guards on a three-day journey to Edo. The
investigations and trial lasted two years, perhaps because of the sheer number of suspects, perhaps because of the bakufu's
determination to emphasize the gravity of the case. The long duration was also a form of punishment, since the kawata were
being detained during that time at considerable cost to their families, their communities, and their health.

― 263 ―

The "verdict" clearly favored the peasants. As is so often the case with "verdicts," the blame and punishment were shared,
although very unevenly. The tendency to punish both parties of a dispute seems to have been part of the concept, or facade
(tatemae), of "justice" meted out by the authorities. The kawata from Nagase village were severely reprimanded, and 102
received the following sentences: gibbetting of the head after decapitation (1 case); simple beheading (1); major deportation
(4); medium deportation after flogging (3, including Tatsugoro[*] ); banishment from Edo (1); expulsion from the community
(1); flogging and banishment beyond ten miles from Edo (1); flogging (1); manacles (85); house confinement (2); and demotion
from kawata head, kogashira (2). In addition, Danzaemon was condemned to house arrest (heimon) in Edo. In stark contrast,
only a few peasants were found guilty. None of them had been detained, and their punishments were not physical, but limited
to manacles (5) and fines: Nagase's headman was fined three kanmon, and the kumi heads were severely reprimanded; in
Ogoseimaichi, the headman was fined five kanmon, and the kumi heads three.

The imbalance between the two verdicts becomes staggering when one discovers that in the course of their two-year
presentencing detainment, sixteen kawata died from "unspecified illness," including Tatsugoro[*] and all of those condemned
to death and deportation. It appears that the most serious sentences, following rather than preceding the punishment, ratified
faits accomplis .

The details of this event are known to us through four documents: two separate court records and two private narratives by
the same writer. In the latter are two memorable quotes that reveal kawata sentiments, if not those of the participants then at
least those of the narrator. According to Nagayoshi, one of the kawata who was condemned to maximum deportation but who
died while still in prison, "This uprising is the first since the beginning of Edo and is unique in Japan (Nihon ichi ). To have
come to stand out as those who started the riot is for us, precious instruments, the fulfillment of a divine promise
(senryo[*]dogu[*]warewaretomo myoga[*]ni kanashi tokoro nari ); if you die once, you do not have to die again; to sacrifice one's
life [for this cause] is our deepest wish." Sadaemon, who shared a similar punishment and fate, threatened revenge after
death: "We are wild gods (arakamisama ) even though we are of vulgar and low status (gesugero[*] no mibun ); if even

― 264 ―

an ant's wish rises to Heaven, wouldn't our resentment reach there too?"[35]

Toward the end of the Tokugawa period, minor incidents could escalate and trigger persecution of kawata by peasants and
authorities that popular prejudice alone fails to explain. It is perhaps significant that the so-called Clog Thongs Riot originated
at a market. The kawata's policing work in their dannaba and in prisons was a performance of official duties by a minority of
the kawata. As vendors, which most of them were, they vied for profit. Official documents often contain complaints that
kawata harassed commoners while trying to sell their wares (see arts. 2 and 5 of Komoro domain's regulations for outcastes
and art. 3 of those of Netsu village, in appendix 5). Ordinary peasants and merchants may have begun to resent the
competition and success of kawata in the market. It is certainly striking that while the local mediator at Hinoya Kiee's house
found Tatsugoro[*] innocent and put the blame squarely on Kiee, Ogoseimaichi village reacted by closing its market to all
kawata. Furthermore, the kawata reacted as a group when one of their own was wronged. The discussion that took place after
Tatsugoro[*] returned home points to a shared consciousness of injustice to themselves as a class and a sense of presumed
rights that needed defending. After the protest at Kiee's had become a massive and successful resistance against an attacking
group of peasants—for it was the peasants who rioted—it was this class action that led to a severe reaction by the authorities,
who blamed the victims, the kawata, for defending themselves. Had this been a matter of one peasant village locking up a mob
of peasants from elsewhere who were seeking to settle a score, the authorities probably would not have reacted in this way.

Resisting Repression at the Village Level, 1856

Small incidents could lead to sweeping discriminatory measures, but sometimes it was the other way around: sometimes
official discrimination triggered resistance. In the following two instances, the kawata confronted the authorities directly and
voiced a reasoned critique of discrimination.

[35] NSSS 14:582. The same proverb about an ant's wish rising to Heaven is also applied to the suffering of horses and bulls, in
article 13 of the Regulations for the Villages of All Provinces of 1649, as adapted by an intendant m Kita-Saku in 1665 (see
appendix 4).

― 265 ―

In early 1856/4 five kawata from Mure village (under the intendant of Nakanojo in Shinano) were caught gambling in
Naganuma village.[36] Within days this led not only to a new prohibition against gambling but also to a series of repressive
regulations, to which the kawata responded immediately in a carefully argued written plaint. They listed the prohibitions and
their responses to them one by one. In particular, the prohibition against the use of regular or high wooden clogs, umbrellas,
or parasols even in snow- or rainstorms (only straw coats were allowed) and the wearing of leather-soled clogs and tabi socks
was countered with the argument that "even among people of [low] status, there is not one who does not experience the four
seasons' hot and cold weather conditions."[37] In other words, they argued that they were humans too.

This was an appeal for fair treatment based on a human commonality rather than on institutional status divisions. The
ultimate victims of status distinctions, the kawata, raised a protest to an unsympathetic audience, reminiscent of Plato's
unmasking of ideology in the Gorgias (523e), where he speaks of the same kind of judges who "have placed in front of their
souls a veil made of their eyes and ears and their whole bodies.[38]

Another prohibition made the kawata aliens in their own country, for it forbade certain cultural practices, namely, the use of
banners and high lanterns at festivals. The kawata argued that these were "symbols of gratitude for the benevolence of the
Great Peace and the radiance of the grace bestowed by the country" and such a policy was alien to "our national tradition." In
other words, they argued that they also were Japanese. They continued, "Although the order to slight gods and

[36] Ozaki, Shinshu[*]hisabetsu buraku , 163-69.

[37] Ibid., 165. Komachi Yuhachi[*] , a Shingaku teacher, voiced a similar sentiment in 1828 when he advocated empathy for
the despised classes by pointing out that when one feels cold or hunger, others will feel the same (see Komachi Yuhachi[*] ,
Jishuhen[*] , in Nihon keizai sosho[*] , ed. Takimoto Sei'ichi, vol. 19 [Nihon keizai sosho[*] kankokai[*] , 1915], 445).

[38] Plato, quoted in The Levinas Reader , ed. Séan Hand (Oxford: Basil Blackwell, 1989), 2 43. Levinas writes that "Plato sets
forth a beyond of institutional justice, like that of the dead judging the dead (Gorgias 523e), as if the justice of the living could
not pass beyond the clothing of men, that is could not penetrate the attributes that in others, offer themselves to knowing, to
knowledge, as if that justice could not pass beyond the qualities that mask men .... In the social community, the community of
clothed beings, the privileges of rank obstruct justice."

― 266 ―

Buddhas may very well happen in some regimes under heaven, such things are [only] typical of foreign, 'ruined' countries."

In addition to asserting their right to be treated fairly as human beings and as Japanese, the kawata argued that, as subjects,
they were no different from the rest of the population. When they were accused of disregarding prescribed segregation by
"passing" as regular commoners, they retorted that the breach had actually occurred from the other side: "In recent years,
peasants have behaved like kawata when eating the meat of cows and horses." To the prohibition against the hanging of
signboards, they replied that such a regulation would depress business, making it difficult to feed their families, especially
since they received no compensation for their police work. Noting that they performed national corvée (kuniyaku) just as
everyone else did, the kawata also protested discriminatory regulations regarding the style of their homes. They concluded
with a political critique, redefining as injustice what was being officially represented as a reform (kaikaku ). By arguing that
reforms were rare, they implied that this was not a proper occasion for one. Moreover, to institute a "reform" in addition to
punishing the gamblers constituted double, and therefore unjust, punishment. They understood the bias of the punishment,
for they pointed out that the peasants in the village had heard nothing about an order for a "reform": "only we, kawata, have
been targeted for such treatment."

The village officials responded to this written protest the next day (the fifteenth) by initiating an investigation into the kawata
involved and by lodging a suit with the intendant because their orders had "met with scorn and were not obeyed." On the
sixteenth, three kawata were summoned to appear before the intendant. They set out the next day but got only as far as the
village of Nakano, where, after calling upon fellow kawata, they disappeared. On the eighteenth a new summons arrived for
other kawata, but when the village officials from Mute went to the kawata hamlet to deliver the summons, there were no men
to be found there. Allegedly, they had all left the night before for an unknown destination. Under pressure, however, it came
out that two nights earlier they had made plans to disappear with other kawata of the area.

Absconding kawata trying to "pass" as regular commoners came to be defined as a major social problem in the late eighteenth
century.[39]

[39] See, e.g., laws issued in 1778, 1795, 1796, and 1799 (Kobayashi, Kinsei horeishu[*] , 151, 215, 217, 238-39).

― 267 ―

Yet, the detailed criminal records of Nagasaki indicate that quite a number of them returned to their original communities.
Unable to overcome the numerous obstacles to entering society, such as the need for sponsors (lineages in many villages and
individuals in towns), the requirement to present identification certificates when settling in a new locale, and so on, they
returned.[40]A fortiori , this was the case with group exodus such as the one in Mure, for the fugitive males returned a couple
of weeks later. Some of them were jailed, others reprimanded, and the local temple petitioned a pardon with unknown (to us)
results.

Toward the end of the Tokugawa period some kawata protested and resisted what they explicitly identified as groundless,
unjust, and cruel treatment, not just by other Japanese but by officials as well. The kawata argued not only that they were
human beings and Japanese but that they were full members of the realm even though the state did not acknowledge their
"citizenship." Contrary to what Neary writes, they did not think of themselves as "impure, not quite human, not deserving
equal treatment," and they said so, not only among themselves but in an eloquent written protest.

Resisting Repression at the Domain Level, 1856

The most famous instance of defiance took place in Okayama domain in 1856.[41] The context here also concerns a "reform."
As a result of the political and economic repercussions following Commodore Perry's opening of Japan in 1853, Okayama
domain in 1855/11 initiated a series of social reforms that included a decree on frugality. The last five of this decree's twenty-
nine articles came to be known as a special decree (betsudan ofuregaki ) because they applied only to kawata. Among the
stipulations were the following: Clothing must be plain, without designs or crests, and either yellowish-brown, persimmon, or
indigo blue in color. Kawata had to remove their wooden clogs whenever they met peasants, and they could not wear them at
all when they went to other villages. These regulations were typical of discrim-

[40] See Harada Tomohiko's remarks in Kobayashi Shigeru et al., "Zadankai," 494.

[41] KDJ 11:912, s.v. "Bizen no kuni Okayama hanryo[*] Ansei sannen shibuzome ikki"; Teraki, Kinsei buraku , 67-69. The
current name, Shibuzome ikki (Persimmon Dye Riot), was given at the occasion of its centennial anniversary, in 1956 (see also
Cornell, "Caste Patron," 60-62).

― 268 ―

inatory laws against kawata, but they also applied to genin and small landholders (in Hozu) and even new titled peasants (in
Tanoguchi) (see chapter 3). This "reform" was to be enforced by the beginning of 1856. On 1855/12/27-28 the village headmen
informed the respective kawata heads of the new regulations and requested their seals of approval.
Shocked by the stepped-up discrimination, the kawata heads insisted on postponing ratification to allow time for prior
discussion in their communities. The hamlet assemblies all opposed ratification, which led to a number of regional meetings of
kawata heads. First the leaders of five urban hamlets convened at the Jofukuji[*] (in Shimo-Ifuku village, Mino district), the
head temple for all the kawata of Bizen province, under the pretext of paying the customary New Year's visit to their temple.
There they decided to mobilize all the kawata communities of the domain at a general meeting in order to press for the repeal
of the discriminatory laws. Seventeen heads of rural hamlets, however, had already agreed at a separate meeting to first ratify
the laws and then press for a repeal. On 1856/1/15 the heads of fifty-three communities held a stormy meeting at the Jofukuji[*]
in which the leaders from the countryside resisted the more radical plan of their urban colleagues. After several sessions the
radicals prevailed and drafted a formal petition to withdraw the laws.

The kawata's arguments against the discriminatory regulations were clearly spelled out: (1) kawata were cultivators, paid
tribute, and were thus "honorable," that is, titled peasants who should be treated like other peasants and not singled out for
separate treatment in a time of crisis;[42] (2) many of their fields were of low quality, increased discrimination would squelch
the incentive for young people to work them, and therefore the fields would lie fallow and produce no tribute; and (3)

[42] In other incidents also the kawata expressed their self-image as onbyakusho[*] , that is, honorable, or titled, peasants (see
Mae Kei'ichi, "Kinsei chu-koki[*] ni okeru 'kawata' no keizai seikatsu," m Burakushi no kenkyu[*] : Zenkindai-hen , ed. Buraku
mondai kenkyujo[*] [Kyoto: Buraku mondai kenkyujo[*] , 1984], 266-70). We saw in chapter z how a whole new community of
exclusively peasant kawata was created in Umaji in 1808 (Igeta, "Kuchi-Tanba," 97); Kisaki village, in Tanba, counted twenty
landholding kawata, some of them extremely wealthy (owning thirty-six koku, some of it in neighboring villages) (see Igeta
Ryoji[*] , "Iriai sabetsu to buraku mondai," Kindai Kyoto no burakushi , ed. Buraku mondai kenkyujo[*] , Kyoto no buraku
mondai, 2 [Kyoto: Buraku mondai kenkyujo[*] , 1986], 171).

― 269 ―

when they could not meet the tax quota, many kawata obtained the remaining cash by pawning clothes, which would now
become impossible if the proposed restrictions on the kinds of clothes they could wear were imposed. This also would
negatively affect their ability to pay tribute. It should be noted that this last point implies that peasants were engaged in
economic exchange with the kawata and accepted kawata clothing for pawn without much fear of pollution.

The domain denied the petition on 4/6. The intendant (gundai) charged with seeing that the regulations were ratified
pressured the village group headmen to secure the signatures of all the kawata heads. One by one the kawata heads
capitulated (some under torture) and signed the documents. On 4/15, however, another general assembly (the fifteenth) was
convened. In the tense atmosphere something unexpected happened. The kawata leadership had thus far remained within the
law by taking the petition route, but now they were criticized by the rank and file, who raised the possibility of making an
illegal, direct appeal (goso[*] ). In subsequent strategy deliberations it was decided not to surprise the domain lord by going to
him directly but instead to approach a certain elder who had a "liberal" reputation. The elder headed a rural office in
Mushiage, in Oku district (some twenty kilometers from Okayama castle), which geographically was perhaps a safer place to
stage a mass protest.

The domain officials, aware of what was brewing, on 6/9 ordered a village group headman to "investigate" the Konoshita[*]
hamlet, which had taken the lead in recommending the direct appeal. This immediately triggered the decision to proceed with
the illegal protest. Mobilization instructions were hurried to all males between the ages of fifteen and sixty of all fifty-three
kawata communities. In the early morning of 6/13, disregarding pleas and threats from the village group headmen and village
headmen, some fifteen hundred to two thousand kawata gathered in a dry riverbed in Yokaichi[*] village (Oku district). The
next afternoon they started their march in the direction of Mushiage. By night they had arrived at Sayama village, where they
put up camp and made preparations for the next day's trek to their destination.

The elder, of course, had been informed about this small army on the march and had prepared troops armed with cannon to
welcome the protesters. He met them at Sayama the following day, however, and participated in negotiations that ended in a
promise that he would forward their petition to the domain's council of elder retainers, who

― 270 ―

would then review the discriminatory laws. When they had received this promise, the kawata disbanded. The domain council,
while appreciative of the compassion shown to the protesters, felt that the honor and authority of the village headmen were at
stake and ordered the immediate ratification by the kawata leadership, which now had no choice but to comply.

Punishments followed betrayal. Twelve of the ringleaders were arrested and jailed; by the time they were sentenced, three
years later (in 1859/6), half of them had died. The entire populations of the communities that had participated in the protest
were sentenced to one to two weeks' domiciliary confinement. A few of the officials—the village group headmen, the village
headmen, and the Oku district commissioner and inspector—were held responsible for the mismanagement of the situation,
and they also were punished with domiciliary confinement. This incident of protest, resistance, and reprisal has inspired the
struggles of burakumin against continued discrimination in the twentieth century.[43]

The Tag Incident: Gorobe-shinden, 1978

Today the burakumin conduct their antidiscrimination drive on many fronts: in the workplace, in the courts, in the media, in
religious organizations, and throughout the education system. Reclamation of their history is as vital to the burakumin as
ethnic studies are to minorities in the United States and elsewhere. An incident in Mimayose, near Gorobe-shinden, led quite
unexpectedly to such a historiographical effort.[44]

On March 10, 1978, a resident of a public housing complex in the village of Mimayose found a small note (7 cm2 ) tacked to his
front door. On it was scribbled: "In Mimayose [this word is written in katakana] there is not even one tsubo [3.3 square
meters] of land occupied by a chorippo [derogatory local slang for burakumin]. Get your face

[43] For examples of discrimination in the twentieth century, see Ninomiya, "Inquiry," 115-25; and Hane Mikiso, Peasants,
Rebels, and Outcastes: The Underside of Modern Japan (New York: Pantheon Books, 1982), 139-71.

[44] In Tokugawa times, Mimayose (which was located five kilometers east of Ken's Makibuse and was the place where Ken
lived with her first husband) was a small village. It bordered on the Nakasendo[*] and the fields of Gorobe-shinden, which we
have come across several times already (see map 2; see also Saito[*] Yoichi[*] , Gorobe-shinden , 16 ff., 151-58).

― 271 ―

out of here. [Signed:] The people of the community (chorippo no sumu tochi wa, Mimayose ni wa hito tsubo mo nai; hayaku dete
ike. Kumin ichido[*] )." The victim was not a burakumin, but his older sister, it seems, had married one, a genealogical pollution
of the sort in the Bonboku Incident two hundred years earlier in Ozawa[*] village, less than ten kilometers to the southeast.
Although Mimayose had no buraku, there were three in the administrative village of Asashina, of which Mimayose was a part.
When the burakumin heard of this incident—the fifth in five years—they decided to take action. The burakumin in Asashina
collected garbage, disposed of hazardous materials, and emptied latrines. When they went on a strike lasting twenty-nine
days, the community was forced to publicly address the problem of discrimination in a number of meetings with the
burakumin's representatives, while its garbage went uncollected.

The burakumin, incensed by the allegation that "there is not a single burakumin in this community," decided to reclaim their
past, which brought them into confrontation with the elite Gakushuin[*] University in Tokyo. During the Tokugawa period, the
three buraku of what is today Asashina village were part of Gorobe-shinden, a village whose history has been studied by a
number of scholars, among them Oishi[*] Shinzaburo[*] ,[45] because of the extraordinary technology involved in the building
of the irrigation system when Gorobe-shinden was developed in the 1620s. Some twenty thousand documents preserved in the
Yanagisawa house of Gorobe-shinden's hereditary headman had been transferred to Gakushuin[*] University; an additional
ten thousand had remained at various locations in the area.[46]

Now the garbage collectors of Asashina started a drive for the return of these documents, which was complicated by legal
issues of ownership and custody. They eventually won their battle, and all thirty thousand documents are now in a memorial
building, the Gorobe Kinenkan, in Asashina. The reclamation of their past has just begun.[47]

[45] Oishi[*] , Kinsei sonraku , chap. 3.

[46] For a photo of the successor to the Yanagisawa house standing next to a small stone shrine built by his ancestor in 1643, as
well as a photo of the wooden water duct (part of a canal that was twenty kilometers long) crossing a river, see ibid.,
frontispiece.

[47] Saito[*] Yoichi's[*]Gorobe-shinden is an early product of this research.

― 272 ―

Dimensions and Developments

These episodes provide a sense of the complexity of the status question surrounding the Tokugawa ancestors of today's
burakumin. The kawata were part of the administrative apparatus: they were state functionaries, but polluted ones. Pollution,
however, was not only discriminatory and prejudicial but also discriminating and selective: in certain areas of daily
intercourse, in certain circumstances, it was suspended in practice; in others, it could be activated. And there is clear evidence
that the kawata's self-perception diverged from the perceptions others had of them, that they related to their position
differently than the non-kawata did. Let us now attempt a more systematic analysis of some dimensions of this notion of
pollution as a basis for a status racism and its evolution during the Tokugawa period.

Pollution, the Elastic Idiom of Discrimination

Demographic evidence of contemporary burakumin communities, such as the Ministry of Welfare's survey of 1935, shows a
strikingly uneven geographic distribution. Although population figures may be too low, by 100 to 150 percent, the data on the
number and location of burakumin communities , harder to overlook, are considered to be fairly accurate.

It is immediately evident from table 21 that the western half of Japan's main island, Honshu[*] (the Kinki and Chugoku[*]
regions in the table), accounts for a disproportionately high percentage of buraku, with the highest concentration in the
relatively small region around Kyoto. The eastern half, settled later, counts far fewer buraku, and in the northeast region there
are virtually none. We shall return to this question of historical diffusion later. Also, in the Kinki area (another name for Kinai)
buraku tend to be rather large, while in the east they are tiny and scattered.

This demographic variation, which goes back to the Tokugawa period, puts into question a theory that links discrimination
against kawata directly to Shinto and Buddhist religious notions of pollution and indirectly to occupations involving the
butchering and skinning of animals. This is not to deny that religious formulations and taboos played a part, representational
or justificatory, in discriminatory practice against kawata. However, the almost total absence of kawata

― 273 ―

Table 21.
Distributio of Buraku Population, 1935

Communities Population

No. per
Region % thousand %

Tohoku[*] (northeast) 9 0 I 0

Kanto[*] (large Tokyo region) 835 15 104 10

Chubu[*] (central; around 751 15 75 8

Nagoya)

Kinki† (Kyoto, Osaka, Kobe) 1,042 19 438 44

Chugoku[*] (southwest) 1,197 22 147 15

Shikoku 797 15 122 12

Kyushu[*] 736 14 112 11

Total 5,367 100 999 100

† Kinki is another name for Kinai.

SOURCE : De Vos and Wagatsuma, Japan's Invisible Race , 117; for a map,
see ibid., inside back cover.

communities in northeastern Japan, a region not particularly less Shinto and Buddhist than the rest of the country, points to
the inadequacy of such an explanation. Some people there must have disposed of dead animals and manufactured leather
goods as well. Those who did may have been despised, as were many others, as senmin and subject to temporary pollutions,
but as far as we know, they were not assigned a special social status on the basis of this pollution or subjected to
discriminatory practices and laws.[48] In other words, the northeast may have been without eta, but it was not without kawata.
[49]

It is important to note that during the Tokugawa period the linkage between certain occupations and kawata status was far
from uniform

48. Watanabe, Mikaiho[*]buraku , 128.

[49] In several of his publications, the medievalist Amino Yoshihiko has argued for marked cultural differences between
eastern and western Japan, one of them being the absence of a taboo against the consumption of horse meat and a weaker
sense of pollution in eastern Japan (associated with the early prehistoric Jomon[*] period) than in western Japan (linked to the
later prehistoric culture of Yayoi). See, e.g., his Nihon no rekishi o yominaosu (Chikuma shobo[*] , 1991), 140-43.

― 274 ―

or consistent. In some towns there were leather workers who were not categorized as kawata;[50] and other occupations that
were considered polluted were not attached to kawata communities.[51] Certain provinces—Tosa, Iyo, Aki, Bingo, Awa, and
Sanuki—had kawata fishermen.[52] Often the decision concerning which groups were to be legally stigmatized as kawata, that
is, specifically categorized as "polluted," rested with the authorities and was reached in court.

That legal pronouncements played such a prominent role in identifying rather arbitrarily the grounds on which certain
groups would be the object of hereditary discriminatory treatment may at first make this kind of discrimination seem quite
different from that resulting from Western racism, based on skin color. But in the West also legal decrees often played a
decisive role in deciding who belonged to one "race" or another, and membership sometimes could be purchased. Thus, in
1783 some blacks in Spanish America could become "white men" by fiat of the king of Spain; after 1795, legal white status was
for sale, which only sharpened whites' sensitivity about race.[53]
Thus, the notion of pollution was not only used to separate certain occupations based on the belief that pollution was inherent
in those occupations; it was also wielded as a weapon in intraoccupational disputes, where its strategic manipulation was
harder to hide, since these disputes were settled by administrative fiat. In 1768, for example, the village group headmen from
a district in Harima were given instructions about categories of polluted carpenters (yogore daiku ), which included kawata
carpenters working as cremators or as fabricators of instruments of torture and others such as basket makers or builders of
houses of prostitution.[54] This taxonomic clarification appears to have

[50] Watanabe, Mikaiho[*]buraku , 100. Thus a decree of 1794, possibly as a response to an inquiry into a dispute, stipulated
that although the making of bamboo hats and bamboo sandals was a kawata prerogative, hamayumi (exorcising bows used in
ridge-pole-raising ceremonies; toy bows and arrows) could also be made by town artisans (Kobayashi, Kinsei horeishu[*] , 208).

[51] Teraki, Kinsei buraku , 189.

[52] Watanabe, Mikaiho[*] buraku , 88.

[53] Leslie B. Rout, Jr., The African Experience in Spanish America , 1502 to the Present Day (Cambridge: Cambridge University
Press, 1976), 156; John Lynch, The Spanish American Revolutions , 1808-1826 (New York: W.W. Norton, 1973), 20-21. I thank
Igarashi Yoshikuni for calling these references to my attention.

[54] Kobayashi, Kinsei horeishu[*] , 139.

― 275 ―

been the result of an intratrade dispute in which certain groups of carpenters felt threatened by others and is, coincidentally,
closely related to an advanced stage in the division of labor. Peasants in Hozu (discussed in chapter 4), frustrated over the
course of a lawsuit by the kawata, obtained a ruling ordering the kawata to deliver their tribute rice on a different day from
the day the peasants delivered theirs, allegedly to avoid creating occasions for trouble.

The notion of pollution was flexible to a certain extent. Activated and propelled by social, economic, or political forces, it was
appropriated and applied in some situations but not others. From the existence of a general notion of pollution and its
application at certain sites, one cannot deduce or predict where else it would also appear: its application could be customary
(but custom is flexible), institutional (but institutions change), or situational (and therefore contestable). In the case of Mon,
the kidnapped hinin woman, the introduction of the notion of pollution changed the tenor of the confrontation between the
two villages. In the Bonboku Incident, the attempt to discredit a villager through association with the world of the kawata and
other lowly occupations failed. Some Tokugawa Japanese raised the question why hunting, skinning, and eating wild boar was
not considered polluting but the kawata were most concretely stigmatized for the skinning and slaughtering of cattle. Hunters
(matagi ), it seems, had succeeded by means of a lawsuit in 1694 to officially differentiate themselves from the polluted
kawata.[55] As Pierre Bourdieu has pointedly remarked, "Practice has a logic that is not that of the logician."[56] One cannot
essentialize pollution and grant it final explanatory value. Rather, one should historicize its functions and examine how and
when it was used and not used.

Logic would lead one to expect that rice grown by kawata was considered polluted. And, indeed, there exists a decree to that
effect, but it was enforced for only a short time. In 1720 the bakufu Office of Finance (Gokanjosho) issued an edict requiring
kawata, who, like peasants, had paid part of their tribute in rice and part in cash, to now pay the total amount in cash and to
store their bags of cash separately, marked eta osame ("eta tax"). The reason for this measure was revealed two years later,
when the law was rescinded, for then it was ruled that

[55] Mase Kumiko, "Ishiki no naka no mibunsei," MK: 274.

[56] Bourdieu, Logic of Practice , 86.

― 276 ―

"eta, hinin, and cremators could again, as before, contribute their polluted rice (kegaretaru mono ni tsuki kome ) as tax."[57]

Here, then, the logic of pollution drained a precious and pure commodity of all exchange value. Analogously, a different
commodity (one that was by nature polluted, one might say), namely, samurai excrement, took on a surplus valorization as a
result of the high social status of its "producers": in the night soil market, samurai excrement fetched the highest prices.[58]
One would expect the daimyo variety to have been even more valuable, but that does not seem to have been the case, perhaps
because it was not for sale. These examples show that the logic of status valorization or devalorization is a contingent rather
than an absolute one.

In 1558, Suwa shrine in Shinano recorded its taboos (monoimi no ki ), which stipulated very circumscribed periods for
bloodshed and contact with death: killing a person polluted the killer for one day, as did the removal of an animal carcass, and
skinning an animal polluted a person for five days. This hierarchy of pollution, perfectly adapted to a country torn by warfare,
suggests that there may not yet have been kawata in the region and that others may have handled dead animals at that time.
[59] Nor was pollution a permanent, existential condition for certain classes of people: it was a temporary and removable stain.

Reference to Shinto's concern with pollution as an explanation for the kawata/burakumin's imputed state of inherent pollution
misses the point that much of Shinto ritual focuses on the purification of individuals and their environment. No pollution was
considered unremovable. Even Deguchi Nao (1837-1918), who founded the new religion Omotokyo[*] in the late nineteenth
century believing that humanity had lapsed into a fallen and therefore polluted condition, did not view this predicament as
permanent or inalterable. Her millenarian vision predicted an imminent restoration of the world.[60] Thus, we need to address

[57] Kobayashi, Kinsei horeishu[*] , 111, 115.

[58] Walthall, "Village Networks," 295.

[59] Banba Masatomo, "'Buraku' no keisei ni kansuru kosatsu[*] : Shinshu[*] ni okeru burakushi sobyo[*] (1)," Shinano 12, no. 5
(1960): 18. In Amino's new interpretation (see above, n. 49), the Suwa taboos would constitute an argument for his thesis that
pollution associated with killing was less developed in eastern Japan than in western Japan.

[60] Emily Groszos Ooms, Women and Millenarian Protest in Meiji Japan: Deguchi Nao and Omotokyo[*] , Cornell East Asia
Series (Ithaca, N.Y.: Cornell University, East Asia Program, 1993).

― 277 ―

not only the question why kawata/burakumin were associated with pollution but also what the permanent character of that
association was.

As mentioned above, kawata joined peasants in some uprisings. They also cosigned with peasants oaths attached to lawsuits
(Kashii village, Izumi, 1708).[61] Yet peasants were also particularly insulted when kawata were mobilized against them. In a
confrontation between peasant and kawata youth in 1772 the peasants apparently shouted: "We can kill five or six eta without
being struck by a curse (tatari ) from anywhere!"[62] Later in the Tokugawa period, a similar devaluation of kawata lives was
expressed by an official. In a well-known incident in 1859, a kawata visiting a Shinto shrine in Edo was killed by a gang of
young men because he was "polluted." When Danzaemon requested the death penalty, the magistrate apparently responded
that a townsman's life was worth that of seven kawata and that therefore the quota to justify the execution of a townsman had
not been reached.[63] This ratio, it should be noted, was also applied to samurai and peasants, at least in rhetoric.[64]

In the seventeenth century, kawata not only served in villages as guards and champêtres but also worked in peasant
households or as day laborers, perhaps replacing the bond servants who had been set free.[65] Yet, in the first decades of the
eighteenth century, we find numerous prohibitions against kawatas' entering commoners' homes. That the segregationist laws
were not an empty letter is shown by the ways they were circumvented and the punishments that followed discovery. Mixed
marriages between majority Japanese and kawata occasionally functioned as channels for smuggling a number of kawata into
peasant and merchant houses, sometimes over distances as far as seventy kilometers, to perform labor as servants. If
discovered, the gobetweens, especially when they profited from the scheme, were punished more severely than the "illegal
aliens" themselves were.[66] The

[61] Teraki, Kinsei buraku , 152.

[62] Ibid., 61.

[63] Inoue Kiyoshi, (Kaihan ) Buraku mondai no kenkyu[*] : sono rekishi to kaiho[*]riron (Kyoto: Buraku mondai kenkyujo[*] ,
1959), 32-33; Saito[*] Yoichi[*] , Gorobe-shinden , 243; Ninomiya, "Inquiry," 98.

[64] Bix, Peasant Protest , 94.

[65] Watanabe, Mikaiho[*] buraku , 96, 107, 122, 132-33 (service); 125 (day laborers).

[66] Hatanaka, "'Kawata' mibun," 333; Igeta Ryoji[*] , "Kon'in sabetsu no rekishiteki shoso[*] : Edo bakufu no hanketsu o
tsujite[*] ," in Kindai tennosei[*]kokka no shakai togo[*] , ed. Mahara Tetsuo and Kakeya Saihei (Kyoto: Bunrikaku, 1991), 12-18
passim.

― 278 ―

mixed marriages, Ozaki Yukiya notes, were not unknown in Kita-Saku before the 1690s, and as a rule they were
unproblematic for the authorities.[67] Gradually, however, increased administrative pressure for segregation, starting in
earnest in the 1770s, led to severe sanctions against mixed marriages, now categorized with adultery; for commoners the
punishment was status demotion to hinin underling (tega ).[68]

Occupations—Degraded and Polluted

The term senmin ("despised people") was associated with certain occupations and has existed since the beginning of recorded
history in Japan. Generally speaking, in all societies, past and present, certain occupations seem to be perceived socially as low
and undesirable. Medieval Japan counted dozens of them. Danzaemon, the "national" head of the kawata/hinin since the
1720s, had fabricated his genealogy (at least in one version) through Heian warrior houses all the way back to China and listed
no fewer than twenty-eight lowly occupations among his ancestors as an argument for his rightful dominant position among
senmin.[69] Many of these occupations were still viable throughout the Tokugawa period, although this does not necessarily
imply that they were hereditary. As mentioned above, Tokugawa Japan counted many more lowly and polluted occupations
than those specifically held by kawata. Conversely, only some of the occupations, communities, and lineages of the kawata can
be traced to late medieval times.

The term eta , first mentioned in texts of the thirteenth century, was written with the characters for "plentiful pollution" only
some two centuries later, referring then to a subgroup of the kawaramono , or "riverbed people." The kawaramono lived on
the riverbanks in Kyoto. They were ranked among the lowest of the many senmin categories.

[67] Ozaki, Shinshu[*]hisabetsu buraku , 177.

[68] Igeta, "Kon'in sabetsu," 15. Satsuma domain, for one, in 1784 forbade intermarriage between peasants and "eta, kengo [?],
etc."; made "eta and the like" live separately from peasants; and shifted its idiom for discriminated people from death to
pollution when it changed the term from shiku (literally, "death pangs") to eta (Teraki, Kinsei buraku , 46).

[69] These genealogies can be found in Kobayashi, Kinsei horeishu[*] , 27-49; and NSSS 14:427-38.

― 279 ―

Riverbanks were not taxed, and their occupants—cleaners, plasterers, dyers, well diggers, gardeners, transporters of goods,
and so on, and many highly skilled laborers—worked for the noble and warrior establishment or were attached to temples as
jobbers. Kawaramono who earned their living cleaning estates, slaughtering animals, or working leather were specified as eta.
[70]

During the wars that engulfed Japan throughout the sixteenth century, leather was in great demand for the manufacture of
saddles, arms, and protective gear to equip the armies. Thus, some of the warlords (like the Imagawa in 1526 and the Hojo[*] in
1538) concentrated leather workers around their castles, subjected them to their immediate authority by cutting their ties with
whoever might have claimed them as dependents, organized them under one head, exacted levies in leather from them, and
forbade the export of leather goods to other domains. Later in the century other warlord houses, like the Maeda, the Ueda, and
the Takeda, followed the same practice.[71]

These changes in the social organization of leather production were part of the gradually developing efforts by daimyo to
mobilize all the sectors of society. Other artisans were similarly concentrated, organized, and forced, as the peasants were, to
provide yaku (service that included materiel as well as corvée) to the new lords. Among those in the "low" occupations, the
leather workers were most in demand. A number of them were moved from the countryside to castle towns, and when none
were in the area, they were imported from the provinces around Kyoto. In Kaga, Noto, and Etchu[*] leather workers were
brought in from Omi[*] and Tanba. The castle towns in Shinano were probably provided with kawata communities by Takeda
Shingen (1521-73). As Banba Masatomo points out, leather workers in these towns may have suffered not only the prejudice
associated with their "polluted occupation" but also that directed against outsiders and newcomers.[72]

Documents drawn up on the occasion of transfers of lords in Shinano in the 1580s and 1590s refer to already existing service
relations between kawaya sogashira[*] or kawata toryo[*] (heads or leaders of leather workers) and lords whom they provided
with leather goods and for whom they also performed policing duties such as guarding prisons,

[70] Teraki, Kinsei buraku , 25-26.

[71] Banba, "'Buraku' no keisei," 9-11.

[72] Ibid., 21-23.

― 280 ―

performing executions, cleaning castle grounds, checking the entering and exiting travelers at way stations, and so on.[73]
These leaders were probably chosen from among the more prominent leather workers in the area. In order to perform the
function of their occupation, these leaders had been moved to the castle towns and been put in charge of other kawaya or
kawata. They were granted the privilege of wearing swords, received tax-free land (sometimes even called chigyo[*] "fiefs"),
and at times received stipends or collected "one sheaf" (monnami hitomasu, momiko, hitowa ine ) per household from the
villages they served. Their names became hereditary occupational names, like Danzaemon in Edo; Magoroku in Matsushiro,
controlling four districts for a stipend of sixty bales; Hikodayu[*] in Matsumoto, in charge of two districts; Kojuro[*] in Ueda.
They formed the elite corps of the kawata, married into one another's lineages (Hikodayu[*] was related through marriage to
Danzaemon), lived in larger houses than their subjects, and enjoyed respect as officials. For instance, when Magoroku traveled
on business to Edo he stayed in the daimyo's nagaya , the long house where samurai were lodged.[74]

Within kawata communities status divisions and concomitant tensions and power struggles developed along the same lines as
in peasant villages. By the nineteenth century leadership positions had become hereditary, the number of main families had
been frozen, and lawsuits initiated by kawata underlings (tega) concerning succession to office

[73] The information that follows comes from ibid., 5-15. As a general rule, the execution of criminals was not associated with
kawata executioners and guards in medieval times. There is mention of an execution performed by a kawaramono in 1488
(Watanabe, Mikaiho[*]buraku , 139).

[74] Banba Masatomo, "'Buraku' no sui-i to mibun kisei no kyoka[*] : Shinshu[*] ni okeru burakushi sobyo[*] (3)," pt. 1, Shinano
13, no. 8 (1961): 33-34; idem, "Mibunsei no teppai e," 35. On the size of the homesteads (yashiki) in the Kamasu buraku, Komoro
domain, in Kita-Saku, see Ozaki, Shinshu[*]hisabetsu buraku , 61-63. Danzaemon's yashiki in Edo was about 1,000 tsubo (3,300
square meters), and he is said to have drawn income from 3,000 koku; other eta/hinin heads are said to have held 300,000 to
400,000 ryo[*] , and a drum maker from Watanabe hamlet near Osaka is said to have held over 700,000 ryo[*] (Kobayashi
Shigeru, "Kinsei ni okeru Buraku kaiho[*] toso[*] ," Rekishi koron[*] 3, no. 6 [1977]: 90-91; Ninomiya, "Inquiry," 103). Danzaemon
in Edo was allowed to posture as a small daimyo on certain official occasions. For instance, when he paid New Year's visits m
senior and junior counselors and to the three commissioners, he was carried in a sedan chair and accompanied by spear
bearers (Mase, "Ishiki no naka no mibunsei," 270).

― 281 ―

and abuses of authority were widespread.[75] It will be recalled that the domainwide struggle of 1856 in Okayama was forced
upon the kawata leadership by their subordinates. In some of these disputes, shares in the privilege of patrolling villages
(dannaba kabu) were at issue. These shares were controlled by the main houses, which were the equivalent of titled peasants
in kawata communities. In Kamasu, for instance, only 30 percent of the households enjoyed such shares, which, as in peasant
villages, could be transmitted or sold.[76]

After leather workers were forced to relocate to the castle towns, they were often moved to the outskirts as the towns'
population expanded (this occurred, for example, in Kaizu, Sunpu, Odawara, and Kanazawa). The reasons were in part
functional, for their work as prison guards or executioners or post attendants for checking travelers made it necessary for
them to resettle outside the town; in part economic, the outskirts being inferior locations; and part cultural, having to do with
notions of pollution.[77] In late medieval times similar separate communities had formed around market and temple towns,
where a floating but semistable population of entertainers, tatami makers, and others had settled.

In the population rosters and land cadasters of the late sixteenth century many people were registered as "kawata," a
reference to the leather workers or their fields.[78] Sometimes whole communities were registered as such, separate but not
independent from peasant villages. Asao Naohiro stresses that on Hideyoshi's land survey, no kawata villages were registered
as autonomous units; they were always appended

[75] For details of internal governance of kawata communities, see Banba, "'Buraku' no sui-i," pt. 1, 34-36; and Ozaki's study of
Kamasu buraku in his Shinshu[*] hisabetsu buraku , 75-76. For intrahamlet struggles in Shinano communities, see ibid., 20, 100-
150, 296-99; Banba, "Mibunsei no teppai e," 36-39; and idem, "Ueda-ryo[*] nai ni okeru 'eta' sosho[*] jiken," Shinano 17, no. 11
(1965): 1-12, and 18, no. 3 (1966): 30-48.

[76] Ozaki, shinshu[*]hisabetsu buraku , 85. Igeta notes that over the years some households succeeded m accumulating quite a
large number of such shares, which did not automatically translate into leadership positions (Igeta, "Kuchi-Tanba," 107).

[77] In Hozu village, for instance (see chapter 4), forty-two kawata households (including 250 people and 12 cattle) were
relocated, not without some monetary remuneration, from the center of the village to a lowland closer to the Hozu River,
which was exposed to frequent flooding (Igeta, "Mikaiho[*] buraku," 112).

[78] For this paragraph, see Teraki, Kinsei buraku , 35, 36, 86.

― 282 ―

to peasant villages.[79] Sometimes the entries were hierarchized, as when kawata were listed after peasants but before aruki
(messengers and jobbers on the payrolls of villages and towns). However, many were listed with their own kokudaka, some as
owners of as much as twenty koku, although such entries are only found in rosters and cadasters west of Shinano. It is not
quite clear whether these entries identified an occupation or what came to be their fixed hereditary legal status. One thing is
certain: these entries were not discriminatory as such; the households thus identified were not subject to special
discriminatory laws, and the kawata had not yet been organized "nationally," for this happened in the 1720s.

Only very rarely were leather workers entered on early Tokugawa records as "eta" (in Kumamoto in 1604, Awa in 1623). In the
1660s and 1670s, however, around the time when a new class of economically independent peasants was vying for political
power in the villages, a great number of domains began to record kawata as eta. It is not clear to what extent the link between
an increased competition among commoners and a stepped-up discrimination against kawata can actually be documented. At
the same time, in some provinces around Kyoto, communities of kawata were relocated to less desirable locations that often
were plagued by flooding or other natural disasters.[80]

Throughout the seventeenth century the bakufu and the various domains were engaged in haphazard efforts to regulate and
order the marginal populations engaged in numerous "lowly occupations." Thus, for instance, the dyers of Kyoto, initially
labeled eta, succeeded in being recategorized separately; in Omi province, however, dyers were organized into a kawata
community.[81] Hunters also succeeded in separating themselves from the kawata. The expanding area of Danzaemon's
control over the kawata, all of eastern Japan by the 1720s, and the extension of his jurisdiction over hinin and entertainers
culminated in the recording of kawata (and Danzaemon's) genealogies in the early eighteenth century, which testifies to this
ongoing effort at regulating the nonpeasant lower population strata. Yet judging from the frequent queries by various local
authorities in other parts of Japan concerning

[79] Asao, "Kinsei no mibun," 36.

[80] Teraki, Kinsei buraku , 45.

[81] Ibid., 189; NSSS 14:267n.

― 283 ―

the locus of authority over kawata or hinin, it is evident that the situation was far from clear.[82]

Starting in 1678, the one kawata household of Kodaira village (Kita-Saku) was entered as eta in the population register. Around
the same time, however, kawata in that area were also referred to as chori[*] ("officials"), often written with the homophonic
characters meaning "removed, separate from the town." Sometimes they were entered as yakunin ("officials") on the temple
rosters for memorial services (kakocho[*] ).

The practice of registering individuals in official records as members of a separate, defiled category of human beings came
also to prevail in Buddhist temples around the end of the seventeenth century, whereby discrimination was extended into the
afterlife. At Fukuoji[*] in Kodaira, for example, the pre-eighteenth-century term yakunin had been replaced by 1725 by eta .
Three decades later discrimination continued to grow: the names of deceased temple members who were kawata were now
put on a list kept separate from the list of deceased peasant members. Moreover, posthumous names (kaimyo ) with the
additional character kaku ("tanned leather," also read as kawa , "hide," "pelt") were bestowed on kawata.[83]
Customarily, posthumous names were honorific titles that ennobled the deceased to varying degrees, depending on the sum of
money paid by the family to their Buddhist temple. In the early nineteenth century the bakufu was sufficiently disturbed
about the inappropriately high ranks commoners were being granted in the afterlife to pass legislation

[82] See the numerous requests for clarifying questions regarding penal jurisdiction over kawata and hinin in Kobayashi,
Kinsei horeishu[*] . Igeta discusses some of these requests regarding mixed marriages and incognito kawata servants in his
"Kon'in sabetsu." See also Tsukada Takashi's interesting article on various suits by conflicting lower occupational groups that
resulted in officializing the hierarchical relationships between them ("Kasomin[*] no sekai," 225-68).

[83] On the numerous ways posthumous names on tombstones identified the dead as separate beings, sometimes as
"humanoids," using the character meaning "to look like [humans]," sometimes even as beasts, chiku , see Kobayashi Daiji,
Sabetsu kaimyo[*] . Much of this study is concentrated on Ueda city and Chiisagata district, adjoining Kita-Saku district in
Nagano prefecture. A temple in Nagano is said to have charged 3 million yen for converting a discriminatory posthumous
name into a regular one (15). The oldest discriminatory posthumous name Kobayashi mentions dates from 1622 (180).

― 284 ―

on the proper posthumous status for kaimyo.[84] The bakufu, however, did not have to worry about the kawata: the Buddhist
derby made sure they maintained their assigned status even in the afterlife. Discriminatory posthumous names were chiseled
on tombstones, allowing for the immediate identification of kawata graves, a practice that continued well into the mid
twentieth century, mainly in eastern Japan, where the tiny kawata communities often share graveyards with majority
Japanese. Kobayashi Daiji reports a case of discriminatory posthumous names as recently as 1980.[85] The Soto[*] branch of
Zen Buddhism in Nagano also had the term sendara in its posthumous discriminatory repertory for kawata, sendata being the
Japanese transcription in characters of the Sanskrit candãala , the name of an outcaste group in India.[86] In medieval texts
sendara had been associated with butchers, labeled evil people (akunin ).[87] Thus did mid-Tokugawa Buddhist monks
recognize and sanction discriminatory practices against the Japanese kawata as similar to the casting out of India's
"untouchables," long before modern scholars recognized the similarity.

The logic of discrimination penetrated the religious domain in other ways as well. Since families usually did not change
temples, modern data on burakumin temple affiliations reflect quite accurately the Tokugawa situation. In regions scattered
with tiny settlements, kawata shared temples with other commoners. Even then, they often went to special temples that served
only kawata (etadera ). A nationwide survey conducted in 1968 of all 1,470 etadera revealed that 91 percent were New Pure
Land (Jodoshinshu[*] ) temples. These "polluted" temples were orga-

[84] For a law dating from 1831, see Narusawa Eiju[*] , "Rekishiteki ni mita mikaiho[*] buraku no kaimyo[*] ," in Shukyo[*]to
buraku mondai , ed. Buraku mondai kenkyujo[*] (Kyoto: Buraku mondai kenkyujo[*] , 1982), 17o-71.

[85] Kobayashi Daiji, Sabetsu kaimyo[*] , 15.

[86] Ozaki, Shinshu[*]hisabetsu buraku , 26-33; see also 329-35. For information on sendara and candãla , see Teraki, Kinsei
buraku , 56. For the twentieth century, see Narusawa, "Mikaiho[*] buraku no kaimyo," 158-87; and Wakamiya Yoshinobu,
"Hisabetsu buraku ni totte no kami to hotoke: Komoro-shi Arabori de no minzoku chosa[*] o chushin[*] ni," Nihon shukyo[*]to
buraku sabetsu , special issue of Dento[*]to gendai , no. 73 (1981): 79- 112. I discuss the question briefly in the context of
ancestor worship in my Sosensuhai[*]no shimborizumu (Kobundo[*] , 1987), 171-72, 179-80.

[87] For example, in the Chiribukuro , a mid-thirteenth-century dictionary; however, m an expanded version of the mid
fifteenth century, the Ainosho[*] , the kawata are associated with pollution (Kuroda Toshio, Nihon chusei[*]no kokka to shukyo[*]
[Iwanami shoten, 1975], 380).

― 285 ―

nized separately into a specialized institutional hierarchy in direct connection, not with the sect's head temple in Kyoto
(Honganji), but with midlevel head temples (chuhonzan[*] ) specifically for the kawata.[88]

There is a striking regional pattern to these temple affiliations. Almost all burakumin temples west of the central region are
New Pure Land temples; only in Okayama and Takayama does one find temples affiliated with other Buddhist sects: Nichiren
and Pure Land. East of Shinano (in the Tokai[*] , Kanto, and northeast regions) one finds a variety of Buddhist sects serving
burakumin, including the Nichiren, Rinsai, Soto[*] , Tendai, and Shingon sects. Moreover, fully 80 percent of the New Pure
Land temples belonged to the Nishi Honganji branch, which claimed descent from the original line of the founder Shinran
(1173-1262) after Tokugawa Ieyasu split up the sect in 1602. Ninety percent of these temples affiliated with Nishi Honganji are
located in the Kinai area around Kyoto; there is not a single one in the central region or to the east of it.

Shinran's populist, egalitarian message appealed to the kawata.[89] Because kawata had been concentrated in the Kinai, where
the most radical followers of Shinran, the Ikko[*] sect, had organized armed resistance (ikki) against warlords in the sixteenth
century, before the Tokugawa period, Teraki Nobuaki tentatively linked kawata to the Ikko[*] ikki even though there is scant
evidence for kawata participation in these ikki.[90] The close association between Nishi Honganji temples and kawata in the
Kinai seems significant within the context of Teraki's argument, since a great number of these temples were established
during the heyday of the Ikko[*] movement.

[88] This and the following information on religious affiliation can be found in Teraki, Kinsei buraku , 55, 192-221.

[89] Shinran (1173-1262) identified his followers, all believers in the sole power of nenbutsu (reciting the name of Amida
Buddha), as human beings with the lowest karma: toko ("killers of dogs and sheep and sellers of sake"); and Nichiren (1222-82)
wrote that he was the son of a sendara (Kuroda, Nihon chusei[*]no kokka , 388).

[90] In eastern Japan, the Hakusan Shinto shrine in Kaga province is especially venerated by the burakumin. Hakusan seems
to have been the protective god of socially marginal members of society, since he is also seen as especially protective of
children (Shibata Michiko, Hisabetsu buraku no densho[*]to seikatsu [San'ichi shobo[*] , 1972], 17-24). Teraki mentions a close
link between the kawata communities and Hakusan belief in Shinano, and he writes of three thousand horsemen having been
dispatched by two large Hakusan shrines in Kaga to join the Kaga ikki in the sixteenth century (Kinsei buraku , 193).

― 286 ―

Toward a Legislated Racism

Just as Karl Marx argued that "the economic" constitutes the hidden or suppressed truth of past societies, now revealed to us
by the workings of capitalist society,[91] one can make a case that racism reveals the quintessential operations of the ascription
of any inherent inequalities, such as those that prevailed in a "society of orders" like Tokugawa Japan. Viewed from this
perspective, which is not that of the logician, racism is not a particular variety of discrimination; rather, all forms of
discrimination produce effects like those of racism. The Tokugawa history of today's burakumin points to practices and
ideologies that betray racist sentiment developed from what was originally a status distinction.

We have seen how asymmetrical punishments meted out for kawata and peasants were following the 1843 Clog Thongs Riot.
By the standards of late Tokugawa juridical practice, this was not unusual. In 1796 the bakufu's superintendent of finances,
who had a seat at the Tribunal (Hyojosho[*] ), answered a query whether or not commissioners should conduct direct
investigations of eta/hinin after preliminary interrogations by lower officials. The superintendent's answer was that there
should be no distinctions in this regard between ordinary commoners (heijin ) and eta-hinin. But, he added, their sentences
should be different.[92] In 1839 it was decided that when peasants and kawata filed joint suits in court, they should be treated
differently (sabetsu , discrimination): peasants should squat on straw mats, kawata on the gravel, or three shaku (almost one
meter) lower.[93] Clearly this could be construed as an official strategy to drive a status wedge into possible solidarities among
commoner groups.

Differential punishments for samurai and commoners were routine throughout the Tokugawa period, but the singling out of
kawata around the turn of the nineteenth century was something altogether new. The first Tokugawa segregation law for "eta,
hinin, chasen , and

[91] Karl Marx, Grundrisse (New York: Vintage Books, 1973), 460. This thesis has been reformulated by Bourdieu, who argues
for the pervasiveness of "interests" even in the symbolic realm; Foucault has substituted power for Marx's economic interests.

[92] Kobayashi, Kinsei horeishu[*] , 217.

[93] Ibid., 337.

― 287 ―

the like" dates from 1778.[94] It expresses concern with the lawless behavior of eta, hinin, and chasen and with their "mingling
incognito with peasants or townspeople at inns, eateries, drinking establishments, etc." This law was followed in 1804 by the
first listing of special penalties for kawata and hinin.[95] The latter document refers to precedents going back to 1771, but not
earlier.

Bakufu regulations before the end of the eighteenth century mentioned kawata but did not legislate discrimination and
segregation in a comprehensive manner. In this the bakufu followed pre-Tokugawa practice. For example, a kawaramono who
in 1518 assaulted a silk merchant did not receive particularly harsh punishment from the Muromachi bakufu court.[96] Where
official prejudice existed, however, it was manifested in various ways. Teraki Nobuaki mentions a common practice dating
from the early Tokugawa period, the cartographic custom of leaving out figures for road sections passing through kawata
hamlets when computing distances between two points.[97] If maps are representations of the land and its occupants, this
practice may have implied that the kawata did not belong in Japan. Similarly, a decree of 1763 required that hinin be removed
from the road to Tokugawa Ieyasu's shrine in Nikko[*] when the Korean embassy passed through.[98] The image of Japan that
was visible on maps (as seen through the eyes of officials) and from sedan chairs (through the eyes of foreign dignitaries) did
not include kawata or hinin.

In reality, of course, kawata were far from being ignored by the authorities. The first reference to kawata in Tokugawa edicts
dates from 1657. They were mentioned only in passing, though, in an addendum to the fourth article of a nine-point directire
regarding the catching of thieves. Kawata figured in a list of socially undesirable elements "one should look out for," the list
subsequently adopted in goningumi village

[94] Ibid., 151-52; TKKz 5:474-75 (no. 3434). In some domains, chasen (one of the many low occupational groups) were
distinguished from kawata; they performed similar functions as guards or at funerals under the direction of kawata leaders
(Teraki, Kinsei buraku , 48; KDJ 9:456).

[95] Watanabe, Mikaiho[*] buraku , 6, 155.

[96] KDJ 3:758, s.v. "kawaramono."

[97] Teraki, Kinsei buraku , 44. This cartographic practice was abolished by law on April 2, 1869 (Ninomiya, "Inquiry," 107).

[98] Kobayashi, Kinsei horeishu[*] , 136.

 ― 288 ―

laws: "Buddhist monks, yamabushi (mountain ascetics), wandering ascetics, mendicant monks with flutes and bells, eta, hinin,
etc."[99] Large cities conducted investigations and introduced reforms (aratame ) and passed administrative measures for
hinin (Osaka, 1644; Kyoto, 1654; Edo, 1674) and kawata (Kyoto, 1715; Edo, 1719).[100] These were efforts to tighten control over
the floating population of senmin, who were neither samurai nor peasants, craftsmen nor merchants.

By the end of the seventeenth century the prevailing practice was to list kawata separately from peasants, sometimes in
completely different population rosters. In Matsumoto domain (Shinano) beginning in 1722 kawata and hinin were even listed
under the separate heading jingai ("outside humans"), written with the same characters, in reverse order, as today's pejorative
term gaijin (foreigners).[101] Starting with Danzaemon in 1719 and continuing into the 1730s, the genealogies of kawata were
being checked in Kyoto, Edo, and a number of domains,[102] apparently to build a data base around the membership of these
classes. In later years these data were used in so-called eta hunts (etagari ) in Edo, Osaka (1798, 1827, 1833), and Kyoto.[103]
One such etagari conducted in Kyoto in 1831 yielded fifty-three kawata who were trying to "pass." In 1722 Danzaemon was
established as a well-pedigreed "national" chief for the hinin and kawata in the eastern provinces (the eight Kanto provinces;
Izu, Suruga, and Kai to the west of them; and the northeastern provinces).[104] In order to establish a clear hierarchy between
hinin and kawata, the former were required to cut off their topknots and were forbidden to wear headgear.

The 1720s and 1730s constitute a watershed for segregation legislation aimed at kawata in many domains, perhaps prompted
by the bakufu's retrenchment reforms collectively known as the Kyoho[*] Reforms.[105]

[99] Ibid., 102. See article 1 of the goningumi rules (1662) of Shimo-Sakurai, Kita-Saku district, in appendix 3; cf. articles z and 3
of the goningumi rules (1640) of the same village in appendix 2.

[100] Teraki, Kinsei buraku , 44, 46.

[101] Banba, "'Buraku' no seiritsu," 56.

[102] Watanabe, Mikaiho[*] buraku , 108, 138.

[103] Kobayashi Shigeru, "Buraku kaiho[*] toso[*] ," 91.

[104] For details of the national governance of the kawata/hinin, see Cornell, "Caste Patron," 56-70; and Ninomiya, "Inquiry,"
99-100.

[105] For an excellent, detailed discussion of these reforms, and the only one available in English, see Kate Nakai, "Kyoho[*]
Reforms," in the Kodansha Encyclopedia of Japan 4:330-31.

― 289 ―

This institutionalization of discriminatory practices against kawata and hinin gathered momentum with each subsequent
reform (Kansei, 1780-1790s; Tenpo, 1830s), eventually culminating in the intra-race racism of the nineteenth century. Earlier
some domains had issued status legislation for the kawata: Awa domain had decreed in 1699 that kawata should dress more
coarsely than peasants, and Choshu[*] domain in 1813 had decreed that "eta, chasen, and the like" should not be extravagant in
their clothing or home design and should not engage in occupations other than those authorized by the domain.[106] During
the period of the Kyoho[*] Reforms a number of domains made a concerted effort to restrict social intercourse between kawata
and other commoners. The domains of Matsumoto, Komoro, Iwamurata, and Okudono (Shinano, the latter three in Kita-Saku)
all issued legislation for kawata in the year 1738.[107] In 1704 the Okudono domain (the location of the Bonboku Incident of
1771) acquired additional villages in Kita-Saku (around Tanoguchi) as a tobichi , or detached parcel. On that occasion the
domain issued a comprehensive law comprising eighty articles, but none of them directly concerned the kawata. In 1738,
however, the domain issued a seven-point directive aimed specifically at the kawata population. Reissued again in 1779 and
1809, this law constituted the domain's policy toward the kawata.

It is worth noting that in 1704 there were only eleven kawata households in the domain, spread over five (seven in 1738) of the
twenty-five villages. By 1838 the kawata population, in ten villages, still numbered only 107 members. In 1765 in Komoro
domain (15,000 koku), with a population of approximately the same size, the kawata population was 438 in ten villages, half of
it concentrated in Kamasu, the largest kawata community in both Komoro domain and Shinano province.[108] How could such
a minuscule "minority" population possibly represent a threat? What motivated domain officials to implement discriminatory
and segregationist directives against the kawata in 1738? Judging from the regulations of that year and later, when they were
reissued, the concerns seem to have been the following.[109]

[106] Teraki, Kinsei buraku , 45; another example is Toyoura domain in 1683 (Kobayashi, Kinsei horeishu[*] , 103-4).

[107] Ozaki, Shinshu[*] hisabetsu buraku , 17, 72-74, 243-44.

[108] Ibid., 52

[109] These rules, all from the Kita-Saku area, can be found in the following publications: for Okudono and Komoro in 1738,
ibid., 75-74, 244-45; for Chiisagata district in 1819 and 1838, ibid., 232-26; and for Iwamurata in 1783, Banba, "'Buraku' no sui-i,"
pt. 2, Shinano 16, no. 9 (1964): 29.

― 290 ―
Standards of segregation, which were perceived as breaking down, were the main concern: the kawata were behaving too
much like ordinary peasants. Concretely this involved reinstituting status-marking practices and creating new ones. Practices
signifying status were fairly consistent throughout the area and the country, with only a few local variations. Laws
consistently stated and restated that kawata could wear only straw sandals; wooden clogs were forbidden, as was the use of
umbrellas or parasols; and clothing with family crests and haori jackets were proscribed. Kawata could not build new-style
houses and were forbidden to hang streamers (nobori ) from poles or roofs during village festivals.

These laws set kawata clearly apart from other peasants, although, as already mentioned, in certain places, such as Hozu
village, titled peasants directed very similar taboos against other peasants. Social interaction was regulated and reduced to a
minimum, usually only when official business was involved. As a rule, kawata were instructed always to be courteous and
deferential toward peasants. Kawata could wear their short swords only when they were on official business, such as guard
duty, and certainly could not parade about with them in other villages, posturing as officials. Even for one-day trips they
needed permission from the village headman. At plays or entertainments they were to appear only in the line of duty, in
accordance with instructions from the local headman, but not among the audience or as participants. Nor could they be seen
at teahouses or drinking establishments. When going to the fields or mountains, they had to take back roads. They were
forbidden to enter peasants' homes or even commoners' shops except on official business. Finally, they were not permitted to
mingle with ordinary people (heijin), a regulation that led, as noted earlier, to a practical prohibition of mixed marriages. (For
a translation of these and other regulations, see appendix 5.)

It was essential that these Japanese be identifiable as "other" at all times. While ultimately ancestry became the criterion for
determining who were kawata, more visible signs were needed to differentiate them. In Tosa, persons of kawata status could
not be in the streets after 4:00

― 291 ―

P.M. (1780).[110] In Matsushiro domain, which had the largest kawata population in Shinano (380 households with a total
population of 2,450 in 1869),[111] all kawata were required to carry a lantern so that they could be identified at night (1841).
[112] As we saw in the Okayama uprising in 1856, they had to wear special clothing; elsewhere they had to wear pelts or pieces
of animal fur.

Since kawata hamlets were set apart from society, fugitives sought refuge there. Others were lured to kawata hamlets by the
opportunities of the sometimes lucrative leather trade.[113] That kawata hamlets were "hotbeds of criminals," as one
Confucian scholar put it, is sheer derision. On the other hand, certain legal decisions criminalized the hinin and kawata
population as a whole. In Osaka hinin status had become criminalized in 1680 by virtue of a decree stating that anyone buying
clothes from hinin would be treated as a thief, the assumption being that hinin, already presumed to be criminals, peddled
only stolen goods.[114] A new twist in the law regarding disinheritance (kando[*] , kyuri[*] ) had a similar effect with regard to
the kawata. Disinheritance of an individual, it will be recalled, precluded the threat of punishment by vicarious guilt that
made social groups (family, neighborhood, village) legally responsible and punishable for crimes committed by persons who
had absconded. When a member fled the village, only after several active searches had failed to produce the fugitive could
disinheritance be implemented. In the late eighteenth century numerous laws denied kawata communities any such
protective clauses. If kawata absconded, a nagatazune ("long search") was ordered that did not end after a certain period, in
other words, a search in perpetuity. Thus kawata communities were always responsible for crimes committed by their
wayward members and held punishable for them.[115]

Popular views of pollution and its association with certain occupa-

[110] Teraki, Kinsei buraku , 45.

[111] Banba, "Mibunsei no teppai e," 41.

[112] Banba, "'Buraku' no sui-i," pt. z, 31.

[113] Hatanaka documents the concentration of wealth in some kawata communities and in-migration there by non-kawata
("'Kawata' mibun," 308-9, 328-35).

[114] Kobayashi, Kinsei horeishu[*] , 103.

[115] For such bakufu laws issued in 1781, 1792, and 1802, see ibid., 162, 189-90, 252-54; for the difference between kando[*]
and kyuri[*] , see ibid., 312.

― 292 ―

tions were officially sanctioned through legislation marking kawata as a visually identifiable class of people to be avoided.
This led not only to perceptions of kawata as lawbreakers and their communities as "hotbeds of criminals" but also to racist
views that naturalized the kawata as literally nonhuman. In Oshu[*] domain in Iyo (Shikoku) all male and female kawata over
the age of seven were obliged to wear a piece of animal fur measuring fifteen square centimeters at all times; each kawata
family also had to hang an animal pelt at the entrance to their home.[116] In Kaga in 1776, Matsumoto's term jingai was used in
a notification stating that "since kawata are originally different from humans, they should not be in places where humans
congregate."[117] As we shall see, a theory was developing that the kawata were indeed racially different from the Japanese.
This view of the kawata as belonging to the realm of beasts was also adopted by intellectuals such as Kaiho Seiryo[*] (1755-
1817), rationalist and mercantilist, who wrote that they "are like animals (kinju[*] dozen[*] ) ... there is no morality in their
heart (kokoro ni zen'aku no naki )."[118] But before turning to the question of racist ideology, we shall take a look at the
economic dimension of segregation.
Economic Suppression

Discriminatory regulations were not only segregationist. Some were explicitly intended to control the kawata's economic
status by either reducing them to poverty or keeping them there. Like peasants restricted to agriculture and required to
deliver tribute (in principle, in rice), the kawata were required to engage in leather works. Represented by the authorities as a
"privilege," this occupational restriction seems to have been fairly well established by the end of the seventeenth century.[119]
Choshu[*] domain reissued such restrictive laws in 1743, 1771, and 1836. There were laws strictly forbidding the kawata to buy
cows or horses and then slaughter them or to engage in other trades that ordinary commoners engaged in.[120] A bakufu edict
of 1612 had already for-

[116] Teraki, Kinsei buraku , 59.

[117] Ibid., 45.

[118] Ibid., 60.

[119] Watanabe, Mikaiho[*] buraku , 104.

[120] Kobayashi, Kinsei horeishu[*] , 127, 142, 334.

― 293 ―

bidden the killing of cattle, and in 1614 a kawata was crucified in Matsumoto because he had broken a similar law.[121] This
prohibition was not new, however, since laws forbidding the slaughtering of horses and cattle had been issued some nine
hundred years earlier, in 676, and again several times in the first half of the eighth century.[122] In Toyota village, reportedly, if
common peasants removed dead horses, they would be considered kawata; elsewhere it was specified that the removal had to
occur only at night.[123] In principle, the owner of a cow or horse lost ownership over it when it died; the carcass, taken to a
designated place in the village, was taken possession of by kawata.[124] Nevertheless, a trade in live cattle for slaughter and
consumption most likely existed anyway, since investigations were conducted, sometimes across whole provinces, into the
slaughtering of cows and horses (e.g., in Bizen, Bitchu[*] , and Omi[*] in 1734).[125]

Status systems function to minimize competition by automatically excluding all holders of other ranks from the benefits or
entitlements associated with a particular rank. This would apply also to the situation of the kawata: they held a monopoly,
guaranteed by overlord authority, in leather working but were excluded from other trades. To explain the absence of kawata
in northeastern Japan, Harada Tomohiko hypothesizes a link between increased discrimination against the kawata and
increased competition among commoners generated by an expanding commercial economy. The commercial economy was
weak in the northeast, and discrimination against kawata was absent there, as were the kawata, but discrimination against
kawata was strong in the more commercialized regions where the kawata were concentrated.[126] Ian Neary suggests that the
Emancipation Decree of 1871, which certainly did not result in equality, had the devastating effect for the kawata (perhaps
intended) of bringing commoners in direct competition with kawata for

[121] Ibid., 99; Banba, "'Buraku' no seiritsu," 55 n. 3.

[122] Ninomiya, "Inquiry," 55, 78.

[123] Kobayashi, Kinsei horeishu[*] , 498.

[124] Teraki, Kinsei buraku , 49.

[125] Kobayashi, Kinsei horeishu[*] , 124-25. Recent archeological excavations indicate the widespread consumption of animal
meat, although not beef, in Edo (see Uchiyama Junzo[*] , "San'ei-cho[*] and Meat-eating in Buddhist Edo," Japanese Journal of
Religious Studies 19, nos. 2-3 [1992]: 302).

[126] Harada Tomohiko, Buraku sabetsushi kenkyu[*] (Kyoto: Shibunkaku, 1985), 230-31. For a cultural explanation, see above,
n. 49.

― 294 ―

Table 22.
Kamasu Buraku/Village: Population and Kokudaka Compared, 1723 and
1868

1723 1868 % Change B/V %

Difference

Kokudaka (koku) B 45 29 (-38)

V 108 108 (-0) 38

Population B 120 237 (+100)

V 155 207 (+33) 67

Households (ie) B 42 49 (+16)

V 32 46 (+43) 27

Taka holders B 27 30 (-3)

V 23 36 (+6) 9

Average holder B 1.7 1.0 (-42)

(koku)

V 4.7 4.0 (-15) 27

Average per ie B 1.1 0.6 (-46)

(koku)

V 3.4 2.4 (-23) 23

SOURCE : Ozaki, Shinshu[*] hisabetsu buraku , tables 5, 7, and 8 (pp. 66, 89,
and 91).

modern businesses such as butcher shops or shoe manufacturing, which logically would have been the kawata's prerogatives.
[127]

Competition for scarce economic goods was reflected by other restrictions. Iwamurata domain prohibited kawata from
increasing their land holdings, and peasants were forbidden to sell them land. Fields owned by kawata were often of low
quality and small, and access to water was often difficult to obtain. Thus, fields were mostly used only as a limited source of
food and supplementary income, especially later in the Tokugawa period. The data for Kamasu, the largest kawata community
in Shinano (located outside Komoro, Kita-Saku), illustrate this well (see table 22).[128]

Kamasu's kawata population doubled from 120 to 237 persons (42 to 49 households) between 1723 and 1868, but the holdings
decreased from a total of 45 koku to 29. The average holding of taka holders (64 percent of all households in 1723, 61 percent
in 1868) dropped from

[127] Neary, Political Protest , 32.

[128] Ozaki, Shinshu[*] hisabetsu buraku , 66, 89, 91.

― 295 ―

1.7 to x koku, which, averaged out over all households in the buraku (taka holders and nonholders), amounted to a drop from
1.1 to 0.6 koku per household.

In contrast, the figures for the Kamasu village peasant population are higher than the kawata figures at the beginning (155)
and lower at the end (207), marking an increase of only 33 percent. (The kokudaka remained the same, 108 koku; the number
of households increased from 32 to 46.) The average holding of taka holders (72 percent and 78 percent of all households)
dropped from 4.7 to 4 koku, which, averaged out over all households in the village, amounted to a drop from 3.4 to 2.4 koku
per household. It should also be noted that in the early nineteenth century the Kamasu kawata were strictly forbidden to
purchase land or even mortgage their own.[129]

Kamasu's population growth was not exceptional. The dramatic universal increase in the number of kawata in comparison
with the rest of the population, which stabilized after the 1720s, has long been noted by historians. The population of the
Watanabe kawata community near Osaka increased fivefold between 1692 and 1856 (from 840 to 4,257), while the population
of Osaka declined by 7 percent (to 320,780) during the same period.[130] In Ueda domain (Shinano) the kawata population
increased from a total of 329 (70 households) in 1726 to 893 (110 households) in 1868.[131]
Historians have offered several explanations for this phenomenon, ascribing to it a combination of outside factors (the influx
of drifters) and natural growth (facilitated by strong religious beliefs against infanticide), diet (kawata ate meat), and the
growth of a market economy from which the kawata would have benefited (the price of leather went up, keeping pace with
other rising prices).[132]

[129] Ibid., 94.

[130] Teraki, Kinsei buraku , 54.

[131] Banba, "'Buraku' no sui-i," pt. 2, 16-17. The kawata population continued to increase well into the twentieth century.
During the first sixty years after the Meiji Restoration (1868) the burakumin population increased by 400 percent, while the
general population increased by only 80 percent (Ninomiya, "Inquiry," 114 n. 15). For a population diagram of three villages
and their attached buraku between 1725 and 1875, see Hatanaka, "'Kawata' mibun," 313.

[132] Steer hides fetched twice as high a price as horse hides, with cow hides in between. The price of steer hides increased
from 0.700 to 1 kanmon (both dates unknown), then to 1.600 in 1781, 1.800 in 1812, and 1.900 in 1821 (Banba, "'Buraku' no
seiritsu," 54-55 n. 2). Using tables m convert copper coins into silver value (Nihonshi jiten [kaitei zoho[*] ], ed. Kyoto Daigaku
bungakubu kokushi kenkyushitsu[*] [Sogensha[*] , 1960], 845) at exchange rates prevalent in Edo and price lists for Higo rice to
be found on p. x of the appendix to any of the thirty-five volumes of Nihon nosho[*] zenshu[*] (Nosangyoson[*] bunka kyokai,
1977-81), one finds that the value of steer hides remained fairly constant, being the equivalent of 0.25 koku of rice. Teraki
(Kinsei buraku , 50) puts the value at 1 koku for the mid Tokugawa period.

― 296 ―

As the figures for Kamasu and Ueda indicate, while the population increased dramatically, the number of households did not.
This was probably because of an attempt by the authorities to stem population increase by limiting the number of households.
In the first bakufu legislation that dealt comprehensively with the kawata (1778), their "illegal (hogai[*] ) population increase"
was noted.[133] A decree of 1834 stated that "the number of kawata is increasing disproportionately in relation to that of the
peasants. Someday this will become a source of problems. Henceforward only one male per household can take a bride. The
other children must marry out. We will try to decrease the population one by one. Permission to marry must be secured from
the appropriate officials."[134]

How widespread such population control measures were is not known. We do know, however, that one way of dealing with
the undesirable growth of the local kawata population was to relocate them to other villages.[135] As we saw earlier, Umaji
village, near Hozu, imported a whole kawata community in 1801 to work fields tenant peasants were trying to boycott. Some
daimyo even had grandiose plans to utilize the kawata on a national scale to colonize Ezo (Hokkaido). Already by 1720 Arai
Hakuseki, a retired shogunal adviser, had sorted out the available data on the island in his Ezoshi . Thus, interest in Ezo's
economic potential antedates by several decades Tanuma Okitsugu's famous plan in the 1780s, based on a recent exploratory
survey, to colonize the island with kawata.[136]

[133] Kobayashi, Kinsei horeishu[*] , 151.

[134] Banba, "'Buraku' no sui-i," pt. 1, 56.

[135] Ibid., pt. 2, 26.

[136] Other books, less trustworthy in detail than Hakuseki's, appeared around the same time: Matsumiya Kanzan's Ezodan
hikki and Itakura Genjiro's[*]Hokkai zuihitsu (KDJ 2:271, s.v. "Ezoshi"). For Tanuma, see John Whitney Hall, Tanuma Okitsugu,
1719-1788: Forerunner of Modem Japan (Cambridge: Harvard University Press, 1955), 67, 100-105.

― 297 ―

The kawata had become incorporated in an emerging "colonial" discourse in two ways. Through a fabricated etymological
relationship between Ezo and eta the kawata were turned into aliens, literally alienated. As alleged descendants of captives
from past wars with the aborigines of Ezo, the kawata were being signified by Tani Shinzan in 1717, among others, as
genealogically non-Japanese linked to a non-Japan, Ezo.[137] Moreover, when the possibility of colonizing Ezo was actually
being entertained, the proposal was made to utilize the kawata for this purpose on a grand scale. Thus in 1786/2 Matsumoto
Hidemochi, the superintendent of finances, presented a scheme to Tanuma Okitsugu, the bakufu's chief political leader, for
turning Hokkaido into one enormous shinden (new rice paddy):

Because the native Ebisu would not be enough for this development project, we have conceived a plan to gather eta
from various provinces and move them to Ezo. When we asked Danzaemon, without revealing the location of the
project, whether he had ever thought of utilizing some of his subjects (tega) to develop new fields, he said that of the
33,000 eta in his jurisdiction, 7,000 eta/hinin could be made available for such purposes. This is not enough, but
Danzaemon's authority, of course, does not extend to all the provinces. However, because he has on occasion issued
rules to the various regional eta leaders, we could put him in charge of all the eta/hinin of the whole country, a
population of some 230,000. This would enable us to move 63 to 70 thousand eta to Ezo, including Danzaemon.[138]

Hoashi Banri (1778-1852), a Confucian eclectic, scholar, teacher, and, at one point, elder in the domain of Hiji in Bungo
province (Oita[*] , Kyushu[*] ), combined the idea of colonizing Ezo with kawata (who he considered to be descendants from
past captives from Ezo anyway) with that of liberating them first at a giant purification ceremony: "The ancestors of the eta are
captives from Ezo. And, although they are assigned the task of apprehending thieves, [their communities] are hotbeds of
crimes. So we should gather them all at the Great Shrine of Ise, purify them, make them heijin [commoners], send them to the
land of
[137] Morita Yoshinori, "Edo-ki ni okeru buraku e no shiteki kanshin," in Kinsei buraku no shiteki kenkyu[*] (1) , ed. Buraku
kaiho[*] kenkyujo[*] (Osaka: Kaiho[*] shuppansha, 1979), 1:317.

[138] Tsuji Zennosuke quotes a section of this proposal in his Tanuma jidai (Iwanami shoten, 1980), 300-302; for the whole text,
see Kobayashi, Kinsei horeishu[*] , 175-77.

― 298 ―

Ezo, and set them to work at agriculture or cattle breeding."[139] The "liberal" view of this Confucian betrays the depth of
prejudice against the kawata. They are viewed as less moral than ordinary people and as property of the state who, like
condemned criminals (with whom they were associated anyway), could be put to work to enrich the state. Interestingly,
Hokkaido was eventually colonized and developed in the Meiji period by inmates condemned to forced labor.

The Contaminated Intelligentsia

Tokugawa discourse on society, while moving in a number of directions and shifting idiom, developed over time on a template
provided by Confucianism. Whether traditional Confucian, neo-Confucian, anti-Confucian, eclectic, or Shinto, this discourse
was always in dialogue with Chinese learning. Prominent in this "Chinese" tradition as it was "translated" in Japan was the
view of society as divided into four classes: samurai, peasants, artisans, and merchants. It should be noted, however, that this
"translation" took some time. Banba Masatomo writes that the shift in categories from the Chinese "ministers, high officials,
gentlemen, commoners (ch'ing, tafu, shih, shujen )" to "samurai, peasants, artisans, merchants" took hold in Japan only around
the 1720s.[140] Moreover, the expression does not occur in the Tokugawa jikki , which uses senmin ("lowly people," referring to
the nonsamurai) or "peasants and merchants."[141]

Tokugawa scholars did not seriously confront the status and class system beyond these categories. Their "four-class" system,
more a gen-

[139] Quoted from Hoashi's Tosenfuron[*] by Teraki, Kinsei buraku , 62.

[140] Banba Masatomo, Nihon jukyoron[*] (Mikasa shobo[*] , 1939), 172-77, esp. 176. Watanabe mentions Banba's interpretation
(Mikaiho[*] buraku , 137-38). For a more detailed analysis, see Asao, "Kinsei no mibun," 14-24. It is interesting that Banba, who
after the war would pioneer new research into the Tokugawa past of burakumin, to my knowledge mentions eta only once in
his book (131). Apparently following censorship guidelines, the word eta is not spelled out fully. Instead, it is printed as xta , an
x replacing the e (meaning "dirt," "pollution"): even the character for "pollution" seems to have been taboo in those years. A
similar practice governed the publication of Tokugawa village laws. In Hozumi Shigeto's[*]Goningumi hokishu[*] zokuben , the
word eta is replaced by two blank squares (see, e.g., 1:97): the eta are erased just as their communities were expurgated from
Tokugawa maps. Perhaps inspired by concern for burakumin sensibilities, the practice nevertheless maintained these
communities as "the unnameables," the ones whose identity could not be spoken.

[141] Watanabe, Mikaiho[*] buraku , 137.

― 299 ―

eral metaphor for hierarchy than a sociological concept, did not encourage discussion of status differentials within these
classes, let alone of subjects like the kawata, who fell outside the system.

Occasional references to kawata were limited to incidental remarks. Texts on government policy and Confucian discussions of
jinsei (benevolent government) do not contain compassionate references to kawata. Kobayashi Shigeru, who compiled bakufu
directives related to kawata and hinin, found only one such reference, in the apocryphal Tokugawa seiken hyakkajo[*]
(Tokugawa Constitution in One Hundred Articles), most likely composed in the late eighteenth century, possibly by a Buddhist
priest.[142] The article reads: "Eta, beggars (hoito[*] ), blind men and women, and indigents who have no one to rely on
(tsuguru naki ), who are outside the four classes, have since the ancient past been treated with compassion so that they can
make a living. One should know that this is the beginning of benevolent government."

Other remarks expressive of a reasoned attitude include the following from early Tokugawa rulers and thinkers. Ikeda
Mitsumasa (1609-82), the famous lord of Okayama domain, was quoted as saying to a vassal whom he took to task for
prejudicial views: "Should we really consider the eta different? They are also peasants. Skinning of wild boars and badgers
and eating their meat—don't a lot of people do that (dare tote mo sumajiki ni mo arazu )? Why, then, do we look with repulsion
only at them?"[143] Whether or not he actually said this, it is perhaps significant that these words were ascribed to a meikun
("model lord"). Kumazawa Banzan (1619-91), the Confucian scholar in Mitsumasa's service and equally famous, expressed a
similar logic, undoubtedly shared by the kawata. In a passage critical of the Buddhist

[142] Kobayashi, Kinsei horeishu[*] , 127. Inexplicably, Kobayashi gives Kanpo 2 (1742) as the date for this one article, although
it is article 85 of the version of the Tokugawa seiken hyakkajo[*] (literally, the Tokugawa treasured legacy in one hundred
articles) reprinted in TKKz 1:59 and in Kinsei buke shiso[*] , Nihon shiso[*] taikei, 27 (Iwanami shoten, 1974), 474 (hereafter NST
27). For a full German translation of the three versions, see Rudorff, "Tokugawa-Gesetz-Sammlung," 4-21. Hiramatsu Yoshiro[*]
writes that the articles "appear to have originated in about the Kansei period (1789-1801), and it is strongly suspected that a
Buddhist priest had a hand in it. There is also doubt whether they even reflect the general Tokugawa legal consciousness"
("Tokugawa Law," trans. D.F. Henderson, Law in Japan 14 [1981]: 5, originally published as "Kinseiho[*] ," Kinsei 3, Iwanami
koza[*] Nihon rekishi, 11 [Iwanami, 1976], 332-78).

[143] Watanabe, Mikaiho[*] buraku , 156-57.

 ― 300 ―

clergy, he argued, "The eta are said to be polluted, but the Buddhist monks are even more so ... people are called 'eta' merely
for handling dead cows and horses. But aren't those who handle dead people (shijin ), consume meat, use clothes made from
animal skins (shihi , "dead skin"), and live next to hundreds of graves big eta (daieta )?"[144] This view was rarely expressed in
Tokugawa writings. Only in texts from near the very end of the period does one come across expressions of this view again. As
discriminatory practices against kawata were institutionalized by official records and legislation, correlative theories
developed to explain and validate this discrimination.

To justify segregation, Ogyu[*] Sorai (166-1727) does not appeal to the sages, as he did to rationalize the four-class system, but
to the "customs of the Divine Country," Japan: "Not sharing fire with kawata is a custom of the Divine Country and unavoidable
(shinkoku no fuzoku[*] , zehi nashi )."[145] The expression "not to share fire," referring to an avoidance of kawata in order not to
share their pollution, also appears in a letter circulated in 1748 by a bakufu intendant prohibiting the apparently widespread
custom of "sharing fire, etc.," with kawata.[146] As we have seen, such prohibitions against "mingling" with the kawata became
increasingly common in the second half of the eighteenth century: the "custom of the Divine Country" Sorai referred to was
being upheld by law; that is, the state became the carrier of "customs," while the people seemed to be more accommodating.
Ultimately, the authorities seem to have dreaded the reality of the kawata itself far less than the prospect of there being none.

On his official journey to the province of Kai in 1706 Sorai passed through a kawata community. In his official travelogue (the
Kyochukiko[*] ), he notes: "We pass through the village of Utsunoya. All are butchers. I wanted to smoke but could not ask for a
light." This is a reworked version of his private travelogue, Furyushishaki[*] , where the entry reads: "When we pass through
the village of Utsunoya, we see women huddled together in a boisterous clamor. Bamboo peels

[144] Teraki quotes this passage from Banzan's Miwa monogatari in Kinsei buraku , 60.

[145] Ogyu[*] Sorai, quoted from his Seidan by Teraki (ibid.). The quotation can be found in Ogyu[*] Sorai , Nihon shiso[*] taikei,
36 (Iwanami shoten), 286 (hereafter NST 36). For Sorai's justification of the four classes by appealing to the sages, see
Maruyama, Studies , 214, 217.

[146] Banba, "'Buraku' no sui-i," pt. 2, 16-34.

― 301 ―

are soaking in the ditches under the protruding eaves of the houses [to be used for making leather-soled sandals, setta ]. They
are houses of butchers. I feel a great urge to smoke, but refined emissaries should not break customs. I could not ask for a
light. I have a poem...." In this poem Sorai refers to age-old customs and to the ten categories for noble and base people that
prevailed in China. In his Seidan , Sorai alludes to the theory that the kawaramono are of a different stock: "Prostitutes and
kawaramono are considered lowly people; this is the same in China and Japan, both now and in the past. Because they are of a
different stock (genrai sono sujo[*]kakubetsu naru ), they are considered lowly and entrusted to Danzaemon's rule. Nowadays,
however, the old law is lost and ordinary commoners sell their daughters in prostitution and kawaramono become merchants,
which is the greatest of evils."[147]

The unproblematic acceptance of the existence of an outcaste group by many Tokugawa scholars, regardless of philosophical
or political bent, represents the rejection of a core tenet of Confucian thought, namely, the idea of a universal human nature.
[148] This rejection derives perhaps from a perception of the rigid status divisions of the time as a "natural" state of affairs, and
one closely linked to sacred customs dating back to Japan's divine creation. Embedded in this social or experiential "reality,"
Tokugawa scholars may have tended to downplay or ignore the notion of the universality of human nature so central to
Confucianism. While they read social reality through Chinese texts simplifying society to four classes, they also read Chinese
metaphysics through the practice of a rigidly differentiated status system that posited in a most radical way different degrees
of "intrinsic worth" for members of different status orders ("six or seven kawata lives for one commoner").

[147] For the quotations from the travelogues, see Watanabe, Mikaiho[*]buraku , 153; for the quotation from the Seidan , see
ibid., 152, and NST 36:283.

[148] Nakae Toju[*] and Kumazawa Banzan seem to have been rare exceptions. For Banzan, see above, n. 144; for Toju[*] , see
Kurozumi Makoto, "The Nature of Early Tokugawa Confucianism," trans. Herman Ooms, Journal of Japanese Studies 20, no. 2
(1994): 359-60, originally published as "Tokugawa zenki jukyo[*] no seikaku," Shiso[*] , no. 792 (1990): 117-18. Kurozumi
discusses the difficulty Tokugawa thinkers had with the concept of a universal human nature (without linking it to social
practice as I do here) on pp. 363-65 and 368-72 of the English text and 120-21 and 124-25 of the Japanese text. I am currently
pursuing this issue and others in a full-length study of the kawata and senmin during the Tokugawa period.

― 302 ―

Sung Confucianism, it has often been argued, constituted an advance in "rationalism" over traditional Confucianism and
Buddhism. One may therefore conclude that the coexistence of Sung Confucianism and the "irrational" practices of segregation
and indeed racism in Tokugawa Japan can only be explained by the faulty nature of this rationalism. The argument has been
made, however, with regard to the development of Western racism that it is a strictly modern phenomenon and directly
related to the concomitant emergence of rationalism.[149] It would be intriguing to explore this thesis for Tokugawa Japan, less
to establish a causal relationship than to establish an affinity that goes beyond the observation that a Confucian-style
rationalism had no difficulty making room for racist segregation.

That Tokugawa Japan's intra-race racism was expressed and fed by the "irrationalism" of cultural nationalism as it developed
in nativist (kokugaku ) discourse is more immediately plausible.[150] Given its emphasis on purity and pollution and the
explicit hatred of foreign "others" (which was China and the West, according to Motoori Norinaga and Hirata Atsutane), it
seems quite logical that kokugaku thought would validate and reinforce the creation of an "internal" group of despised others.

About the Dutch and the Russians, Atsutane writes: "The slenderness of their legs also makes them resemble animals. When
they urinate they lift one leg, the way dogs do. Moreover, apparently because the backs of their feet do not reach to the
ground, they fasten wooden heels

[149] Christian Delacampagne, "Racism and the West: From Praxis to Logos," in Goldberg, Anatomy of Racism , 83-89.

[150] While pointing out the connection between nativist ideology and the crisis of village leadership, Harry Harootunian
leans far toward an uncritical acceptance of that ideology, which, he writes, "substituted friendship, affection, and reciprocity
for the fragmentation and conflict reflecting the impersonal relationships of the authority system." His focus on the
relationship between the village and the overlord causes him to overlook the discrimination that was increasing at the village
level precisely when (and where) nativist thought became ruralized (see his Things Seen and Unseen: Discourse and Ideology in
Tokugawa Nativism [Chicago: University of Chicago Press, 1988], 230-32, esp. 232). The province of Shinano, where traditional
dogo[*] families were very powerful in village affairs, counted the highest number of Hirata Atsutane's students (see
Katsurajima Nobuhiro, "Hirata Kokugaku to gonoso[*] ," in Shukan[*] Asahi hyakka, Nihon no rekishi 91 [1988]: 9/74; see also
Fukaya Katsumi, "Bakuhansei shihai to mura yakuninso[*] no kokugaku juyo[*] ," Shikan 91 [1975]: 13-23).

― 303 ―

to their shoes, which makes them look all the more like dogs. This may also explain why a Dutchman's penis appears to be cut
short at the end, just like a dog's. Though this may sound like a joke, it is quite true, not only of Dutchmen but of Russians."[151]
It is interesting to note that Atsutane had to insist on his learned authority in this matter, anticipating that his readers might
dismiss this information as a joke. It is also interesting to compare Atsutane's characterization of foreigners with an early-
twentieth-century "popular notion" that burakumin had "one rib-bone lacking; they have one dog's bone in them; they have
distorted sexual organs; they have defective excretory systems; if they walk in moonlight their neck will not cast shadows;
and, they being animals, dirt does not stick to their feet when they walk barefooted."[152]

As Etienne Balibar notes, "the racial-cultural identity of the 'true nationals' remains invisible, but it is inferred from (and
assured by) its opposite, the alleged, quasi-hallucinatory visibility of the 'false nationals.'"[153] And as deconstructionists are
wont to say, (national) identities can only be established impurely, by incorporating through negation their others, which in
due course they produce. Hence, it is perhaps no coincidence that in Motoori and Atsutane's discourse exclusion served as a
central structuring device and that the discourse developed at a time when efforts were being made to reinstitute and
reinforce kawata segregation through discriminatory laws and regulations.

As mentioned earlier, pre-Tokugawa explanations of the term eta , which are still used widely today, were occupational and
linked to some low occupations or religious taboos because they were associated with bloodshed or the consumption of meat.
The term eta (initially not written with the characters for "plentiful pollution") was first mentioned in a mid-thirteenth-century
dictionary, the Chiribukuro , which explained it in association with kiyome , or purifiers, who removed polluted items from
sacred places, and linked it etymologically to etori , feeders of hawks, and associated the Indian outcast candãla with butchers.
These explanations were repeated in later versions and sup-

[151] Donald Keene, The Japanese Discovery of Europe, 1720-1830 (Stanford: Stanford University Press, 1952), 170. For a long,
virulent analysis and indictment of antiforeignism in Tokugawa thought, see Sajja A. Prasad's three-volume The Patriotism
Thesis and Argument in Tokugawa Japan (Samudraiah Prakashan, 1975-84).

[152] Quoted in Ninomiya, "Inquiry," 56 n. 18.

[153] Balibar, "Paradoxes of Universality," 285.

― 304 ―

plements of the Chiribukuro with a Buddhist slant (the Ainosho[*] of 1446, using for the first time the characters meaning
"plentiful pollution" for eta , associating the kawata with pollution, and the Jinten ainosho[*] of 1532) and throughout the
Tokugawa period in various writings.[154]

Not until the beginning of the eighteenth century was there mention of a different origin for the kawata. In 1712 Terajima
Ryoan[*] , a medical doctor at Osaka castle, mentioned that kawata were butchers, gave them the (Sinified Indian) caste name
sendara , and noted that one should not share fire with them, that they were of a different stock (seishi o koto ni su ).[155] In
1725 Sorai alluded, as mentioned above, to a theory that kawaramono had a different parentage or lineage or nature from that
of other Japanese (genrai sono sujo[*]kakubetsu naru ). Sujo[*] , the term rendered as "parentage," "lineage," "nature" (literally,
"seed-nature"), indicates, according to Kuroda Toshio, that in medieval times differential human qualities were attached to
status by birth, the nobility being the kishu ("noble seed"). On the other hand, the nobility were sometimes spoken of as "not of
human seed but partaking of the imperial nature." At the other end of the social scale, the kawaramono were described in one
fifteenth-century text as "humans, but like beasts." According to Kuroda, the medieval social imaginary was genealogical
throughout and originated in, and was maintained by, the Buddhist karmic world-view. Kuroda sees sujo[*] as none other than
a translation of the Indian term for "caste."[156] In this context, it is worth noting that other derogatory terms for classes of low
people also originate in Buddhism; Kuroda mentions bonge (commoner) and hinin .[157] Yet, we should keep in mind that there
was considerable mobility among the various lower strata, occupations, and statuses of medieval society.

This genealogical imaginary, as we shall see in a moment, continued to be applied to status groups in Tokugawa times.
However, while most of the karmic connotations such constructs may have had in medieval times were eventually lost, to
ancestry was added a racial dimension. As already mentioned, some writings suggested that the origins of the kawata were in
Ezo, in a racially distinct group. In the 1720s and 1730s,

[154] Teraki, Kinsei buraku , 25-26; Morita, "Buraku e no shiteki kanshin," 308, 318; Kuroda, Nihon chusei[*]no kokka , 380.

[155] Morita "Buraku e no shiteki kanshin," 311.

[156] Kuroda, Nihon chusei[*]no kokka , 361, 372, 375, 381, 388, 391.

[157] Ibid., 376, 388.

― 305 ―

when Danzaemon fabricated, in one of his genealogies, a Chinese ancestry for himself, stories about the foreign origin of the
kawata started to circulate. Thus Kyoho[*]sewa , a record of daily life of that period, explains that the kawata were descendants
from Chinese refugees who had lived in the wild and eaten animal and bird meat and were considered polluted. For that
reason they were avoided by good people and forced to live apart.[158] Nativists did not have to specify where the kawata
came from. For them it was sufficient to know that they were not Japanese. Writing in 1795, Tamada Naganori wrote that
"butchers, although living in the divine country, are not of divine descent (shinson ); they therefore do not venerate the gods,
do not wash their hands after relieving themselves, do not use mourning garments for relatives, and do not spit when they see
impurities."[159] Not to discriminate against butchers would be against the Way, wrote Ban Nobutomo in 1847. He maintained
this orthodoxy, although he clearly understood that such discrimination may have had historical explanations and that people
may have had mixed feeling about such practice in his own day:

Butchers of wild animals came to be treated as if they were not human beings. If one inquires about their ancestors,
many are found to hail from respectable families, and it is regrettable that it is impossible for them to become good
people (ryomin[*] ). If, however, today we would not despise them as defiled, this would be against the Way, because
this would be the same as not being repulsed by the pollution of the meat of wild animals.[160]

The transition from a medieval genealogical to a racialized social logic was apparently completed in the early eighteenth
century—around the same time that the nobility in France argued its racial superiority over the common Gaulois because of
their Germanic-Frankish descent![161] It should be noted that by then a quasi-genealogical or

[158] Morita, "Buraku e no shiteki kanshin," 312; see also 317.

[159] Ibid., 318.

[160] Ibid., 330.

[161] André Devyver has traced a similar genesis of racial philosophy of "pure blood" among the French lesser nobility
between 1560 and 1720, culminating in the racist writings of Henry de Boulainvilliers (see his Le sang épuré: Les préjugés de
race chez les gentilshommes français de l'Ancien Régime [1560-1720 ] [Brussels: Éditions de l'Université de Bruxelles, 1973]).
According to Noël du Fail, an earlier contributor to this discourse, one of the worst marriages a country noble could contract
was with the wife of a butcher, which would make his offspring bloodthirsty (181). Unlike in Japan, the French "noble race"
was not said to be indigenous, since its origins were allegedly Frankish: they were the fifth-century conquerors of the
aboriginal Gallic Celts, the ancestors of the common Frenchmen. Historically, the nobility traced itself back to Germany,
mythologically sometimes to Abel (10, 11, 27).

― 306 ―

racial concept of Christians had also been codified in Tokugawa law. In 1687 and 1695 laws were issued against former
Christians who had publicly abandoned their faith but were suspected of having done so only for the sake of appearances.[162]
Therefore all agnatic and affinal relatives to the fourth degree and direct male descendants to the sixth generation were put
under the close supervision of local officials, who had to report twice a year on their occupations, marriages, deaths, divorces,
travels, and so on. The taint of Christianity, like the stigma of pollution, could not be removed and was passed from generation
to generation. Similarly, in sixteenth-century Spain the Inquisition paid special attention to conversos , who were always
suspected of not having become genuine converts.[163] The French nobility (especially the lower rural nobility), Spanish
converted Jews, Japanese ex-Christians, and the kawata—all more or less contemporaries—were all minorities, either
threatened (the French nobles by the ennoblement of many bourgeois) or threatening (economically competitive Jews and
kawata), racialized in order to set them further apart from an otherwise undistinguishable majority.

The kawata did not accept being called eta, and they did not subscribe to the genealogical and racial portrayals of themselves
by majority Japanese. Using the same discourse but inverting its values, they pictured themselves as genealogically noble and
racially Japanese. The major difference was that their counterdiscourse was private. We find mythological genealogies kept in
the important houses of the kawata, but only very recently have they been studied seriously as avenues to the consciousness
and self-representation of this oppressed minority.[164] In the early seventeenth century the kawata from the Kinai region
traced their ancestry back to Indian nobility, a certain prince Entara—phonetically close to eta and sendara —whose line was
punished because he

[162] Yokota, "Kinseiteki mibun seido," 72, 76; see also KDJ 4:441, S.V. "Kirishitan ruizoku shirabe."

[163] Devyver, Le sang épuré , 50-55.

[164] See Mase, "Ishiki no naka no mibunsei," 270-72; Wakita, Kawaramakimono .

 ― 307 ―

had committed cannibalism, or to the Japanese mythical figure Sominshorai[*] , a poor man who had become wealthy because
he had lent his humble abode to a god. In the eighteenth century imperial ancestors (the emperor Ojin[*] or Suzaku) and
divine origins (Hakusan or Ebisu) were invoked. In the nineteenth century, far removed from the original sin of cannibalism
that constituted a complicitous theo- or sociodicy of their low status, the kawata imagined themselves descendants from the
Minamoto shogun (ancestors of the Tokugawa) or considered themselves simply to be codescendants of the gods with all
Japanese.[165]

People locked into a status system usually seek to improve their position within it, yet their struggles often strengthen the
system that oppresses them. A status system will keep on reproducing itself unless some independent institutions develop and
a counterhegemonic practice and philosophy or ideology is produced. Cataclysmic social or political events force fundamental
questions about the system's utility and rationale. Tokugawa Japan had neither institutions such as churches nor independent
courts to challenge the principles of the prevailing political practice. In principle and practice, reason could not prevail over
law, nor law over authority. The collapse of the regime in 1868 was perhaps the cataclysmic event that offered the most
opportunity for change. But the change that occurred fell far short of what the kawata might have hoped for. On the surface
(tatemae), the 1871 decree declared that that official class status was abolished and that in terms of occupation and social
standing the kawata would be like other commoners. But in reality the discrimination against them continued: they were
entered in the early Meiji population registers, not as citizens like everyone else, but as "new commoners (shinheimin )." A
public record of difference was thus established that provided an official ground, and the only one, upon which to base future
discrimination. An important step toward the elimination of such a basis for genealogical racism would be to do away with the
Meiji records that document it.

Until recently, access to these records was open to anyone. In 1976, through the amendment of Section 10 of the Family
Registration Law, "the government restricted access to family registries to family members, their legal representatives, and
officials whose job required it." This, however, is another instance of a tatemae aspect of law, for en-

[165] Mase, "Ishiki no naka no mibunsei," 271-72.

― 308 ―

Plate 7.
Advertisement for Genealogical Tracing. The investigative services
advertised here include "confirmation of nationality and family registration,
including matrimonial history," which practically targets Koreans and burakumin.
Japan Law Journal 6, no. 2 (1993): 2.

forcement is very lax; "local registrars do not consistently enforce the prohibition on securing copies of unrelated individuals'
family registries."[166] Plate 7 shows an advertisement in English for genealogical tracing services.

One late Tokugawa voice on behalf of the kawata was Senshu[*] Fujiatsu's (1815-64), a loyalist rebel who was ordered to
commit seppuku because of his anti-bakufu activities. In a very short tract, Eta o osamuru gi (How to govern the eta), Fujiatsu
articulated a vision of a new Japanese social order that included the emancipation of the kawata.[167] Like the agricultural
utopia that figured in the writings of Ando[*] Shoeki[*] (1703-62), Fujiatsu's imagined new social order made no

[166] Taimie Bryant, "For the Sake of the Country, for the Sake of the Family: The Oppressive Impact of Family Registration on
Women and Minorities in Japan," UCLA Law Review 39 (1991): 120, 121. I thank my colleague Bryant for referring me to the
advertisement shown in plate 7.

[167] Fujiatsu, a retainer of the Kaga domain, was an instructor at the Hayashi College in Edo and then in his domain school,
and finally he served as rotor to the successor of the daimyo in his domain. He was one of the leaders of the small band of
Kaga loyalists. A short biographical sketch serves as the preface to his tract on the kawata, which is published in NSSS 14:565.

― 309 ―

immediate political or social impact; his tract was not discovered until after the Tokugawa period, and it did not appear in
print until 1924. Like Shoeki[*] , Fujiatsu provided evidence of views opposed to the Tokugawa status system, and to the
position of the kawata within it.

Unlike Shoeki[*] , who constructed a counterideology that was complex and decidedly utopian, the work of an intellectual,
Fujiatsu, in his indictment of discrimination against kawata, offered a critical analysis of its sources and a realistic program
for its abolition, the work of an official. He listed theories and rationalizations (historical, genetic, and cultural) used to justify
discriminatory practices and rejected them all. For example, he dismissed as a myth the idea that the kawata originated in Ezo
(Hokkaido); and he countered the idea that the kawata were not human but animal in nature by asking, "Would Heaven and
Earth produce such a thing? If they are not humans, then they would be beasts or birds, grass, trees, dirt, or stones, but how
could they have an animal nature if they have a human body?" Fujiatsu argued against the popular notion that those who
engaged in "low" occupations were by association polluted or base: "In our country there are prison officials, but that does not
make them necessarily base; burials are taken care of by monks, but that does not make them necessarily base; people in the
mountains kill wild boars as an occupation, but that does not make them necessarily base ..." He believed that nothing
comparable to the situation of the kawata existed in the West and concluded that the kawata must be the result of the
government's status system. He then recommended that the status system be abolished and that the kawata be registered as
"good people"—advice that was not followed by the Meiji government—and he urged public assistance for the economic
emancipation of the kawata through agricultural start-up programs.

Conclusion

One could say that the Tokugawa status system, like status systems everywhere, served to retain "privilege among the
powerful and power among the privileged."[168] Inherently discriminatory, it was developed and expanded for specific
economic and political purposes by and for those who had the power and privilege to determine those who should have less or
none. Status stratification sets limits on what people can

[168] Berreman, Caste , 198.

― 310 ―

realize; consequently they are induced to look upon those immediately above them with deference and envy and upon those
immediately below them with contempt and fear.

Such dispositions are the result of man-made laws and rules of social practice, which are effective only in that they are
misrecognized as something other than sheer coercion. Prescriptions in the guise of descriptions transform originally political,
economic, and legal objectives into the "natural" order of things. The ought of the law slides under the is of an imaginary order
structured according to symbolic categories such as nature, descent, purity, and pollution, which are presented as direct
readings of reality but are in fact nothing but political values or social norms. Etienne Balibar notes in a discussion of racism
that "the criteria of differentiation cannot be 'neutral' in practice; they incorporate sociopolitical values that are often
challenged and that have to be imposed via the detour of ethnicity or culture.... Classification and hierarchy are above all else
operations of naturalization, or more accurately, the projection of historical and social differences onto an imaginary nature....
The nature of racism is not one of proportional causes and effects, immanent regulations: it is a nature that is 'inherent,'
'immemorial,' 'always already valorized.'"[169] Like the title "titled peasants," such symbolic constructs have the potential to
develop a power within their own logic and to withstand alternative, critical readings of reality.

The escalating discrimination against the kawata in Tokugawa Japan provides an excellent illustration of this logic of practice
and representation. Based on popular prejudice, accentuated by economic rivalry, articulated further and given a new layer of
reality by legislation, rationalized by racial explanations and racist theories, discrimination and segregation produced, and
kept on producing, a great amount of particularly painful social suffering that from today's perspective it is hard not to see as
akin to racism. That the predominant idiom in Japan was and is pollution rather than the pseudoscientific notion of race,
although that aspect was not absent, makes no difference to those who suffered and continue to suffer from its effects. Ritual
pollution, the central concern of Shinto, may be contingent and circumscribed, but pollution ascribed to bestial origins is an
existential, inalterable predicament.

[169] Balibar, "Paradoxes of Universality," 290.

― 311 ―

Pollution is not an adequate social marker of difference, for it is still an imaginary, spiritual state. The construction of racial
and racist theories was an attempt to make sense of the world as fashioned by bakufu laws. The Tokugawa authorities
legislated visible but nonexistential, sartorial, "attachable" (and therefore also detachable) marks of social status; that is why
"passing" was a major concern. Ultimately, only official records of ancestry and birthplace could identify sites of pollution;
status racism became a state racism because bakufu law constructed the legal ground for an intra-race racism.

Only with the formal collapse of the Tokugawa system did emancipation become a possibility. By then, however, the
intensification of discriminatory and segregationist legislation had succeeded in giving new authority to values and practices
that are now viewed as part of Japanese culture. Somewhere near Shimoda, at the site of the first foreigners' residence after
Japan was opened in 1853, a memorial plaque enshrines in Japanese and English a lie that establishes an official truth,
suggesting also perhaps that gaijin belong to the realm of jingai:

This monument, erected in 1931 by the butchers of Tokyo, marks the spot where the first cow in Japan was
slaughtered for human consumption (eaten by Harris and Heusken).[170]

[170] Townsend Harris (1804-78) was the first consul general to represent the United States in Japan. He arrived in Japan in
August 1856 and established his post in a temple in Shimoda. Henry Heusken (1832-61) was Harris's secretary and served as
his Dutch-English translator.

― 312 ―
6
The Tokugawa Juridical Field and the Power of Law
"Unreason < Reason < Law < Authority < Heaven"
Tokugawa dictum

"I'll go to Edo"
Ken; Kami-Kaize village officials in the hinin kidnapping case; Chuemon's[*] relatives in the Bonboku Incident;
Others

In this study, the intravillage/extravillage dichotomy, especially the ways in which the intramural and extramural were
mutually imbricated, is a major theme. Villages had two sets of regulations, laws with an extramural origin and internal codes,
and the relationship between them changed over time. With regard to village justice, penal practices in principle existed by
sufferance from the overlords and often functioned in "illegal" ways. Certain informal practices suggest that the overlords
silently acknowledged some of these irregularities. Probably more often than not, village and overlord authorities were in
mutual agreement about administrative matters, whether openly or silently.

Order, Justice, Instrumentalism, and Formalism

The principal concern shared by overlord and village authorities was the maintenance of order , which did not necessarily
always mean the same thing to both. This parallel interest also governed and informed legislation and judiciary practice. As
the first epigraph above indicates, authority does not yield to law. At the same time, however, commoners who were
dissatisfied with that order as it affected them personally often threatened to go or indeed did "go to Edo" in the hope of
finding justice there.[1]

[1] While this perspective of justice informs some of the cases analyzed in this study, I came to see its importance clearly
through discussions with Maeda Ichiro[*] and other participants in the seminar at Omni University in Kyoto in the summer of
1994 and with Christine Schoppe at UCLA, whose contributions I gratefully acknowledge here.

― 313 ―

Order and justice—these two Tokugawa perspectives on the role of judiciary practice also implicate the village in the wider
field of power, but more important, they represent two opposing views of the relationship between the juridical domain,
including legislative and judiciary practice, and the field of power. For the rulers (and scholars who, with them, view
Tokugawa society from the top down), an instrumentalist view is almost inescapable: the juridical domain is foremost a means
to maintain the power distribution over which the rulers preside. Obvious on a macro-, social scale, this is also valid at a
micro-, local level. The internally generated "village" rules regarding status in Hozu village and the whole Mino region (for
genin and small peasants) or Tanoguchi village (for newly established titled peasants) are clear examples. The system of
"titled" peasants, "shares," and status differentials among peasants served the same purpose.

On the other hand, commoners turned to petitions, suits, and the courts to seek redress, right wrongs, and, ultimately, find
justice, perhaps hoping that something else—reason, good sense, empathy?—might prevail besides or against the authorities'
vested interest in order. The imagined locus for finding this formal , meta-institutional justice was most stereotypically Edo,
but no petition or suit was devoid of that expectation. Thus, the assumption underlying the practice of many commoners was
that the juridical field was (should be) somehow independent of worldly powers and institutions; otherwise, why bother
petitioning or suing at all? In the course of this study, we have encountered a number of such instances. Ken threatened to go
to Edo, as did Kami-Kaize's village officials in the kidnapping case of Mon, the hinin woman. When a certain Seigoro[*] was
voted into jail without evidence that he was the Higashi-Kami-Isobe village arsonist, his relatives marshaled all possible
resources to set him free. The kawata also resorted to petitions and suits: Nagase hamlet (the Clog Thongs Riot) brought a suit
against Ogoseimaichi village; before staging their protest, the kawata of Okayama domain resorted to petitions; Hozu's kawata
sued the small peasants with good effect; and kawata sued one another.

At a more general level, an institution like the kokudaka system, a means used by overlords to systematically extract their
subjects' sur-

― 314 ―

plus, was used as leverage by the peasants to obtain greater fairness in the allocation of burdens at the village level. The
plaintiffs' sense of justice only rarely caused them to target status institutions as such, as it did in the instance of Hozu's
peasants' demand that they cease to be treated as hereditary vassals by the gomyo[*] , who considered themselves rural
samurai; or when a group of clients (kakae) in Kodaira village demanded that the system of main and branch houses (kakae
and kakaeoya) be abolished; or when kawata questioned the official discriminatory policies, arguing that they were, after all,
tribute-paying peasants.

At first sight one might be inclined to generalize the contrast between the institutional status order and the personal sense of
justice as one between particular and universal values if it were not for the fact that any explicit justification one might
identify as universal or natural is to be found on the side of the status order This sense of justice, rarely expressed discursively,
is limited almost exclusively to the level of practice. I have attributed this practical sense to assumptions about the possibility
of finding formal justice in the juridical sphere apart from the administrative dictates of social order.
Structurally, of course, the juridical domain in Tokugawa Japan was not independent from the administrative apparatus but
an integral part of it: administrators also functioned as judges. In terms of the categories Bourdieu uses to characterize a
longstanding dichotomy in the interpretation of the status of the jurisprudential field—instrumentalism and formalism—
historians of Tokugawa Japan would readily side with the instrumentalists. The instrumentalists conceive the world of law as
a direct reflection of the world of power and therefore as an instrument of domination pure and simple.[2] This view holds
that the juridical field, a mere superstructure without autonomy, permeated throughout by heteronomous social power,
generates no specific force of its own. In contrast, formalists posit (modem) law as a domain closed off from the social world
and hence as functioning solely according to its own specific dynamic, logic, or rules.

The value to a discussion of Tokugawa Japan of the polemical scheme that has structured a debate about the relationship
between modern society and law may seem limited. Here, all will agree, the issue is clearly settled structurally and
institutionally in favor of the instrumentalist interpretation. Yet, further questions need to be addressed.

[2] Bourdieu, "Force of Law," 814-17.

― 315 ―

First, it remains to be seen not whether but how an explicitly instrumentalist juridical institution contributes in its own
specific, symbolic but real way (as transformed, surplus social power) to the social forces that created it. In this respect,
Bourdieu speaks of oblique and misrecognized ways in which outside (heteronomous) forces affect juridical activity that in
turn deploys functions that are not identical but homologous to the operation of these forces. Second, at times the
instrumental function clashed with the commoners' practice springing from a desire for a kind of justice (which I call formal
or meta-institutional) that was not simply a conduit for domination. Attention to such areas of practice will modify our natural
inclination to view the juridical as an instrument of power in premodern, "feudal" societies. Third, since juridical practice,
even when it is instrumentally overdetermined, as in Tokugawa Japan, generates its own specific symbolic power, one has to
examine (1) where such development starts to function through the formalization and routinization of its own practice, as a
brake on the arbitrary use of heteronomous power, and (2) how this symbolic power was put to use by plaintiffs for their own
struggles, which had nothing to do with the rulers' instrumentalist intentions.

Formalist legal history pays scant attention, if any, to law's embeddedness in society. Its narratives typically do little more than
trace the genesis of modern juridical concepts, seemingly for the exclusive entertainment of practitioners of law. Strange as
this may seem today, all prewar and most postwar Japanese legal historians, following German models, organized their
analyses in this fashion.[3] Postwar social historians, however, bringing the study of law within the ambit of social history and
fully aware, from a modern perspective, of the underdevelopment of an autonomous judiciary in premodern (and prewar)
Japan, have aligned themselves until recently with the instrumentalist (Marxist) position.[4] Since the 1980s, Mizumoto
Kunihiko and other historians have explored another dimension. One can loosely theorize their

[3] See, e.g., Nakada Kaoru, "Tokugawa jidai ni okeru mura no jinkaku," Kokka gakkai zasshi 34 (1920): 8, reprinted in
Hoseishi[*]ronshu[*] , vol. 2. (Iwanami shoten, 1938), 963-90. See also the introductory remarks to chapter 4, above.

[4] Kodama Kota[*] , in his review of Maeda Masaharu's classical study of village laws, Nihon kinsei sonpo[*]no kenkyu[*] ,
seriously questions the latter's thesis of even limited autonomy mainly in terms of the origins of village laws, which, as Maeda
readily admits, were skillfully used by the authorities and gradually absorbed by them (see Kodama's "Hihyo[*] to shokai[*] :
Maeda Masaharu, Nihon kinsei sonpo[*]no kenkyu[*] ; furoku sonposhu[*] ," Shigaku zasshi 60, no. 9 [1951]: 66-70). Any grand-
scale analysis of the function of the Tokugawa legal system winds up as instrumentalist almost by necessity; see, e.g.,
Mizubayashi Takeshi, Hokensei[*] ; and idem, "Kinsei no ho[*] to kokusei kenkyu[*] josetsu," Kokka gakkai zasshi 90, nos. 1-2, 5-6
(1977), 91, nos. 5-6 (1978), 92, nos. 11-12 (1979), 94, nos. 9-10 (1981), 95, nos. 1-2 (1982).

― 316 ―

position as closer than the positions of the instrumentalists and formalists to Bourdieu's notion of field and field-specific
power used from below for specific local interests not immediately related to the forces that shaped the field in the first place.
[5]

Although Bourdieu emphasizes the relative autonomy of the juridical field and its specific "world-making" power, he
nevertheless sees this field's output almost exclusively as feedback to society sanctioning the power structure of that society.
He would probably argue that any subversive power that law might possess is itself ultimately subverted by the social and
class embeddedness of the juridical field—an argument that would hold a fortiori for more neutral uses of the law from below.
Because legal action has to follow the rules of the juridical field in order to be legal action, and insofar as these rules
themselves serve heteronomous power (which for Bourdieu is very far, if not all the way), there would be "in the last instance"
no "neutral" use of the law.

In this final chapter, then, I shall try to specify how the juridical field, on both the legislative and the judiciary side, exercises
its instrumental function and where and how that instrumentalism meets internal and external limits.

Contextualizing Tokugawa Legislation

Let us start with some very broad strokes at the risk of stating the obvious. Legislation of the Tokugawa period was Tokugawa
legislation, which means that it was created over the course of two and a half centuries. This suggests two things: qualitatively,
that it was by and large new, and quantitatively, that there was much more of it at the end, in the 1860s, than in the early
1600s.

Tokugawa legislation was new because it replaced earlier codes. The eighth-century Ritsuryo[*] codes, supplemented by the
Kamakura and Ashikaga shogunal formularies of the thirteenth and fourteenth cen-
[5] Asao Naohiro reports on this new trend in Japanese scholarship in his "Kinsei no mibun," 24, 35, 38.

― 317 ―

turies, were partially superseded by the military provincial laws of the sixteenth-century daimyo, all of which, in turn, were
replaced by Tokugawa legislation.

It is no exaggeration to say that virtually all of the warrior codes from the sixteenth century were abandoned and replaced
with new legislation, which no doubt retained certain of their stipulations; conspicuous are cruel punishments. Harafuji
Hiroshi cites just one example as an exception, a house law enacted in 1555 that was applied in the late Tokugawa period.[6] It
should also be mentioned that the Ritsuryo[*] model had some formal influence on Yoshimune's Kujikata osadamegaki of 1742,
according to Dan Fenno Henderson "the high point of written native law in Japan," and on the Tokugawa reformatting of the
imperial institution.[7]

Tokugawa legislation surpassed that of the past also in volume. Various factors account for this increased legislative activity,
which was aimed at regulating ever more areas of daily practice. Legislation became calibrated to a much more finely tuned
status society. This set severe limits on its potential to universalize, because what could otherwise have been general norms
required multiple refractions, a striking example being the discriminatory legislation introduced during the eighteenth
century against the kawata. Moreover, legislative growth was stimulated, mainly in the second half of the period, by the
requirements of an unprecedented expansion of commerce, which could not be accommodated by status legislation. In
addition, laws and regulations were issued not only from the higher authorities, shogun, daimyo, and their representatives,
bannermen, or intendants (lordly law), but also from various constituent social units, such as villages or status orders

[6] On these points, see his "Han Laws in the Edo Period with Particular Emphasis on Those of Kanazawa Han," Acta Asiatica
35 (1978): 51-53.

[7] Dan Fenno Henderson, "The Evolution of Tokugawa Law," in Hall and Jansen, Studies , 204. See also idem, "Introduction to
the Kujikata Osadamegaki (1742)," 539-40. For Mizubayashi Takeshi's recent theories on the importance of the Ritsuryo[*]
system for a proper understanding of the imperial institution in Japanese history, including its particular institutionalization
under the Tokugawa, see his "Bakuhan taisei ni okeru kogi[*] to chotei[*] ," NNS 3:120-58; "Kinsei tennosei[*] kenkyu[*] ni tsuite
no ichi kosatsu[*] : kinsei tennosei[*] no sonzai hitsuzensei ni tsuite no shogakusetsu no hihanteki kento[*] ," Rekishigaku
kenkyu[*] , nos. 596 (1989): 18-27, 597 (1989): 19-33, and 598 (1989): 57; and "Ritsuryo[*] tennosei[*] ni tsuite no ichi kosatsu[*]
(1)," Tokyo-toritsu Daigaku ho[*]gakkai zasshi 30 (1989): 1.

― 318 ―

("corporate law"), where, moreover, practice was also regulated by local tradition (customary law).

The Tokugawa period no doubt witnessed not simply a multiplication of laws but the constitution of an increasingly complex
juridical field that was structured quite differently from either its modern or its medieval counterparts if only because of the
different relationship it had with sociopolitical power. As stated above, in Tokugawa Japan, rulers were also both legislators
and judges. This fact alone, one may presume, reduces the autonomy of the juridical field to virtually a zerodegree level, and
the very notion of such a field to little more than an artificially (anachronistically) constructed category of dubious heuristic
value. The assumption, however, is that the application of sufficient analytical pressure via the notion of "juridical field" will
yield more insights about the absence in the Tokugawa legal system of a separation of powers or a formal rationalism.

Max Weber's elementary typology of law distinguishes between an "interpretation of law on the basis of strictly formal
concepts" ("formal rationalism") and "adjudication that is primarily bound to hallowed traditions" (substantive rationalism).
He argued that for settling ambiguous cases the latter is supplemented by (a ) charismatic justice, by oracle or ordeal; (b ) kadi
justice, "informal judgments rendered in terms of concrete ethical or other practical valuations"; and (c ) empirical justice,
based on analogical reasoning and reliance on precedents.[8] Weber warns against a simple identification of modern legal
systems or democratic systems of justice with formal rational adjudication: kadi justice is still very prevalent in England, as
empirical justice is in America. But he suggests that formal rationalism in the juridical field does not develop without a
bureaucratization of the polity.[9] These precautions

[8] Weber, Economy and Society , 2:976. Weber does not type traditional justice as a form of "substantive rationalism." In view
of the way he defines formal and substantive rationality in terms of economic action (ibid., 1:85-86), however, the term
substantive rationality is appropriate for describing traditional justice since it refers m the application of "certain criteria of
ultimate ends, whether they be ethical, political, utilitarian, hedonistic, feudal (ständish ), egalitarian or whatever, and
measure the results of economic action, however formally 'rational' in the sense of correct calculation they may be, against
these scales of 'value rationality' or 'substantive goal rationality.'" A few lines further on he adds "social justice" as a possible
value (see Bourdieu, "Force of Law," 825 n. 33, 842 n. 59).

[9] Weber, Economy and Society , 2:976-78.

― 319 ―

by Weber notwithstanding, Bourdieu suggests that Weber idealistically assumed the possibility of "rational law," but Bourdieu
argues that predictability and calculability should be measured in terms of the consistency and homogeneity of a historically
specific habitus rather than against an abstract "rationality."[10] For Bourdieu, therefore, there exists only one kind of
rationalism, substantive rationalism.

The juridical field of Tokugawa Japan was marked by substantive rationalism. Lordly law was a mixture of kadi justice and
empirical justice. In village law and practice, as we have seen, a form of charismatic justice was often added to this mixture by
the custom of "ordeal by voting" in the absence of any concrete evidence on the identity of thieves or arsonists. Formal
development, however, was not totally absent. It was made possible, but not necessary, by the development of a bureaucracy
during the Tokugawa period.

The newness of the Tokugawa juridical field can be explained in a straightforward functionalist manner by pointing to the
new configuration of forces that forged a sociopolitical field that differed from the one that preceded it. In the sixteenth
century institutional arrangements and power relations that had prevailed since the early Middle Ages, considerably
weakened since the fourteenth century, were completely scrambled and reconstituted on a new basis, first by daimyo and
then, toward the end of the century, by superdaimyo such as Oda Nobunaga and Toyotomi Hideyoshi. The sociopolitical field,
traditionally structured as a tripartite power block of nobles, clergy, and warriors, who were buttressed economically by the
shoen[*] , or estate, system and legally bound to the Ritsuryo[*] system and shogunal law, was replaced by the daimyo's
monopolistic hold on power and territory, secured through sheer military force.[11] The Tokugawa system guaranteed the
preservation of this monopoly and gave it a hierarchical structure under the hegemony of a shogun who was more powerful
than any daimyo at the time or any shogun (or emperor) of the past.

The Tokugawa system is thus perhaps better rendered as a regime of conquest than as a feudal order. The shogun, daimyo,
and warriors by

[10] Boardieu, "Force of Law," 833.

[11] Neil McMullin reports that during the Muromachi period more than 25 percent of the land in Japan belonged to temples
and shrines, while during the Tokugawa period the number of temples increased but land shares dropped to a mere 2.5
percent (Buddhism and the State in Sixteenth Century Japan [Princeton: Princeton University Press, 1984], 23, 251).

― 320 ―

and large constituted a new class that achieved domination of the whole national territory and population through the
application of unprecedented military force. And they sought to secure their gains through a peace settlement whereby
coercive violence was reduced to a threat and a remedy of the last resort, supplemented by the authority of law—a form of
symbolic violence that was also of unprecedented magnitude.

The context of conquest, needed to properly understand some of the dominant features of Tokugawa society and law, such as
an obsession with the maintenance of hierarchy, population control, social order buttressed by an overall lack of juridical
autonomy that lasted throughout the whole period, is in fact a context of two conquests. First was the complete domestic
victory of warriors over all other warring and nonwarring groups (other warriors, religious institutions, organized peasants,
autonomous towns), a victory accompanied in the last three decades of the sixteenth century by specific institutional
arrangements and measures of social control, which were spelled out in local legislation. Second, in the late 1580s and 1590s,
with the formation of Hideyoshi's paramount military power and the extraordinary emergency requirements for foreign
conquest (aimed at China but successful only briefly in Korea), regulatory creativity, both institutional and legislative,
expanded and took on a national character (within limits set by the division of power between Hideyoshi and the daimyo).

Many commoners, although often victims of the first conquest and shouldering the inevitable sacrifices of the second,
probably felt relief at the stabilization that followed a century or more of civil war. There is little doubt that control deepened
as a result of the national effort of foreign war and the extraordinary need to secure the domestic peace for the production
and procurement of military supplies and manpower. And these emergency measures did not end when the troops returned
home, but remained in place throughout the Tokugawa period, when they were articulated further.[12]

[12] Fujiki Hisashi's contextualization of Hideyoshi's "Peace Laws," which laid the basis for the Tokugawa system (Toyotomi
heiwarei to sengoku shakai [Tokyo Daigaku shuppankai, 1985]), has drawn new attention to the military character of
Tokugawa governance. Chapter 2, above, presents Takagi Shosaku's[*] elaboration on the theme of Tokugawa Japan as a
"garrison state." Many Tokugawa thinkers shared the view that Tokugawa governance (institutions, tax burdens, and so on)
had come into being in the context of the warfare of the sixteenth century and were inappropriate for peacetime. For Arai
Hakuseki, see Kate Wildman Nakai, Shogunal Politics: Arai Hakuseki and the Premises of Tokugawa Rule (Cambridge: Harvard
University Press, 1988), 293-94; for Sorai, see J.R. McEwan, The Political Writings of Ogyu[*] Sorai (Cambridge: Cambridge
University Press, 1969), 33, 96; for Motoori Norinaga, see his Hihon tamakushige , in Motoori Norinaga zenshu[*] , vol. 8
(Chikuma shobo[*] , 1972), 336 ff.

― 321 ―

Consider for a moment some elements of the ruling apparatus assembled during those years: data on the size of the
population and the productive potential of the land, legislation binding peasants to agricultural production, monopoly on
juridical power, and the exercise of violence by the highest authorities, rights that earlier had rested with local communities
and guaranteed them varying degrees of self-determination. This whole ensemble of regulations, arrangements, and measures
was implemented on a national scale to launch the Korean expedition. It was with that aim (or under that pretext?) that
Hideyoshi could inform himself of the economic potential of the whole country, get accurate estimates of available supplies for
his armies, make sure that there were producers to produce them, and secure domestic peace so that his military adventure
abroad would not be hampered by interruptions or trouble at home. Thus, institutions developed in the extraordinary context
of conquest and accordingly provided with high degrees of control (surveillance, harshness, and alarmism turning mundane
tasks and activities into duties for the public authority, the country, in short, "patriotic" duties) became permanent features of
Tokugawa normalcy. Two hundred and fifty years of virtually unchallenged military hegemony and pressures of commercial
development attenuated somewhat the tenor of institutional and legal arrangements launched under such extraordinary
circumstances. Still, the Tokugawa system continued to bear its birthmarks until its demise.[13]

The general allocation of power under this regime, its dispersion and limits within the social field, especially in the early
period, is directly relevant for an understanding of the juridical field and its own specific
[13] Takagi Shosaku[*] argues that samurai status rankings were rooted in the military functions of very early Tokugawa times
(NRT 3:203-23). And Kurushima Hiroshi demonstrates, as mentioned in chapter 2, that perceptions by commoners of their
duties of military support and assistance, duties originally defined in the 1630s but nominal for much of the next two
centuries, were still very strong and functional in the civil war that brought about the fall of the Tokugawa (Kurushima,
"Kinsei gun'yaku").

― 322 ―

structure and power.[14] Once the wars were over, the matter of ruling the realm, that is, of ensuring the continuity of the new
order fashioned after the victors' interests, could not be dispensed with (as the Mongol or Manchu conquerors did in China) by
relying on already established institutions or legislation. There were no national government organs that could do the
conquerors' bidding. They had to invent the state— anew, one might add, after its first institution in the seventh century.[15]

Like a colonial conquering power, these would-be rulers had to operate within the limits imposed by the existing social
infrastructure, which even the most totalitarian modern regimes are incapable of refashioning completely at will. The only
social groups that were eliminated, or forced into underground existence, were religious organizations such as the Catholic
Church and the Buddhist Fujufuse sect. This limitation on power's downward reach was further aggravated by the decision,
made in order to better block at the source possible challenges from within its own ranks, to hold the ruling class hostage to
itself (alternate attendance), which was misrepresented as host to itself, in urban centers.

These circumstances greatly determined the shape of the field of power, which was allowed—indeed, forced—to follow the
general contours and inner divisions of the social field. Lordly power first of all added legal sanctions to already existing social
divisions and thereby claimed the authority to control them. In this way, overall social stability was built by strengthening
inwardly the separate spaces of all the constituent units—the villages, domains, towns, status groups—and by controlling these
spaces indirectly from the outside by ensuring that

[14] Relational concepts of power fields allow one to bypass the academic debate concerning whether, or to what extent,
Tokugawa Japan was feudal or absolutist. For the latest additions to this debate, see Mary Elizabeth Berry, "Public Peace and
Private Attachment: The Goals and Conduct of Power in Early Modern Japan," Journal of Japanese Studies 12, no. 2 (1986): 237-
71; and James White's response, pointed and m the point, "State Growth."

[15] In a series of recent articles, Mizubayashi Takeshi has drawn attention to the importance of the imperial institution for
the organization of the Tokugawa polity. He argues that the restitution was saved and resurrected (in its religio-ritual aspects
only at the very end of the seventeenth century) by the bakufu to meet specifically structural needs (see his works cited in n. 7,
above). This supplements my treatment of this subject in my Tokugawa Ideology , esp. 28-35 passim, 51-55, 162-73.

― 323 ―

coopted prominent insiders were held responsible for intragroup law and order, shouldering the costs thereof.

The superordinate powers of the shogun and daimyo subcontracted members of commoner groups for "administrative"
purposes, foremost for supervising the production and delivery of tribute, which supported the rulers; at the same time, they
handed over to prominent local families the burdens and privileges of maintaining local law and order. More than
"subcontracting" was involved, however. Legally sanctioned because they were connected to ruling power by official duties,
these groups were transformed into something more than they had been, one of the practical effects of "world-making" typical
of state power.[16] The state, or more accurately, the lords, did not merely subsume groups but subjected them to a social and
economic logic, telling them what they were to be and what they were to do. Lineages were still effective, but now their
contours acquired new significance by way of new institutions and leadership positions (goningumi, kumi heads, titled
peasants). The outside appointment of village heads set in motion internal power struggles.

These structural constraints distanced the dominant class, the warriors, from the sites of production, the villages, while they
made direct infrastructural reach unnecessary. At the same time, the rulers pushed as far as was necessary and feasible the
principle of a single sociopolitical hierarchy; it was in their interest that there be only one such hierarchy and that they control
it. The position of the emperor made such an effort problematic at the top. While the emperor was politically neutralized and
controlled by shogunal regulations, his very existence raised the question of the hierarchical apex: should it culminate in the
shogun or in the emperor? As John Haley has remarked, "Neo-Confucianist views of suzerainty were more than inadequate in
rationalizing the respective roles of emperor and shogun ... law sourced in the will of the shogun ... law and lawmaking [were
transformed] from simply an expression of legislated regulation, custom, and precedent into a manifestation if not source of
legitimacy as well."[17]

[16] On the dialectic involved in legal "world-making," of creating or instituting a social reality that already exists, see
Bourdieu, "Force of Law," 838-39, and his Ce que parler veut dire: L'économie des échanges linguistiques (Paris: Fayard, 1982),
125 ff.

[17] Haley, Authority without Power , 54, 55.

― 324 ―

Two solutions were proposed to end this ambiguity—clearly testifying to the desirability of a single hierarchy. In the early
eighteenth century Arai Hakuseki's effort to raise the shogun's legal status to that of a king and Ogyu[*] Sorai's proposal that
the bakufu devise its own rank system independent from the court were obvious attempts at clarifying the position of the
shogun vis-à-vis the emperor by making the former unambiguously the pinnacle of a single hierarchy.[18]

Through hierarchization, the shogun and daimyo fixed directly the positions of all constituent groups, including the nobility
and the clergy (no longer partners in government with the warriors as they had been in medieval times), and through them
indirectly made (new) subjects of everyone. This strategy entailed two consequences: (1) the unavoidable and essential
reliance on regulations, laws, and practices to determine and express hierarchical status while representing it as natural; and
(2) the dual origin for laws and practices, namely, the bakufu and lords, and the corporate (territorial and status) units.

The lords' primary concern lay with any intergroup jurisdictional problems—among domains or villages, between court and
warriors, among kawata or temples, and, for commoners, among occupations— because such discord tested the ability to
maintain the principles of hierarchy, and with the serious and potentially explosive problems of maintaining intragroup order,
because they could affect the position of the constituent units within the one hierarchy. The corporate groups also produced
internal laws and regulations to the extent that they were entrusted with juridical matters.

It should be noted, however, that the infrastructural reach of superordinate (public, "state") power, although subject to the
above flections, extended deeper during the Tokugawa period than previously. In medieval times, self-redress, the armed
variety in the form of vendettas, was a socially accepted and legally sanctioned remedy, whether among individuals, clans,
villages, village leagues, towns, temple communities, city wards, or warrior bands. Warrior courts and public investigations
were available, but only upon the request of the parties involved: "even

[18] On Hakuseki, see Nakai, Shogunal Politics , chap. 8; on Sorai, see McEwan, The Political Writings of Ogyu[*]Sorai , 90-94. On
the neutralization of the emperor during the first Tokugawa century, see my Tokugawa Ideology , chap. 5.

― 325 ―

a murder in front of the jail [i.e., under the eyes of the authorities] should not be investigated unless someone initiates a
lawsuit."[19]

The Tokugawa as "state," however, monopolized in principle the use of coercive force: armed self-redress was outlawed, and
all serious crime had to be reported to the lordly authorities. In the case of violent quarrels, the authorities, while insisting on
their monopoly over law and violence, never played the role of administrators of social justice; they refused to examine the
relative merits of parties involved (as was done in pre-Tokugawa courts) and instead punished both parties according to status
and regardless of respective innocence or guilt (kenka ryoseibai[*] ). One could say that it was a case of martial law applied in
peacetime.

For personal vendettas, official permission was needed, and private settlements functioned through a mechanism whereby a
subject was temporarily entrusted with the penal authority that in principle rested only with the higher authorities. One of the
legal flaws of the celebrated action by the forty-seven ronin[*] from Ako[*] was not that the murder of a bakufu official was a
vendetta but that it was an illegal execution, permission having been neither sought nor granted. In addition, there was the
question of legal definition: was it a violent quarrel (in consequence of which both parties, including Lord Kira, should have
been punished rather than only Lord Asano's being ordered to commit seppuku), or was it an attack by one party only (Lord
Asano, since Lord Kira had not struck back when wounded, in which case the rule of kenka ryoseibai[*] would not apply and it
would have been proper that only Lord Asano be punished).[20]

We can summarize the structural characteristics of the juridical field of Tokugawa Japan as follows. First, its location within
the social field as an integral part of the ruling (military) apparatus made for its thorough penetration by heteronomous
power, making the possibility

[19] Mizubayashi Takeshi, "Kinseiteki chitsujo to kihan ishiki," in Chitsujo , Koza[*] Nihon shiso[*] , 3 (Tokyo Daigaku shuppan,
1983), 114. See also Haley, Authority without Power , 40: "Law enforcement by the Kamakura authorities could only be
activated by outside complaint or accusation generally, and the process remained subject to the initiative and direction of the
litigants."

[20] This argument is central to Tahara Tsuguo's discussion of the incident in his Ako[*] shijurokushi-ron[*] : Bakuhansei no
seishin kozo[*] (Yoshikawa kobunkan[*] , 1978).

― 326 ―

of developing an autonomous juridical field virtually impossible, this in contrast to the medieval conjuncture, and in even
greater contrast to that of modern times. Second, the reach of the law was greater than that of its medieval counterpart and,
arguably, even its modern counterpart. This was because (a ) the greater imbalance of power between the centers (shogun and
daimyo) and their areas of control (villages and towns) reduced local autonomy, this only in comparison to medieval Japan;
and (b ) the law treated as public, and hence subject to status control, many matters that were considered legally neutral and
private in both pre- and post-Tokugawa times. Third, since the expansion of legislation, although the result of a heavier
concentration of sociopolitical power by a military force than in medieval times, did not emanate solely from one source, the
Tokugawa juridical domain was imperfectly unified, constituted as a complex network of mutually reinforcing and partially
overlapping subfields.

Since this study focuses mainly on legal and juridical matters with regard to village practice, the structural and historical
analysis of courts, jurisdictions, and legislative centers as they pertain to specific social formations such as cities, temples and
shrines, and specific status groups need not preoccupy us here.[21]

The Tokugawa Juridical Field Within The Field of Political Power

The undeniably strong presence of heteronomous (political) power in juridical matters predetermines the relative
possibilities, findings, and strengths of formalistic and instrumentalist interpretations. The latter are particularly forceful
since lordly law was foremost a continuation of war by policy, a substitution of coercive (symbolic) violence for real violence
aimed at maintaining the structure of domination established by military conquest. Hence differential punishments for the
same crime depending on whether the parties were rulers or ruled, and more gen-

[21] Good descriptions and analyses of these matters can be found in Henderson, "The Evolution of Tokugawa Law"; idem,
Conciliation , vol. 1, chap. 4; Hiramatsu, "Tokugawa Law"; Haley, Authority without Power , chap. 3; Steenstrup, A History of
Law in Japan , chap. 4; and, for the kawata, Cornell, "Caste Patron," 56-70, and Ninomiya, "Inquiry," 99-100. For a general
description of the route petitions and suits took, see the introduction.

― 327 ―

erally on whether they were of superior or inferior status in relation to one another.[22]

Thus, even a formalistic study of the dynamics of Tokugawa law could not but expose the levels and lines of impact of the
forces of domination. Such an analysis of Yoshimune's great legislative codification of the 1740s, the Kujikata osadamegaki ,
would reveal that many of the articles are no more than casuistic status cases, a multiple unfolding into every nook and
cranny of a penal code according to the fundamental hierarchizing "imaginary" that informed the Tokugawa system.[23]

Very concretely, a comparison with legislation a few decades earlier exposes the operation of this logic, since it shows that the
range of serious crimes against status hierarchy was increased by moving up certain categories to make them subject to more
severe punishment. The Kujikata osadamegaki increased the level of seriousness of certain crimes against superiors and
decreased that of certain crimes against inferiors: the penalty for the murder of a former lord moved up two notches, from
simple crucifixion to crucifixion preceded by exposure in a public place (a bridge in Edo); of a parent or teacher, up two
notches, from gibbeting to crucifixion; of a relative of one's lord or master, one's father-in-law, uncle, aunt, elder brother, or
elder sister, up four notches, from beheading to gibbeting after ambulatory exposure. But the penalty for the unintentional
killing of a child, younger brother, younger sister, nephew, or niece moved down one notch, from beheading to banishment to
a remote island and confiscation of the criminal's land, house, and personal effects.[24]

As mentioned in chapter 4, a formal diachronic analysis of village codes indicates that although punishments were lessened as
time went on, for example, by substituting fines for physical punishments such as

[22] See article 71 of the Kujikata osadamegaki (on punishments for killing and wounding) in Hall, "Japanese Feudal Laws Ill,"
766-71 There was also a Tokugawa saying, already mentioned, that the life of one samurai was worth that of seven
commoners, and another one that the life of one commoner was worth that of seven kawata.

[23] Cornelius Castoriadis, The Imaginary Institution of Society (Cambridge: MIT Press, 1987). Castoriadis uses the term
imaginary to refer to a (socially) unquestioned social metaphor that functions as a template for common sense, or, to use
computer language, a default category through which a society at a particular time organizes, perceives, and partly
misrepresents itself.

[24] Mizubayashi, "Kinsei no ho[*] to saiban," 163-65.

― 328 ―

mutilation or banishment, the codes converged increasingly with lordly authority in the second half of the period, precisely
when, it is generally assumed, the lords were losing their grip on local communities.[25]

Imperfect Instrumentalism

As this late convergence suggests, village codes in the first century of Tokugawa rule were perhaps less immediately
responsive to, or less in tune with, lordly laws than in the second half of the period. Thus, an unqualified instrumentalist
understanding of Tokugawa law as always serving directly the interests of the superordinate powers may suffer from
overinterpretation. We can identify at least four areas in which that may be the case: village codes, the interplay between
village codes and lordly law, commercial law, and the emergence of legal specialists.

The village codes . We do not have in the archives today sixty thousand village codes, one for each Tokugawa village; many
have been lost, many seem never to have been written down, and many were rewritten several times. Those that remain are
rather short, consisting often of fewer than ten articles—much shorter than the written village laws that the lords drafted and
that to a degree influenced the codification of village codes in the eighteenth century. Yet all villages were governed by norms,
whether written or oral, the latter undoubtedly covering a much wider area of practice than the former but not necessarily an
older stratum of practice; codification often signals not the formulation of new rules, but the recording of existing ones. It is
thus very likely that most of the extant village codes, at least those recorded in the seventeenth century, pertain to pre-
Tokugawa practices and thus were not consciously geared toward heteronomous lordly interests.

Peasants, however, were far from insensitive to the possibilities for manipulating intra-village power distributions by
contesting village codes or using them to advantage, a fact overlooked by those who stress the autonomous character of village
laws in relation to heteronomous lordly power. In Hozu village and in the whole Mino region, for instance, village codes
configured relationships between (ex-samurai) titled peasants and the other peasants on a samurai model of vassalage.

The relation between lordly law and village codes . The traditional instrumentalist position has held that village codes were
simply an extension

[25] Oide[*] Yukiko, "Kinsei sonpo[*] ," 98-110.

― 329 ―
of lordly law, and a tool of lordly domination. This View has been challenged by a semiformalist interpretation, but by failing
to problematize power relations within the village, this interpretation has come to understand village codes as unmediated
expressions of the survival needs of the village as an organized social community (Kyodotai[*] ) that was indifferent or even
opposed to lordly power. Recent scholarship, such as Mizumoto's, however, has broken with (antagonistic) essentialist views of
law by positing a space amenable to strategic manipulation rather than the presence or absence of a fixed barrier between the
"autonomous" juridical subfield of village codes and the far larger "heteronomous" juridical subfield of lordly law.[26]

While in general village codes did not contradict lordly law, the latter was not necessarily applied; in practice, village officials
could exercise discretion in the implementation of lordly laws, especially if noncompliance could be legally camouflaged.
Village officials could activate lordly power without actually implementing lordly laws by the mere threat of using them; thus
obliging obedience. Thus, paradoxically, lordly law was implemented at the same time that its formal power was being
manipulated by being redirected, "borrowed" as the mere threat of implementation.

Commercial law . From the beginning of the eighteenth century shogunal courts were forced to (very reluctantly) address civil
suits, whose volume had increased enormously with the expansion of commerce. Civil cases (property or money suits) can be
considered somewhat neutral to the ruling class's political or economic interests because they were irrelevant to the
maintenance of hierarchy; their adjudication constituted, therefore, an unwelcome burden and expense. Yet in this area the
rulers were forced to add to their role as rulers the role of administrators. Hence commercial law constituted a juridical
subfield for which a simple instrumentalist interpretation falls short. Indeed, as Dan Fenno Henderson has argued, it is the
level at which one can identify a formal development of a juridical field,[27] propelled by its own dynamic, and where the
interference of heteronomous social power was weakest.

The adjudication of violent quarrels followed the kenka ryoseibai[*]

[26] Mizumoto, Mura shakai , chap. 6; idem, "Kogi[*] no saiban."

[27] Henderson, Conciliation , 1:11; idem, "Introduction to the Kujikata Osadamegaki," 525.

― 330 ―

principle; no evidence of either party was relevant, and both sides were punished, an unmitigated application of coercive
force. In criminal cases the only decisive evidence was the suspect's confession; hence the use of torture, an extreme form of
coercive force. In civil cases, however, attention to the relative merits of the litigants through an examination of documented
evidence became important. Max Weber noted that in England "the royal power introduced the rational procedure of evidence
primarily for the sake of the merchants," a development that was, therefore, "strongly influenced by structural changes in the
economy."[28]

In this regard one can draw attention to two points. First, every effort was made, even during preliminary hearings (i.e., after
the judicial process had been initiated), to reach a settlement before arriving at a final verdict.[29] This emphatic preference
for conciliation on the part of the authorities testifies to their reluctance to function as judges in cases that did not affect the
power of the dominant class directly. In other words, overlords used their authority to order a binding private settlement, the
terms of which were to be determined by the litigants with or without a mediator. As we saw in the ostracism case that came
before the attendant (chapter 4), such practice also prevailed in noncommercial disputes. Second, the form this conciliation
took, often under threat of dire punishment if the law were applied to the letter—revealing the force of law as such—was as
an application of the kenka ryoseibai[*] principle, dividing the blame, although not necessarily evenly, between plaintiff and
defendant to the detriment of the nonguilty. In the end both sides were guilty. Yet the application of this principle here was not
directly the result of an open application of coercive force, because conciliation was achieved by the parties involved—
tatemae—under the paternalistic prodding of the judge, abdicating momentarily his official position as judge to adopt the
posture of a private go-between. But this was momentarily only, because the next moment he could threaten to let the case
come to a much more dreaded final verdict: honne.[30]

Ishii Ryosuke[*] has drawn attention to one particular aspect of con-

[28] Weber, Economy and Society , 2:977.

[29] Henderson, Conciliation , 1:7, 127, 147, 150.

[30] See ibid., 128, 150, 154, and esp. 156, for attempts at reconciliation in a court case analyzed by Henderson; 149, 153-55, for
instances of the use of threat in the same case. The similarities between the present-day practice of conciliation and the
Tokugawa variety have been explored by Henderson in vol. 2 of Conciliation .

― 331 ―

ciliation, namely, that the fact that it is often ordered under threat is revealing of its coercive nature. The "interrogations in
both the inquisitorial procedure and in the Adversary Procedure (deiri-suji ) were handled by the same officials. Thus the
thinking behind the inquisitorial procedure [where torture was an option] ... influenced the conciliatory procedure of
adversary proceedings; and it was held that in any event, a private settlement cannot be established without basing it on the
free will of the parties. Consequently, 'provisional imprisonment' was used to coerce it."[31]

The emergence of legal specialists . The existence of a corps of legal specialists is necessary for an autonomous juridical field.
Several developments during the Tokugawa period led in that direction. Starting as early as the last decades of the seventeenth
century, the recruitment of investigative and recording officers for the shogunal court was narrowed to the personnel from
one office within the shogunal bureaucracy; these positions ultimately became hereditary, shared among a few families.[32]
Their accumulated knowledge of precedent furthered consistency in verdicts, for practice turned these officers into experts,
and their opinions became authoritative not only regarding the results of the investigations they conducted but also regarding
the penalties; they functioned increasingly as judges. Yoshimune tapped into this experience when he started to codify
standards for judgment in shogunal courts, which had a formalizing impact in the daimyo courts as well.[33]

Since Yoshimune's Kujikata osadamegaki was intended only for internal use in bakufu courts and hence was not publicized,
knowledge of its codes filtered out to commoners indirectly through court decisions, slowing the regulatory effect the
predictability of justicial outcomes might have on social behavior. Yet regularity was enhanced by two practices that limited
arbitrariness: internal queries by intendants or commissioners were sent to higher authorities, and commoners began to make
use of legal consultants (suit inns).

When bakufu intendants or commissioners had to pass judgment on difficult cases, or when the proposed penalty exceeded
their penal

[31] Ishii, review of Conciliation , 223.

[32] Henderson, Conciliation , 1:66-67, 69, 139; Hiramatsu, "Tokugawa Law," 10-11.

[33] Henderson, "Introduction to the Kujikata Osadamegaki," 506; Hiramatsu, "Tokugawa Law," 20-21.

― 332 ―

powers, they queried higher-ups, which ultimately meant the members of the Senior Council. The senior councilors, in
consequence, gathered together all the serious cases from the bakufu courts, examined the proposed sentences to determine
whether precedents were applied consistently, and sometimes consulted further with the Tribunal (Hyojosho[*] ), which
functioned to some extent to standardize the verdicts of all courts throughout the country.[34] Such formalization put
constraints, however limited, on arbitrary judgments.

Igeta Ryoji[*] analyzes a number of queries regarding penalties for kawata who had married commoners and, in addition,
often helped channel other kawata as illegal labor into commoner households. In the latter case, the Tribunal almost always
lowered the proposed sentences.[35] However, according to Igeta, the verdicts alone give a slanted picture of juridical practice
for two reasons. First, in the case of the kawata in the Kinai region, which was beyond Danzaemon's jurisdiction, decisions on
penalties and their execution were entrusted to the elders of the kawata community from where the offender came; since they
produced no documentary evidence, we do not know whether these sentences were harsher than those suggested by the
bakufu Tribunal. Second, the juridical process included provisional incarceration of suspects until a verdict was reached. A
sizable percentage of the suspects died in prison, an understandable deterrent that added considerable coercive power to the
"power of the law" as such. Among those who died during pre-sentencing detainment were 16 of 102 kawata in the Clog
Thongs Riot over a span of two years; 6 of 12 in the Okayama Riot over three years; and 8 of 55 in a mere two months after the
"eta hunts" in Kyoto in 1831.[36]

In the final seventy years or so of the Tokugawa period commoner specialists in legal matters who were sufficiently
knowledgeable in complicated Tokugawa legal procedures and in substantive law introduced their skill and knowledge to the
market. Litigants came to avail themselves of the business of these consultants at their kujiyado (suit inns) near the shogunal
courts in Edo. The litigant in the ostracism case analyzed in chapter 4 seems to have successfully relied on these spe-

[34] Hiramatsu, Edo no tsumi to batsu , 46.

[35] Igeta, "Kon'in sabetsu," 9.

[36] For the general statement and the figures for 1831, see ibid., 19; for the other figures, see chapter 5, above.

― 333 ―

cialists for drafting his suit. The existence of these private legal assistants testifies to the advanced degree of formalization
reached in at least one sector of the juridical field, although their function was limited to cases coming before the shogunal
court. The authorities, although concerned about abuses,[37] understood the benefits commoner assistants offered for the
smooth functioning of the courts and enrolled their services by sanctioning their activity with a license.[38]

Levels of Formalism

The degree of legal formalization and codification is one indicator of the juridical field's degree of autonomy. When discussing
legal formalization, it is useful to distinguish between levels of formalism. An initial formalism, to which great attention was
paid in Tokugawa Japan (as in China), relates to the recognition that cases should be grouped and filed under specific
categories (water and border disputes, money suits, etc.).[39] This first procedural decision was to determine the nature of a
case in order to channel it to the appropriate judicial authorities and to apply to it the appropriate law. As Henderson remarks,
this could not but also reflect "the relative substantive value attached to ... [the] underlying claims [of each type of case]."[40]

More fundamentally, one could say that this is the point where a dispute or incident gets converted into legal matter, the
creative power of legal naming reveals itself, and law has its field-specific effect on social reality. The decision to allow some
events to become legal matter and to deny this to others creates a quasi-ontological divide in social reality;[41] and the effect is
not always to enhance the importance of the former over the latter. Thus, the stipulation that murders and border disputes
had to be brought to court certainly turned such incidents into public events and introduced a new element of power into
them. An-

[37] Henderson, Conciliation , 1:145 n. 59.

[38] Ibid., 167-68. The status of the suit inns contrasts sharply with the illegal status of pettifoggers in China (see Melissa
Macauley, "Civil and Uncivil Disputes in Southeast Coastal China, 1723-1820," in Civil Law in Ch'ing and Republican China , ed.
Kathryn Bernhardt and Philip C.C. Huang [Stanford: Stanford University Press, 1994], 85-121).

[39] See Henderson, Conciliation , vol. 1, chap. 5, for the main categories of "civil suits."

[40] Ibid., 102.

[41] Bourdieu, "Force of Law," 837-40, 846.

― 334 ―

other example is the prosecution, started in 1773, of mixed marriages between kawata and commoners as "adultery," resulting
in status demotion from commoner to hinin, whereas such marriages had been legally neutral prior to that date.[42] That
courts refused to consider cases initiated by inferiors against their superiors (lower samurai against higher samurai, sons and
daughters against their fathers—Osadamegaki , art. 65), however, did not mean that such "authority-related" cases were
deemed less weighty than, say, border disputes; rather it signaled that a certain reality, namely, authority, was above the law.
After all, "unreason does not prevail over reason; reason not over law; law not over authority; authority not over Heaven."[43]

A second type of formalism concerns the bureaucratic procedures whereby an incident, once transformed into a case, must be
submitted in proper form to appear in court. This is the all-important area of legal language and format, which became quite
intricate in Tokugawa Japan, although, and in part because, it was by no means made uniform. The procedures themselves
could be lengthy and complex. Dan Henderson's detailed analysis of a money suit by a villager reveals that a settlement (not a
verdict) was reached after four months, after attempts to settle the suit at the village level.[44] First the village headman's
endorsement was secured, then the local overlord's;[45] subsequently misphrasings underwent correction, the petition had to
be endorsed by the court, and the suit initiated, and after twelve postponements, five stalemates, one hearing before the
commissioner, and six hearings before the recording officer, only a settlement was reached. In another case, involving eight
commissioners, over the establishment of a new market, Henderson

[42] Igeta, "Kon'in sabetsu," 10.

[43] Mizubayashi, "Kinsei no ho[*] to kokusei kenkyu[*] josetsu," Kokka gakkai zasshi 90, nos. 1-2 (1977): 6 ff.; for a different
interpretation of this dictum, critical of Mizubayashi's, see Nanba, "Hyakushö ikki no ho-ishiki[*] ," 47-62. The "reason does not
prevail over law" part of the dictum was incorporated in the third article of the Warrior Code (Bukke s ho hatto ) of 1615 (NST
27:454; for the English translation, see Lu, Sources , 1:201-3).

[44] Henderson, Conciliation , vol. 1, chap. 6.

[45] Ishii Ryosuke[*] (review of Conciliation , 216) questions Henderson's assertion that the headman's approval was needed,
which Henderson also argues on p. 129 of his book. The headman, however, had to endorse with an accompanying letter any
suit by a villager. As we saw in the Kasuga village case (chapter 3), it was considered an abuse of authority by the headman to
forward such a case without his endorsement, which meant that the case would be immediately rejected by the intendant.

― 335 ―

estimates that plaintiffs from three villages located the distance of a five-day trip from Edo must have traveled some thirteen
hundred miles before obtaining a (negative) final verdict.

Similar documents differed from one another depending on whether they recorded the testimony of a commoner (kuchigaki )
or a warrior or priest (kojosho[*] ) or an offense by the former (called kyokuji ) or the latter (otsudo ), or depending on the
commission involved (the Temple and Shrine Commission called complaints not meyasu but sojo[*] ).[46]

The complexity of the above two types of formalism created a special world having a very specific effect on any commoner
who entered it. A veritable arsenal of symbolic devices was marshaled to turn him or her into more of a subject than he or she
was in daily life. First of all, one could not enter this world simply on one's own volition. Endorsements from the authorities
(headman and overlord) were needed, and the language of that specialized world was formulaic and determined in an
authoritative way, severely restricting spontaneous, quotidian expressions. Its tenor was one of supplication and submission,
osorenagara , or "[submitted] with fear and trembling," being the ubiquitous formula.[47] The effect of distance from an
accustomed world increased the subject's isolation and vulnerability, pure subject confronting pure authority. The distance
was reinforced by the layout of the court (called shirazu , "white gravel"), where commoners literally had to grovel in the
gravel one level below the officials, who were seated on a raised wooden or tatami floor.[48] In the late Tokugawa period,
when peasants and kawata appeared in court together, the gravel (or a place located three shaku below the peasants) was
reserved for kawata, while peasants squatted on straw sacks.

A third type of formalism concerns the predictability of the judicial reasoning that leads to legal conclusion. The higher the
degree of formalism and the clearer and more publicly spelled out the principles applied or weight given to precedent, the
more circumscribed the realm of arbitrariness or discretion of the judgments will be and the greater the likelihood that
political power will be bound by law.

Yoshimune's Kujikata osadamegaki (1740s) constituted the apex of

[46] For the two cases, see Henderson, Conciliation , 1:137, 161-62, 166; for the terminology, 166 n. 98, 140 n. 42; 165 n. 95.

[47] Ibid., 151-52.

[48] Ibid., 145-46.

― 336 ―

Plate 8.
Court at an Intendant's Residence. Hanging from the wall at the left
are instruments of torture used to obtain confessions of guilt in criminal cases.
From Ando[*] Hiroshi, ed., Tokugawa bakufu kenchi yoryaku[*] (1965). Reprinted
with permission from Seiabo[*] Publishing Company.

legal systematization during the Tokugawa period. Yet the degree of formalization in the third sense was very limited because
it was not a publicized "code" but a manual to which access was restricted to shogunal officers of justice, especially with
regard to determining penalties. This meant that (1) its principles were intended only for shogunal courts, and not for
application in village or domainal justice; (2) it was "secret" knowledge, not to be divulged in the commoner world; and (3) its
stipulations were no more than suggestive guidelines for judges that were partly based on precedent but did not limit their
penal powers.

The net effect (likely purpose) of the restricted nature of this formalization was the preservation of a discretionary leeway for
judicial officers, which meant, of course, the preservation of the power of the warrior class. Nevertheless, insofar as
investigative and recording officers developed a "legal" habitus, however weak, simply by acquaintance with

― 337 ―

precedent—knowledge that found its way into the weighty responses of the Senior Council or the Tribunal to queries from
below—formalization was a legal, if limited reality.

Practitioners of deconstruction, with their predilection for discursive moments of undecidability, have argued that the one-
time arbitrariness at the origins of laws, meant to eliminate future arbitrary coercive violence by substituting the rules of law,
is in fact reactivated when decisions are made concerning the (degree of) applicability of the laws in each case that comes
before a judge. Typically, however, they have nothing to say about interests orienting the judge through this moment of
undecidability toward closure. Bourdieu's solution to this problem is to be found in his notion of the social construction of the
legal habitus.[49]

Thus, courts were foremost an instrument of the rulers to preserve order, ensuring a slanted distribution of sociopolitical
power. Hence the emphasis on "settling" disputes by restoring peace rather than righting wrongs. Since there was no break
between heteronomous power and the juridical field, "truth" and "justice" accordingly were quite openly secondary
considerations.

Expansion of The Juridical Domain

The extension of the legislative and juridical reach into ever more areas of daily life through prohibitions of all sorts
pertaining to both production and consumption, affected all commoners. People thereby somehow became "officials," as
Suzuki Shosan[*] (1579-1655) and Ogyu[*] Sorai (1666-1724) put it;[50] they were forced to alter their comportment and lifestyle
to fit their assigned status, which was related to their ability to contribute to production (tribute) in support of "the realm" (the
rulers). This affected virtually all items of material and symbolic culture. The logic propelling this legislation of the shogun and
daimyo was one of gradual and reactive comprehensiveness, covering more and more items, in an effort to keep up with
cultural developments that came to be perceived as potentially challenging to the formal status order.

[49] See Jacques Derrida, "Force of Law: The 'Mystical Foundation of Authority,'" in Deconstruction and the Possibility of Justice
, ed. David Gray Carlson, Drucilla Cornell, and Michel Rosenfeld (New York: Routledge, Chapman and Hall, 1992), 13-14, 23-24.
Note the similarity between Derrida's title and Bourdieu's "The Force of Law."

[50] Ooms, Tokugawa Ideology , 131.

― 338 ―
Regulations concerning social practice included prohibitions against the production of luxury items such as tobacco and
certain publications, restrictions on consumption of certain items, such as rice, or the use of palanquins and horses; dress
codes specifying the allowed number of layers of cloth, the cut, and the quality of material (cotton or silk) in a garment;
architectural specifications for peasants regarding the size and/or number of their dwellings, whether or not they could have
gates and what kind, whether or not they could have ceilings, and so on; formalized language necessary for forms of address
and forms of documents (such as petitions); restrictions and prohibitions on forms of entertainment (kabuki, gambling,
Western playing cards); the outlawing of certain religious sects and prohibitions against "new teachings" or the improper
granting of posthumous names; and control of the movement of people traveling, making pilgrimages, or changing domicile.
[51]

Hierarchy seems to be present in all societies at all levels as the result of an informing ideological principle, the effect of the
exercise of power, or the combined effect of both; undeniably, Tokugawa Japan presents us with an extreme example. In this
society hierarchy was, to borrow Castoriadis's term, the "central social imaginary," based consciously or not on the paradigm
of the army. The military model, one could say, provided a far more refined hierarchizing operator than the five relationships
of Confucian teachings could ever have devised.[52] Obviously, this does not mean that practice followed the law or that laws
never went too far in hierarchizing penalties. Laws were often broken, sometimes not enforced, occasionally rescinded. This
did not detract from the pervasive formative force of hierarchizing operators, which were

[51] It would be nearly impossible to give even elementary bibliographical references for this vast area. Suffice it to refer to
two unusual items on this list. For the prohibition against Western playing cards during the Genroku period, which was later
rescinded, see NRT 3:151; for the 1831 prohibition against using improper posthumous names for the deceased, showing
disregard for their official status, see Narusawa, "Mikaiho[*] buraku no kaimyo[*] ," 170-71.

[52] In this context, it is perhaps worth noting that the comparative sociologist Stanislav Andreski, in his search for a qualifier
for a typically feudal society, a "type of society where a stratum of warriors dominates unarmed masses," decided to call it
"bookayan, and the dominating warrior stratum a bookay," deriving these terms "from buke—the Japanese word for military
nobility" (see his Military Organization and Society [Berkeley: University of California Press, 1971], 144).

― 339 ―

sufficiently powerful to permanently mark a great number of salient features of Tokugawa society and culture.

The various grades of banishment from Edo, introduced in the Kujikata osadamegaki in 1742, differentiated between major,
medium, and minor banishment from Edo and gradations of three, five, seven, ten, or twenty miles (ri), although they were
virtually nonexistent prior to 1742. An examination of 974 verdicts handed down between 1657 and 1699 reveals that the most
common distance of banishment from Edo was ten miles. Thus in 1745, only three years after the introduction of these
multiple gradations, previous practice was reinstated by decree, declaring that henceforth the distance of banishment from
Edo for commoners would always be ten miles.[53]

The net overall social effect of such an imaginary was the conversion of an extensive field of (private) cultural practice into an
explicit (public) politically informed practice. Legislation and the accompanying juridical pronouncements have a
transformative power, an "illocutionary" force that by naming reality in legislative utterance or juridical verdict thereby
creates, reinforces, or transforms it.[54] In this way, discursively unarticulated (dispersed) "elements" of private cultural
practice whose meaning was previously heterogeneous, open-ended, and unstable are by edict transfigured into articulated
"moments" of a public space, their meaning now homogeneous, closed, and predictable.[55] This, at least, one may perceive as
having been the intention behind many of the cultural regulations that multiplied during the Tokugawa period, although no
power could ever achieve signifying closure that would prevent the generation of new signifieds, a point to which we shall
return.

The modality of legal expansion in the Tokugawa period is homologous to that of the expansion of knowledge in the Sung
Confucian tradition. Both reveal an unlimited capacity to recreate (through legislation) or rediscover (through knowledge)
again and again in ever new sites, social hierarchy in the former case, natural hierarchy in the latter.[56] It can be argued,
however, that legislative or juridical practice

[53] Tsukada Takashi, "Kinsei no keibatsu," 96-97.

[54] See n. 16, above.

[55] For the distinction between politically free-floating "elements" and politically mobilized and structured "moments," see
Ernesto Laclau and Chantal Mouffe, Hegemony and Socialist Strategy (London: Verso Press, 1985), chap. 3, esp. 105-6, 127-28.

[56] Ooms, Tokugawa Ideology , 95, 240-41.

― 340 ―

was more powerful than teachable doctrine because law restructures the context and thereby reconstitutes the nature of
social practice, including the production of representations specific to that practice, that is to say, social knowledge.

It goes without saying that any power to fashion sociocultural practices and representations, a power to create society through
"vision and division,"[57] must be backed by the necessary weight of coercive force if it is to secure a sufficient degree of
implementation. On the one hand, the power of law, in its thematizing and unifying effect of social vision and division,
extends well beyond the range of social practice as merely regulated by law enforcement. The comprehensiveness and the
thematic consistency of Tokugawa law more than made up for an alleged "shallowness" (Henderson)[58] because it gave public
meaning to so much of daily practice. The purpose of this strategy was not to provide a supplement to private meaning but to
erase it and produce a formal substitute. Dress codes were not an addition to fashion, but were supposed to replace sartorial
expressions of individual idiosyncrasy, taste, or wealth, of privately imagined selves, with the markings of a hierarchical
position in the body politic, expressions of publicly assigned selves. Yet semiotic reality is such that signifiers such as clothing
always produce potentially ambivalent or polysemic signifieds: what the authorities produced, intentions notwithstanding,
were not substitutes but supplements of meaning.

This ruling strategy, in imposing a single signified (hierarchy) on a large number of heterogeneous signifiers, established an
authoritative orientation for the production of cultural meaning. In other words, the hierarchizing principle expanded beyond
juridical matters through the initiative of nonrulers, which converted more and more heterogeneous signifieds and made
them converge into the one signified of the central imaginary. For example, as mentioned earlier, there was no law fixing the
price of samurai human waste in the market of fertilizers higher than that of commoners, yet it was the case that the exchange
value of human excrement was not determined "rationally" by the market or by

[57] Bourdieu, "Force of Law," 838.

[58] Henderson, speaking specifically about shogunal law rather than Tokugawa law in general, says, "The coverage of
Shogunate law was quite shallow, both in terms of territory controlled directly and persons affected directly" (Conciliation
1:62; see also 100).

― 341 ―

its use value but was measured according to the status hierarchy of its producers: samurai night soil fetched a higher price
than that of commoners.[59] Similarly, the production of etiquette booklets as guidelines for proper hierarchical human
interaction indicates the necessity and aspiration to conform to one social model, that of the samurai.[60]

The complicity between subjects and authorities is also illustrated by the readiness of some parents to redefine unfilial
behavior as the authorities had done, from ethically reprobate behavior to a legally punishable crime. In the second half of the
seventeenth century a number of parents requested that their unfilial sons be temporarily jailed, and in the eighteenth
century parents often asked that they be temporarily committed to a penal colony or stockade.[61] In Kodaira village, for
example, as we saw, commoners stretched the limits of hierarchizing status further than the authorities intended; statuses
could be split almost endlessly to simultaneously accommodate efforts aimed at social mobility while safeguarding the power
elite against such ambitions. Competing proximate occupational groups often went to court hoping to have the authorities
settle conflicts over territory or hierarchy to their advantage. Status thus became a weapon and a prize, for example, for the
carpenters from a district in Harima who requested clarification on which of their subcategories were polluted; in disputes in
Azumi district between Ebisu, lion dancers, and kawata; in suits whereby hunters and Kyoto's dyers succeeded in dissociating
themselves from the kawata—all examples of commoners seeking advantage within the rules of the rulers' game of social
ordering, which by publicly, legally sanctioning groups thereby created status positions for them.[62]

To a large extent, the net social effect of the incorporation of legal constraint into practice was the naturalization of social
practice thus constructed. In the hierarchical structure that was sanctioned, recreated, and expanded in this way, each status
position, because locally specific

[59] Walthall, "Village Networks," 295.

[60] Eiko Ikegami, "Disciplining the Japanese: The Reconstruction of Social Control in Tokugawa Japan" (Ph.D. diss., Harvard
University, 1989).

[61] For the jailing, see Tsukada Takashi, "Kinsei no keibatsu," 100-101; for the penal colony (the stockade, or yoseba ) of
Ishikawajima, an island in the lower Sumida River in Edo, see Hiramatsu Yoshiro[*] , "History of Penal Institutions: Japan,"
Law in Japan 6 (1973): 11, 15.

[62] For further examples, see Asao, "Kinsei no mibun," 35, 37-38; and Tsukada Takashi, "Kasomin[*] no sekai."

― 342 ―

and immediately differential only with regard to the positions immediately above and below it, narrowly limited what was
possible or impossible, forming a localized, circumscribed habitus. Limited aspirations precluded ambitions challenging
superordinate powers more effectively. The farther removed the status in question was from super-ordinate powers, the more
effectively limited aspirations precluded ambitions challenging those powers. We have seen a number of "emancipation suits"
that illustrate this kind of mechanism.

Structurally, one can say that a tenant peasant (tribute producer ) might aspire to succeed to becoming a tribute-paying peasant
with land, but not a village headman, at least not immediately. And he feared becoming a hinin or registered beggar. A
landowning peasant might covet the post of headman or membership in the shrine association (miyaza) or "nominal" samurai
privileges (wearing one sword and using a surname) but not full membership (not even in the lowest strata) in the ruling class.
The peasant's greatest fear, of course, would be to end up landless. A commoner's daughter might hope to serve in a samurai
household or to marry into a lower-class samurai family, but she also dreaded being handed over to such a family as a virtual
slave as part of the punishment for a crime committed by her father or husband.[63]

Moreover, in regard to a regime where authorities at all levels— shogun, daimyo, as well as village elite—invested much
symbolic power in explicitly shaping cultural and social practices, and where that effort was met halfway by the colluding
practices of the dominated (their aspirations revealed by suits and petitions), the question whether legitimacy was an internal
matter of acceptance or consent or an external matter of coercive power and resistance becomes difficult to answer. Indeed, it
may be irrelevant if one considers the objective relations in this conjuncture of legal imposition from above and practical
complicity from below.[64]
This publicly created (political) equivalence of all signs—through their restricted single signified, the division between rulers
and ruled, and

[63] On legal stipulations concerning, and real cases of, wives, daughters, and sometimes sons of criminal commoners given as
"slaves (dohi )" to samurai families, see Tsukada Takashi, "Kinsei no keibatsu," 101-2.

[64] Concerning this line of argument, see Bourdieu, "Force of Law," 841, 844. On the effects of universalizing values promoted
by juridical formulations, discussed in the following paragraph, see ibid., 838, 845; and idem, Ce que parler , 72-73.

― 343 ―

its numerous refractions and reflections in "civil society" as the separation into superior and inferior—expanded the spheres
of domination. At the same time, however, it also increased the system's vulnerability, because it multiplied pointillistically the
sites where behavior could be interpreted as illegal or rebellious behavior. If etiquette "manages to extort what is essential
while seeming to demand the insignificant [and] concessions of politeness always contain political concessions,"[65] by the
same token, the essential can be denied while refusing the insignificant. Thus, within this setting, cultural practices could as
easily be redefined as deviant illegal behavior because of the implied rejection of the universally signified. It suffices to recall
the geta incident in Hozu village (chapter 4), in which both sides held precisely opposing views regarding what was essential
and what was insignificant in the peasants' disregard for the prohibition against wearing wooden clogs. Hence almost
anything could become subversive through the creation of a new signified or the restoring of a suppressed one for any
publicly marked cultural signifier.

To suppress the social tumult caused by the sexual and erotic commotion surrounding the early kabuki theater, the bakufu
tried to eliminate a signified (provocative sexuality) by doing away with its signifier (female actors), assuming that there was
only one possible signifier (females) for that signified, when it banned women actors from the stage in 1612. Yet, since
gendered sexuality is not necessarily linked to sex, female impersonators (onnagata ) circumvented the effect of the ban with
clamorous success. The bakufu then thought that it was striking at the heart of erotic signification when it required, in 1652,
that all actors shave their forelocks, which at first made them look, in the eyes of one contemporary, like "cats with their ears
cut off."[66] The actors, however, re-created erotic allure by wearing scarfs, caps or even wigs. By legislating political values
into cultural practice, the bakufu multiplied sites for flaunting the law and displaying contempt for authority, yet
infringements of cultural prescription constituted no ultimate threat to the order of domination.

To the extent that it was accepted that behind Tokugawa norms lay a norm of norms, namely, the natural order portrayed by
Confucian

[65] Bourdieu, Logic of Practice , 69.

[66] 66. My interpretation is based on data found in Donald H. Shively, "Bakufu versus Kabuki ," in Hall and Jansen, Studies ,
236-39.

― 344 ―

teachings or the divine order taught by Shinto,[67] the bakufu was safe from the critique implied in cultural practice's
deviation from those norms. Of course, the crucial question is to what extent this "to the extent that" is applicable, especially
since a "natural" or "divine" order is but a description elucidating a world-view. As Lyotard has pointed out, description and
prescription (norm) are two intrinsically different "language games" that can be engaged only sequentially by shifting from
one game to the other, without description's ever, strictly speaking, by itself yielding prescriptive force or providing its ground,
because the two are incommensurable. Since prescription contains its raison d'être and authority within itself, its oblique
reliance on description makes it, at least in principle, vulnerable to questioning. This postmodern viewpoint is not different
from Derrida's deconstructionist denial of any foundation for authority; see his qualification of it, mentioned earlier, as
"mystical." Such interpretations are generated in part by the linguistic theory of performatives. Bourdieu has incorporated this
view, since he writes of "performative utterances" as "magical acts which succeed because they have the power to make
themselves universally recognized." On the other hand, he speaks elsewhere of the power of a "norm of norms."[68]

Moreover, insofar as prevailing linguistic usage was an embodied expression of order in a practical, unreflective, and
"natural" way, in contrast to the conceptual, reflective, and disputable Way of teachings, language constituted an additional
rampart of defense against ideological challenge. As Anne Walthall's recent research indicates, women could step beyond the
bonds of propriety and speak out politically on their concerns; protesters could use "vulgar" language to address the
authorities in the heat of protest, attempting to force the addressees to their own level.[69] Yet, these attempts by commoners to
submit the

[67] Ooms, Tokugawa Ideology , 250-62, 281-86; idem, "Yamazaki Ansai no 'Kamiyo no maki' ni okeru kaishakugaku—tenkeiteki
ideorogii-keitai to shire," Shiso[*] , no. 766 (1988): 15-18.

[68] Jean-François Lyotard, Just Gaming (Minneapolis: University of Minnesota Press, 1985), 45, 52-54, 59, 99; Bourdieu, "Force
of Law," 838.

[69] Anne Walthall, "Representations of Women in the History of Japanese Peasant Uprisings" (paper read at the Western
Conference of the Association of Asian Studies, Long Beach, Calif., October 1989); idem, "Edo Riots" (paper read at the Southern
California Japanese Seminar, UCLA, December 8, 1989).

― 345 ―

authorities to their own rules of communication and signification by disrespectfully rejecting the order of respect and power
were doomed to failure not only because such attempts were sporadic and limited or gambled against overwhelming odds but
(perhaps mainly) because the very modalities of reversal or trespassing were dictated and controlled by the field of power as it
was structured by the authorities. In other words, the same rules dictating the maintenance of the field of social forces also
informed the shape of attempts to reorder it. Women were socially stigmatized by such behavior as "beyond the pale" or as
behaving "like men," that is, behaving "unnaturally." Paradoxically, hierarchy was reaffirmed through acts of denial that were
rejected as socially unacceptable.

Only critiques that took as their target the historically contingent unequal distributions of power or the malleability of the
social imaginary of hierarchy (informing legal or natural representation) escaped the stigmatizing hold of hierarchy, because
such critiques unmasked and exposed the misrecognized interested power behind ideology. The first attempt of Kodaira's
kakae peasants to do away with the patron system and the demand by Hozu's small peasants that the titled peasants of the
gomyo[*] stop treating them as vassals partake of the nature of a practical critique. Senshu[*] Fujiatsu provided a succinct but
radical critique of the status system, which he singled out as the sole cause of the kawata's plight. At the level of discourse, a
student of the famous Confucian scholar Sato[*] Naokata (1650-1719) recorded his teacher's remark that Hideyoshi and Ieyasu
were like two brigands who took what did not belong to them by killing others.[70] In the writings of Kamo Mabuchi (1697-
1769), the renowned nativist scholar (kokugakusha ) patronized by Yoshimune's son Tayasu Munetake, one can find a similar
sentiment hidden in a text attacking the idea of religious retribution:

In former times,... everyone was waging war with everyone else and killing people. Those who killed nobody then is
today just a nobody [literally, "a commoner"]. But if one killed a few, one got to be called "bannerman"; and a few
more, "daimyo"; and even more, "super daimyo of a whole province." However, he who killed an exorbitant number
became the most highly regarded person [shogun] and flourishes gen-

[70] Sato[*] Naokata, "Sato[*] Naokata sensei setsuwa kikigaki: 5-hen unzoroku[*] , maki 2," Mukyukai[*] Collection, Tokyo, 30b.

― 346 ―

eration after generation. Where is retribution in all this? We see that it comes to the same whether one kills men or
insects.[71]

Ando[*] Shoeki[*] (1703?-62) is well known for his full-scale attack on the principle of hierarchy itself:

From Fu Hsi to Confucius, there are eleven men called sages. They all violated the true way of nature. Their desire
to rob the world and the state have plunged the world into war....It is indeed lamentable that the sages, being
ignorant of nature, established private laws ... beginning with Fu Hsi, ... the Yellow Emperor, ... the Duke of Chou ...
and all the sages, saints, and Buddhas down the generations ... all the scholars ... including Hayashi Razan and Sorai,
were nontillers and violated the way of heaven by robbing the common people who engage in direct cultivation.
They have gluttonously devoured the people's surplus grain ... they distinguish between high and low, and
proclaiming this to be the law, they claim to possess princely characters and place themselves above others ...
teachings that are based on separating everything into two [correspond to the] world of law, [of gluttonous
noncultivation]."[72]

Yet it is significant that none of these critiques were public statements. A later reader of Naokata's remark, jotted down by a
disciple, wrote in the margin, "This ought to be erased." Mabuchi's critique was published only after his death, and Ando
Shoeki's[*] writings were made public only in the Meiji period, as was Fujiatsu's political essay. Tokugawa censorship, implicit
or legislated, had the power to control public discourse that was critical of the social imaginary; here the force of law was
effective.[73]

Conclusion

What are we to make of the development of the vast juridical domain in Tokugawa Japan? A structuralist perspective will
almost inevitably lead to an instrumentalist conclusion of the function of law; a strictly modern perspective would question
the significance of the Tokugawa field

[71] Kamo Mabuchi, Kokuiko[*] , in Kinsei Shintoron[*] ; zenki kokugaku , Nihon shiso[*] takei, 39 (Iwanami shoten, 1972), 388-
89.

[72] Ando[*] Shoeki[*] , quoted in Maruyama, Studies , 254-55, 258, 260.

[73] Jacques Joly would disagree with my attribution of critical importance to Ando[*] Shoeki's[*] writings. See his "L'Idée de
shizen chez Ando[*] Shoeki[*] " (Diss., Université de Paris VII, Unité Asie Orientale, 1991), and idem, "Spontanéité et nature: Le
cas d'Ando[*] Shoeki[*] . Comparatisme et récupération," Revue Philosophique de Louvain 92, no. 4 (1994): 546-69.

― 347 ―

of law because of the lack of even a minimal relative autonomy; and positivists would be negativists, pointing out the great
discrepancy between the number of laws on the books and the questionable level of enforcement. By focusing, following
Bourdieu, on "world-making" and "social consecrating" effects of the symbolic power of law, however, we may appreciate in a
new way the importance of the creation and growth of a Tokugawa juridical field.

In this regard, James White, in reopening the old question whether and in what sense the Tokugawa regime constituted an
absolute state, has made a useful distinction between actual (state) development or growth and capacity. He posits that the
"central assertion of a claim to such [legitimate] control [of coercive resources] ... which either is not widely questioned or is
actively confirmed by significant political actors constitutes state growth, even if de facto realization [i.e., capacity] of the claim
takes some time and even if the claim is effectively institutionalized for less than the lifetime of the regime in question. Such
claims, even if unexploited, are the stuff of which modern nation-states are made."[74] He argues that "centralization and state
development continued" in Tokugawa Japan, although "state capacities deteriorated over time."[75] White's assessment can be
read as a positive evaluation of the symbolic importance of Tokugawa law.

The claim, tatemae, to legitimate control of coercive resources, whether violent or symbolic, by the superordinate powers was
never widely or publicly questioned in Tokugawa Japan. This does not mean that the situation was one of an authoritarian
regime in supreme control of its subjects. In practice, this claim did not translate into a direct monopoly on the exercise of
authority by the warrior rulers. Many commoners were "subcontracted," that is licensed, delegated, or simply allowed to
oversee the maintenance of law and order. Positions of authority below that of the intendants—those of their assistants,
village group headmen, village headmen, village officials, in the early period even including a number of intendants
themselves—were usually filled by coopted prominent local bosses, who were often economically better-off than their
dependents but always buttressed by relatively ele-

[74] White, "State Growth," 9.

[75] Ibid., 11. Philip Brown makes extensive interpretive use of White's categories for state development and state capacities in
his Central Authority , 12, 25-26, 231 ff.

― 348 ―

vated status and concomitant privilege. Although the rulers were warriors, they delegated to others even the ultimate use of
violence, the killing of life (when no valor or honor was involved): hunters were licensed for "killing life," and kawata were
called upon to take part in armed confrontations with peasants and to execute criminals.[76]

Similarly, laws and regulations often incorporated existing commoner practices , such as ways of locally adjusting individual
tribute contributions by households when villages were affected internally and unevenly by partial crop failures; crime voting
and banishment (if reported at least after the fact) in the realm of penal justice; the proliferation of status distinctions among
commoners. Thus the rulers had to compromise not only with local elites but with their practice as well. World-making by the
rulers often followed the contours of the world commoners had already made, admittedly in the process consecrating that
world.

Non-elite, common commoners had little choice but to play the social game by the rules set for them by their immediate and
distant superiors. Both levels of authorities were not merely concerned with maintaining law and order; they were obsessed
with it and extremely sensitive to any potential disruption of the peace. This obsession may explain the martial-law solution of
punishing both parties in violent quarrels and its extenuated form as a propensity to spread guilt to both sides in conciliation
settlements and the practice of nipping occasions for private vendettas in the bud by producing perpetrators of theft or arson
through the ballot box. Nevertheless, commoners, undeterred by such practices, which restored order but failed to satisfy their
demands for justice, forced the authorities to spend far more time and energy than they wished responding to petitions and
lawsuits; because these practices were officially discouraged, those who initiated them were stigmatized as troublemakers.[77]
The presumption by all subjects, including the kawata, that they had a "right" to petition or sue was perhaps a result
(unintended by the authorities) of the fact that, unlike in pre-Tokugawa times, everyone was now recorded in the population
registers and thus identified as a subject with public duties.

Local elites and common commoners, by making the system work to

[76] The need of a license for hunting ("the killing of life") is mentioned in village laws; see art. 20 in appendix 3.

[77] See arts. 16 and 54 in appendix 3.

― 349 ―

their own advantage as best as they could (often in opposite directions), thus contributed to widening the scope of the
warriors' role in society beyond what they originally intended. Law, in its individual use in petitions and suits, must thus be
seen as something more than a pure instrument of domination, although structurally there is no denying that it was
predominantly instrumentalist.

― 351 ―

Appendix 1
Settlement of a Dispute between Kumi Heads and Small Peasants, 1760 (Iribuse,
Kita-Saku District, Shinano)
1. Since the ancient past, the offices of kumi heads and peasant representative in the village of Iribuse in Saku-gun have been
filled from among the twenty-four titled peasant households. Now the small peasants are fighting to establish their own
[representatives], and a dispute has erupted between the two sides. The shogunal office has conducted an investigation.
Tachu[*] , from the way station Hiraga, and Tokorozaemon, headman from Gorobe-shinden, submitted a petition for settling
the dispute that this office, having heard both sides, endorses. The content of the agreement is as follows.

2. Goningumi. According to the law, the twenty-four titled peasants have been divided into groups of five to form five kumi. In
each kumi these five titled peasants, together with the small peasants of each kumi, have elected their kumi head from among
these five titled peasants. Of course, the small peasants cannot insist that one of theirs be elected. Now, it has been decided that
when it becomes unavoidable that a titled peasant relinquish his office as kumi head, the transfer has to occur with the full
consent of the titled peasant; it cannot be done without his consent.
3. When five kumi are made out of the twenty-four titled peasants, one of the kumi will have the headman; which of the five
kumi this will be has to be decided by the titled peasants and the small peasants in a meeting, and the kumi from which the
headman is selected will be the headman's kumi; the other four kumi decide on their heads as men-

The Japanese text can be found in NAK-KS2 (1): 963.

― 352 ―

tioned above. Among these four kumi, one kumi head's seat will be empty and the three other kumi leaders will function
instead, the vacant post shifting every year to another kumi [this vacancy related only to the village council, not to intra-kumi
matters].

The stipend for the kumi heads is, as stipulated in the past, four bales per head, which makes twelve bales for three heads. The
headman is especially important; hence the titled peasants and the small peasants, without discrimination, shall elect him
from among the twenty-four titled peasants at a meeting of the whole village. From now on, the kumi head will be decided on
by the kumi.

4. The peasant representative: to be selected by the small peasants from now on and to rotate among them. His stipend will
remain at four bales, as it has been in recent years, and we shall not go back to the two bales of eight years ago; this by a
decision of all the peasants. Until now the positions of headman, kumi heads, obviously, and the peasant representative have
been limited to the twenty-four titled peasants, and the small peasants could not serve in these functions. This is not according
to the law. From now on the peasant representative shall be selected from among the small peasants, and the office will rotate.
In anything that does not relate to business with the shogunal office, he will function as an equal to the kumi leaders. Of
course, in relation to the village budget, utmost care has to be taken to avoid unnecessary expenditures; an earnest effort has
to be made that quarrels do not erupt again and again.

― 353 ―

Appendix 2
Goningumi Rules, 1640 (Shimo-Sakurai, Kita-Saku District, Shinano)
1. Goningumi should not be formed only with kith and kin; membership should be diverse.

2. Re : The yearly investigation of affiliation with Christianity. There is not a single Christian in this village. The peasants are
investigated every month, and comings and goings are checked with the pertinent temple in each case to verify affiliation.
Therefore, should there be a Christian in this village, not only his goningumi and the headman but the entire village will be
punished.

3. If there is a thief in the village, it should be reported to the shogunal authorities. Presently there are none. If one is found,
the person will be apprehended immediately and turned over to the shogunal authorities.

4. The term of indentured servants is limited to ten years, as ordered. Anyone exceeding this term by even one month will be
punished by the shogunal authorities. A person whose term is completed should know that he cannot come and go or deal
with his employer [as if he were still employed].

5. Prohibitions against traffic in human beings have been issued. Henceforward both parties in these transactions will be
punished.

6. Wells should be dug and rivers cleaned in due time; cultivation engaged in without delay; yearly tribute paid in full.

The Japanese text can be found m virtually the same form in Ichikawa Yuichiro[*] , Saku chiho[*] , "Shiryoshu[*] ," 1-3; and
Hozumi Shigeto[*] , Goningumi hokishu[*]zokuhen , 1:18-23. This is one of the oldest goningumi laws from a bakufu territory
(Hozumi lists one other, comprising six articles, dated 1629).

― 354 ―

7. Turns for guarding the storehouse should be faithfully kept to avoid arrears in tribute rice; in case of fire or theft, loss has to
be made up by the entire village.

8. If someone commits the offense of not paying tribute in full or absconds without paying it because he or she cannot make
ends meet, the entire village should share the unpaid portion and pay it promptly in full.

9. Nothing that has to be turned in to the shogunal authorities can be sold to others.

10. Wives of headmen and peasants cannot wear starched silk or silky fabrics; peasants can only wear clothes made of cotton
or thick paper at any time.

11. Roads and bridges have to be fixed regularly to keep them in good repair.

12. Bamboo cannot be cut without official permit.

13. Gambling, lotteries, and games of any kind are prohibited.

14. Business trips have to be reported to the headman and goningumi before departure, stating the purpose and the places
where one will stay; if no such notification occurs and an investigation is made by the shogunal authorities during one's
absence, this offense will be punished.

15. All orders from the shogunal office must be obeyed; circular letters from the intendant's assistant should be passed on
without the slightest delay; negligence in delivery will be punished.

16. The width of silk and pongee should be of standard size.

17. Wounded individuals cannot be given lodging.

18. Peasants who fix up their hair in bamboo-whisk style, wear long swords, disobey their parents or master, or do not engage
in cultivation but spend their days in idleness must be reported to the authorities and cannot be concealed even if they are
one's children or servants.

19. People with complaints can lodge suits only on their own; they cannot encourage others to do this, invoking gods' names;
as a matter of course, they cannot become involved in others' complaints elsewhere.

20. This document should be rewritten every month, and all villagers, gathered at the headman's, should affix their seals to it,
and the certificates should be handed over to the intendant's assistant.

― 355 ―

21. Oak, lacquer, and mulberry trees, obviously, but also any kind of tree and bamboo should be planted with due diligence.

The above articles should never be violated.

[Date.]
[Signatures.]

We have grouped all villagers, including semi-independent branch houses (kadoya), into goningumi. If we have missed
anyone, we understand that we shall be punished according to the shogunal law.

― 356 ―

Appendix 3
Goningumi Rules, 1662 (Shimo-Sakurai, Kita-Saku District, Shinano)
1. Re : Christians. Following the investigations of the past, each and every one, down to the last person, has been thoroughly
examined: not only house owners, [but also] men, women, children, servants and semi-independent branch houses (kadoya),
renters, fully established branch houses (kakae), down to Buddhist monks, Shinto priests, yamabushi, ascetics, mendicant
monks with flutes or bells, outcastes, common beggars and hinin, and so on—to make sure that there is not a single Christian
in the village. If villagers see or hear about Christians, regardless of whether they are in the village or somewhere else, they
must report

This text can be found in Ichikawa Yuichiro[*] , Saku chiho[*] , "Shiryoshu[*] ," 3-8. Ichikawa lists six identical goningumi laws
from Kita-Saku, dated between 1635 and 1760, for other villages from bakufu domains, bannermen fiefs, and the Tokugawa
Kofu[*] domain (among them one for Makibuse dated 1687); and another four that have supplemented the former goningumi
law with the same number of articles (pp. 17, 29). The translation is based on Ichikawa's text.

Hozumi Shigeto[*] , in his Goningumi hokishu[*] zokuhen, vol. 1, has the same text, with some slight variations, for Shimo-
Sakurai and two other villages, one from Omi[*] province; for Shimo-Sakurai he has seven texts, the last one from 1869, all of
them consisting of virtually the same fifty-six articles. This means that the same goningumi law governed village life in some
locales for nearly two hundred years.

The basic text can be found on pp. 96-103. There are only two remarkable differences between Ichikawa's and Hozumi's
versions. First, in the addition to art. 38 Hozumi's text does not mention the headman as the source of exploitation because of
interest rates he charged peasants at the end of the year on funds he advanced for corvée expenditures. Second, kawata are
not mentioned in Hozumi's text; where they should appear, two character spaces are left blank. I do not know whether this
reflects the original texts or official Meiji/Taisho[*] "sensitivities" (probably the latter).

― 357 ―

them to the authorities at once and receive the appropriate reward. If intentionally no report is made and then someone else
reports it, not only the headman and the goningumi but the entire village will be punished.

Add : Those employing people that move in from the outside or bond servants (genin) should investigate their religious
affiliation and secure a certificate from the Buddhist temple where they are registered.

2. Christians, obviously, but also murderers, thieves, or other rascals should be investigated, and all the villagers should arrest
them; if it is beyond the community's capacity [to do so], the village should see to it that they do not escape, and this should be
reported without delay.

3. Masterless samurai (ronin[*] ) of unknown provenance should not be put up; the headman and goningumi should
investigate them, and if they are certified and guaranteed by relatives or in-laws, they should arrange to seek permission from
the shogunal authorities.
Add : This applies also to people moving in from the outside.

4. If there are people who have killed someone, no matter whom, or strangers who hang out about shrines or in mountain
forests, the villagers, together with people from neighboring villages, should arrest them, tie them up, and hand them over to
the authorities. If it is difficult to catch them on the spot, then you have to pursue them, however far, and catch them where
they settle. But no matter what kind of people they are, you cannot kill them.

5. If evil parties come fleeing to the village from out of nowhere, you have to detain and report them quickly.

6. Gambling, lotteries, and all kinds of games are forbidden.

7. Traffic in humans is forbid. den, and terms of indentured servants cannot exceed ten years.

8. The permanent sale of paddies and fields is forbidden; sales should be made for fixed terms, and contracts must have the
seals of the headman and goningumi.

9. As a precaution against thieves, a lookout post should be built in every village in a good spot for night watches. If a thief is
discovered, whether in the village or in a neighboring village, one should shout out loud, and everyone should gather and
catch him without negligence following the usual arrangement of headmen and peasants.

Add : One should always investigate very well whether Buddhist monks, Shinto priests, yamabushi, ascetics, mendicant monks
with flutes

― 358 ―

or bells, outcastes, common beggars, hinin, and so on, shelter thieves or are among their number.

10. Re : Peasant dress. Headmen can wear woven silk and cotton cloth, other peasants only the latter; silk cloth cannot be worn
even in collars or sashes, and nothing can be dyed purple, crimson, or plum.

11. Palanquins cannot be used by headmen or peasants, male or female.

12. All luxury items are prohibited. Trespassing one's status is not allowed even in religious ceremonies, festivals, rituals, and
so on.

13. Apply your mind to cultivation; the yearly tribute, as ordered, should be paid in full promptly before the fifteenth of the
twelfth month.

Add : Pay the kuchimai tax as decided earlier.

14. If peasants seem set on absconding without paying their tribute, the goningumi should investigate, consult with the
headman, and stop them; if they are negligent and do abscond, their arrears should be made up by the goningumi and a
search organized.

Add : When it obviously concerns a peasant who cannot make ends meet and has not paid his taxes in full, then first the
goningumi and next all the peasants will be ordered to pay the tribute for him. This has to be done in full by the due date.

15. Re : Shogunal business. No matter who transmits it, the business should be carried out without the slightest delay; circular
letters have to be delivered promptly to their next destination.

16. People who are not engaged in cultivation, trade, or any other occupation, or withdraw from consultation with the village,
or like quarrels and lawsuits, or do all kinds of bad things should not be hidden [but reported].

17. No peasants should engage in banding together to plot evil things.

18. In the case of quarrels and disputes, the locals have to gather, put a stop to them, and settle the matter; if no private
compromise can be reached, both parties have to report their complaints. Of course, if injuries occur, the parties involved
have to be restrained immediately.

Add : The dead whose provenance is not known need to be reported.

19. Bad spots not only on highways but also on roads and bridges have to be repaired without negligence.

20. No killing of life [hunting] is allowed except by those who have a permit.

― 359 ―

21. Expanding fields by making ditches or roads narrower is forbidden.

22. Under no circumstance can absconded peasants and fugitives be sheltered, not even by relatives or in-laws.

Add : Buddhist monks, Shinto priests, yamabushi, ascetics, mendicant monks with flutes or bells, common outcastes, beggars,
hinin, and so on, and strange persons cannot be rented rooms even for one night.

23. Aside from those who come for business, people who often come from elsewhere without having some business in the
village cannot be put up.
24. When one has to stay overnight somewhere else, whether on business or on a pilgrimage, one has to report the details to
the goningumi.

25. If a single peasant cannot work the fields because of illness, the whole village, starting with his goningumi, has to engage in
mutual assistance, work the fields, and pay the tribute.

26. No peasant, no matter how poor, can arbitrarily be banished.

27. One cannot force a well-established peasant into bankruptcy and then add his fields to one's own; even in the case of death
without a successor, one cannot add those fields to one's own without first seekig permission from the shogunal authorities.

28. That the fixed tribute has to be paid on the main paddies and fields goes without saying, but the same is true for waste
fields and newly developed fields: all, down to the last bu [four square meters] have to be reported and taxed. No matter how
low the quality of a piece of land, it should not be left waste.

Add : If there is grassland or old wasteland that should be turned into fields or paddies, permission for its development ought
to be secured from the shogunal authorities.

29. Before one turns a paddy into a field because the plot is unfit for wet cultivation, shogunal permission has to be sought.

Add : Permission is also needed to convert fields and paddies into a homestead, or vice versa.

30. In the event of fire within the village, everyone has to rush to the storehouse to protect it. Of course, if the storehouse is
safe, everyone should go to where the fire is, and everyone should always be very fire-conscious.

31. Trees and bamboo in the shogunal forest cannot be cut; this applies also to undergrowth, bushes, and grass.

― 360 ―

Add : Peasants should not lay waste bamboo and other trees around their house.

32. Favoritism by intendants' assistants (tedai) and irregularities all the way down should be reported.

33. No presents whatever should be given to assistants.

Add : It goes without saying that under normal circumstances, and a priori, when intendants' assistants visit the village for
official business, the peasants should not entertain them.

34. When one is to supply men and horses for any official business, one has to secure a note from the intendant's assistant
before proceeding; also, any official document has to be signed by the assistant and bear the headman's seal.

35. At the time of the allocation of tribute, all the peasants, including those with fields in other villages, shall gather, reach
decisions together, and affix their seals to the document to signify that they have seen it and deposit it in the headman's office.

Add : The peasants should also have a copy of the document.

36. When the tribute document, based on a visual inspection of the fields by the shogunal authorities, arrives [from the
intendant's office], all the peasants have to gather, divide the amount due, and affix their seals to a proviso saying that they
have seen and examined the allocation and return it to the intendant's assistant.

37. When the rice and cash tribute registers are closed, the headman and four or five small peasants have to affix their own
seals and secure the seal of the intendant's assistant. Each time someone pays his tribute, he shall affix his seal to the register
under his name and be given a receipt by the headman.

38. None of the items that comprise the corvée cash tax (busen) can be lumped together with the allocation of rice tribute.

Add : It happens that headmen do this, advancing money and then charging interest on the amount. This causes hardship to
the peasants; therefore, corvée cash tax should be allocated at each occurrence and settled at that moment.

39. When tribute rice is stored, the storehouse has to be padlocked and sealed with the seals of the intendant's assistant, the
headman, and the peasants. The assistant keeps the key, and when payments are made, he has to come and open the
storehouse. Of course, if any rice is handed over to someone, a receipt with the amount has to be recorded.

40. In the case of a minor as head of a household, the relatives,

― 361 ―

headman, and goningumi have to confer and through a deed entrust the paddies and fields to someone and have that person
pay the tribute and perform corvée. When the head of household reaches adulthood, the entrusted goods will be handed back
in toto to the person in question, who then will become [registered as] a peasant.

41. A name register has to be presented [to the authorities].

42. The peasants should have a copy of the name register, and the tribute allocation should be specified in detail.

43. Not only all accounts of the intendant's assistant, headman, and the peasants but also the smallest transactions have to be
balanced with the bills.
44. Re : village headmen. Since all the peasants in consultation put up a man they verify to be reliable, if irregularities occur
with the balance of the tribute or other matters, responsibility falls on all of them.

45. Suspicious expenditures related to corvée cash or any other matter and allocated by the headman should be investigated
immediately in every case.

46. What goes for disputes with people from other places applies also to difficulties among fellow peasants: the headman and
the titled peasants (osabyakusho[*] ) have to conduct a joint investigation and settle the matter. They have to report cases in
which no reasonable private settlement can be reached; favoritism is not allowed, not even for relatives and in-laws.

47. No negligence is tolerated in one's rotation duty to guard the storehouse.

48. No owning Of guns is allowed except by those licensed for official use.

49. The storehouse should not leak; obviously, attention should be paid to the walls and even to the matting.

Add : In stormy weather, the storehouse should be checked and any damage reported to the intendant's assistant and repaired.

50. When an intendant's assistant passes through the village on official business, one horse and one boatman will be made
available to him. Without permission, men or horses cannot be put out for anyone else. Of course, if an assistant passes
through on private business, then he has to be charged the fixed rate for a pack horse.

51. At times, in some places, there are horse thieves about. If there are suspicious-looking people passing through with roped
horses or cattle, whether by day or by night, one should inquire about their des-

― 362 ―

tination, and if incongruencies are noted, one has to correctly inform the village and subsequently the headman and
goningumi of their destination.

Add : No cattle or horse trading should occur without a bona fide go-between.

52. If it is reported that people of the village or from another locale have seen or heard about a thief's stored goods, the
headman and goningumi have to get together without delay and investigate the matter.

53. If in some way things have been taken by thieves, or things have been found, or suspicious things occur, they have to be
reported.

54. All peasants should endeavor to take the shogunal rules to heart and always investigate to keep things on the right track so
that lawsuits do not develop.

55. Replacement of seals affixed to this document is not allowed; the same seals must be used for all other deeds. If a seal gets
lost, the owner presents the new one to the intendant's assistant in charge of certifying seals with a note indicating the date of
replacement, and he affixes it to the goningumi roster.

56. A copy of this document is deposited with the headman; it should be read aloud whenever the peasants meet in council.
The shogunal law should be kept in everything. The above articles are not to be deviated from in the slightest. Therefore, they
have been examined in the presence of the headman and the peasants, and the goningumi present this document to the
authorities. If anyone breaks these rules, then, starting with the goningumi, the headman and the titled peasants can expect
any punishment. In witness whereof, we set our seals hereunto.

Addressee: the intendant.

[Date.]
[Signatures, seals.]

― 363 ―

Appendix 4
Regulations for the Villages of All Provinces—The Keian Edict, 1649 (and 1665)
Introductory Comment

This is the famous Keian Edict of 1649 (Keian 2).[1] The translation is based on an expanded version (the expanded portions
given here in italics) dated 1665, from the Mikage intendant's office in Kita-Saku, which Ichikawa Yuichiro[*] found in the
house of the descendant of the headman of (Maki)buse village.[2] The bracketed second sentence of article 5 (against buying
firewood) is the only section from the 1649 document left out in the 1665 version; it is replaced by the long section enjoining
peasants to manage things so that they can sell firewood.

Recently some scholars have questioned the authenticity of this important edict. Maruyama Yasunari argues the most radical
thesis, that this "forgery" was produced in the second half of the eighteenth century.[3] According to him, at that time some
intendant would have felt the need to shore up the status and authority of titled peasants and wrote an exhortative tract, the
Hyakusho[*]mimochigaki , which was compiled from early and mid Tokugawa edicts, goningumi laws, and educational
booklets. This tract, which became quite well known, was expanded and eventually reworked as the famous Keian Edict. And
it was as such that it found its way into the Tokugawa jikki and the Tokugawa kinreiko[*] .
[1] TKKz 5:159-64 (no. 2789).

[2] Ichikawa Yuichiro[*] , Saku chiho[*] , "Shiryoshu[*] ," 21-26.

[3] Maruyama Yasunari, "'Keian ofuregaki' ron no sui-i to sono sonpi o megutte," in Kinsei-kindaishi ronshu[*] , ed. Kyushu[*]
Daigaku kokushigaku ken-kyushitsu[*] (Yoshikawa kobunkan[*] , 1990). Fukaya Katsumi refers to this text (see the following
note).

― 364 ―

Fukaya Katsumi, on whose refutation of this thesis I rely, presents the Maruyama argument as follows.[4] The Keian Edict does
not figure in two famous collections of Tokugawa edicts. The edicts of 1642 and 1643 are reflected in goningumi laws, but there
is virtually no trace of the Keian Edict in these laws. The Keian Edict does not touch upon the proscription of Christianity or
the prohibition of the sale of land; it is not often quoted in manuals for local administrators (the few exceptions being the
Bokumin kinkan , the Jikata hanreiroku , and the Minkan seiyo[*] ), and it is silent on sartorial distinctions between peasants
and headmen, which figure in many goningumi laws of the time. The extant versions are both rare and late (nineteenth
century).

Fukaya points out, however, that no revisionist has explained why a forgery should be given the date Keian 2.[5] In addition,
the absence of any mention of the office of peasant representative (hyakushodai[*] ) in an edict allegedly on village governance
may indicate that it was composed before that office existed.[6] His main argument, however, rests on a close analysis of the
content.

The Keian Edict is an agricultural treatise, not a manual for local administrators; most of its articles are addressed directly to
the peasants.[7] Its emphasis, as the analysis in chapter 4, above, also indicates, is on individual responsibility and economic
improvement. Unlike most Tokugawa edicts, it is not prohibitory (and hence does not figure in collections of such edicts).[8]

Text

1. One should obey the shogunal law, respect district officials and intendants, and consider headmen and goningumi heads as
one's real parents.

2. Those functioning as headmen and goningumi heads should respect district officials and intendants, pay tribute in full,
observe the shogunal law, and see to it that the small peasants behave properly. If

[4] Fukaya Katsumi, "Keian ofuregaki no kosei[*] to ronri," in (Ronshu[*] ) Chukinsei[*]no shiryo[*]to hoho[*] , ed. Takizawa Takeo
(Tokyodo[*] shuppan, 1991), 345-72; see 346 and 348.

[5] Ibid., 349.

[6] Ibid., 370.

[7] Ibid., 350.

[8] Ibid., 345.

― 365 ―

headmen and goningumi heads fail to perform their duty or are in poor standing, they will not earn the respect of the small
peasants, who will not obey even shogunal orders. Therefore, headmen and goningumi heads should always behave properly
and try to be successful in life.

3. Headmen should not mistreat people, including those with whom they do not get along well; no favoritism should be shown
to anyone, no matter how close a friend. Headmen should be close to the small peasants and allocate tribute and corvée to
each of them equally and fairly. On the other hand, the small peasants should not disobey the orders of headmen and
goningumi heads.

4. Apply your mind to cultivation; if you take as good care of paddies and fields as you would of trees, keep fields from
becoming grass-covered by weeding them often, and plow the land often between cultivations, you can expect a promising
crop and a rich harvest.

Add : Plant soy and adzuki beans and such along the edges of fields; you should cultivate even the smallest empty plot.

5. To go buy and drink sake or tea is prohibited; this applies also to wives and children. [Plant bamboo around the homesteads,
collect fallen leaves, and see to it that you do not have to buy firewood.] It makes sense to plant saplings in the village and grow
a forest and bamboo grove. At first this may be troublesome for the peasants. However, the fallen leaves and wood can be used for
various things, such as digging wells, cleaning rivers, building houses, and making bridges. Furthermore, in the end, when the
forest is full-grown, one can sell firewood from it and make a profit. Moreover, economically pressured earners can have tools
made for peasants to build houses. One who makes it in this business can be successful for generations, passing his wealth on to
his offspring, because peasants, unlike officials, who change positions, continue living in the same place. If one plants mulberry,
lacquer, and oak trees at the borders of paddies and fields, one can also plant tea. In twenty or thirty years this will result in many
tea groves and full-grown mulberry, lacquer, oak, and other trees in the mountains. This will bring great [economic] ease to the
village. Therefore, the headman and goningumi beads should consider planting trees as often as possible .
6. All seeds should be selected carefully at the beginning of the fall season. Only good seeds should be sown; otherwise one
cannot expect a rich harvest. Also, it is important to plant rice seedlings suitable to the soil. After rice has grown and been
harvested, the ears should be

― 366 ―

checked for the number of grains. Check the possibility of single- and double-crop farming, which depends on the quality of the
soil: some fields are good for soy and adzuki beans, others for foxtail millet, barnyard grass (millet), greens, radishes, and
potatoes. It is important to consult often with skilled farmers, raise the crops in season, and harvest them on time .

Add: Some crops need humidity, others don't; take this into consideration, and you will be able to harvest even from low-grade
paddies and fields .

7. One should sharpen hoes and sickles every year before the eleventh day of the first month. If paddies or fields are cultivated
with a dull hoe, little progress will be made. The same is true for dull sickles.

8. A peasant must take special care in preparing manure and ash; therefore, he should make a large night-soil reservoir in
such a way that it keeps the rain out.

Add : If a couple lives all alone and cannot afford to make a toilet, they will have no night-soil reservoir (pot). In that case, they
should dig a hole in the yard, three shaku [about one meter] wide and two ken [about four meters] deep and relieve
themselves there; they can also add weeds from the roadside and lead water into it to make manure and use it for cultivation.

9. Peasants lack judgment and do not think ahead; hence they carelessly feed their wives and children rice and other grains
when these are harvested in the fall, but they should always take as good care of food as they do in the first, second, and third
[winter] months of the year.

Add : While cereals are the main products, wheat, foxtail millet, barnyard grass, greens, radishes, and any other grains should
be planted so that rice is not depleted by eating it in great quantities. If you keep famines in mind, then the leaves of soybeans,
adzuki beans, cowpeas, and even the fallen leaves of potatoes will not be wasted.

10. [Everybody], from the house owner down to his children and servants, and so on, should normally eat plain food.
However, at times of demanding labor, like planting rice seedlings or harvesting, slightly better food should be made available
in large amounts. If one shows such consideration regarding food, everyone will be content and work hard.

11. Wealthy peasants, such as those who employ seventy or eighty male and female servants, often limit themselves to giving them
orders

― 367 ―

about work but leave everything to their servants' judgment and do not check things for themselves. While they are well-off, their
weak points do not reveal themselves, but when their ease and comfort have been taken advantage of and their wealth has
declined, they are forced to reduce the number of servants, and finally they have to cultivate by themselves. The comfort they bad
in the past now kills them. Keeping this in mind, they should, as a rule, get up in the morning, wake up the servants, and send
them to weed during the morning; after they return, take them to do field work and be with them to direct them. Then the servants
will apply themselves to their work with diligence .

12. For instance, when one goes out to the field and feels that it might rain the next day, one should be engaged in harvesting rice
or wheat, sowing in the dry fields, or doing something first that normally cannot be done on a rainy day. This sort of planning
should be done on a daily basis so that one can work without delay: on a rainy day one cultivates or mends paddies, and on a fine
day one cultivates or weeds fields . In the evening one should plait rope or weave straw rice bags. Be always diligent in
whatever you do.

Add: When servants have to do back-breaking work, show concern for their bard work and give them some time off .

13. One should make efforts to have a good-quality bull or horse. The better the bulls and horses, the better the manure.
Needless to say, the poorer peasants should take this to heart. Furthermore, these animals need solicitous care (fukaku
awaremu). Think about them once in a while. They cannot ask for water when thirsty or food when hungry. They cannot complain
about the pain in their backs from a rough saddle even when it penetrates to the marrow. Sometimes they collapse near death
under a heavy load on a mountainside. Since they cannot throw off their burden by themselves, their suffering must be
unspeakable. There is a saying that even an ant's wish is granted by Heaven, but a horse or bull's wish must be far more profound
than an ant's. It is said that sometimes people get seriously ill or even die from an unexpected disease because of a curse (tatari)
from horses or bulls. Furthermore, when a poor peasant kills a horse, which costs three bu to one ryo[*] , he cannot compensate
for the loss even after five or seven years. In fact, nobody gains from shrugging this off with the saying that this is the horse's or
bull's karma and my loss. Therefore, make saddles comfortable for horses and feed them well. If poor peasants cannot afford to
feed them ample fodder and soybeans, they can simply take good

― 368 ―

care of them and understand their suffering. Sometimes young children and servants may not feed them as they should, although
they may tell their parents and master that they have fed the animals. In that case, the poor animals have to be without food all
day and all night. This may happen often if one does not consider feeding them rice bran and fodder oneself and personally
checking the stable often. It is regrettable to neglect the animals, when they contribute to support the peasants with their
endurance . Also, in early fall a peasant should prepare food to be given to the horses and bulls during the spring season.
14. Both husband and wife should work for a living: the husband in the fields, the wife at the loom and preparing the evening
meal. If a wife neglects her husband, drinks a lot of tea, and likes pleasure trips, she should be divorced even if she is good-
looking. The only special cases are when a couple has many children and when a husband has a prior debt to his wife.
Furthermore, a husband should take good care of a wife who is not good-looking but is devoted to him.

15. The shogunal law must be observed in everything. Specifically, a criminal whose provenance is not known cannot be
patronized (kakaeoku ) in the village. If someone secretly shelters criminals such as burglars or lawbreakers and this is
reported, he will be taken to the shogunal office and examined. The entire village will be put under severe strain if the
investigation takes long. Also, one should behave honestly and faithfully so as not to be scorned by the headman, goningumi
heads, titled peasants, and the entire village.

16. Peasants can wear cotton cloth only; nothing else can be used even in sashes and collars.

17. One should try to raise one's living by becoming a little business-minded. That is to say, when one sells or buys grains to
pay the tribute, one will be taken in if one does not have a business mind.

Add: There is an old saying that it is better to draw water to each field near you rather than attempt business far away. If water is
drawn to each field, it will fertilize the soil, make the rice stumps rot, and eventually produce a rich harvest .

18. Apart from the rich, the poor who do not own large fields should think well about a means of living throughout the year;
for instance, if there are many children in a family, some can be given away [for adoption?] and some can be sent out as
servants.

19. Peasants should learn early not to spoil their children with love. They should know that too much love makes them play and
be

― 369 ―

worry-free. When children grow up, they may not know how to cultivate fields, or they may associate with outlaws and cause
trouble to their parents. Given the instability of this world, they may be separated from their parents when still young. Then they
will be frightened like monkeys that are separated from the trees. They will try to perform their duties on the fields and budget the
housekeeping, but because they lack training in such things, often they will lose everything and never attain old age. If one truly
loves one's children, one must nurture their individual talents when they are young, and when they grow up they should be taught
housekeeping. They have to be made to get up in the morning and check around the fields with the servants. Indeed, if they can
experience rough times when they are young, they will not be worried when a parent suddenly dies. Then that will be the parents'
great blessing to their offspring .

20. The yard should be kept clear and face south. This is because if the yard is not clean, when a peasant winnows rice and
wheat, threshes soybeans, and handles cereals, soil and sand will be mixed in with these grains. In that case, he will only get a
low price for them, and he has no other resources.

21. It depends on the location, but [as a rule] one should underestimate the area of a prospective wheat field. Then, if one
expands wheat fields one after another, they will bring a greater profit to the peasant in the end. Once a village succeeds in
developing land as wheat fields, the neighborhood villages will follow this example.

22. One should always attend to one's health by getting a moxa treatment in spring and autumn. No matter how hard one is
willing to work at cultivation, if one fails ill, one cannot be engaged in work that year and eventually becomes bankrupt. One
should give this the greatest consideration. The same applies to wives and children.

23. Smoking is prohibited. Tobacco is no substitute for food; on the contrary, it causes illness. In addition, it takes time and
money and constitutes a fire hazard. It is disadvantageous to everyone in every respect.

24. Regarding the allocation of tribute: the amount is allocated either per tan in the case of tan units or per koku in the case of
taka units, and the tax roster is issued by the intendants. Accordingly, peasants should apply themselves earnestly at
cultivation. If they reap a rich harvest, it becomes their profit. If they harvest poorly, it will be their loss.

― 370 ―

25. Regarding full payment of tribute: when they have no other alternative to make up for the shortage of rice, peasants often
borrow from other villagers in the amount of five or six sho[*] to one to [between 10 and 18 liters]. At the time the tribute is
due, everyone becomes short of rice and cannot loan to others. It does not make sense to sell a child, horse, or bull for a small
amount of rice, such as five sho[*] or even one sho[*] . Although good tools and cutlery can be sold, it is loathsome to sell them
for one bu or five or six sho[*] of rice. Some without anything to sell have to borrow at a high interest rate, and then their loss
becomes larger. At the time of the allocation of tribute by the intendant, peasants should estimate the amount of tribute and
borrow the shortage in advance. Getting an advance results in a low interest rate, and then selling something for a loan
payment is easier. It is better to pay the tribute promptly. If one waits to pay, it may result in a large loss due to rats, theft, fire,
or any other damage. One should dry the unhulled rice completely and thresh it to grains; otherwise rice often gets cracked.
Keep this in mind.

26. If one has to borrow rice, say, two bales, because one failed to act properly or in order to supplement one's yearly tribute
payment, after five years the principal and interest of that loan will be fifteen bales because of the yearly compounded
interest. It will be over one hundred fifty-five bales within ten years , and by that time one will be bankrupt. One will sell one's
wife and children and even oneself, and future generations will suffer for it. One must consider this well and act properly. At
first two bales of rice seems little, but the interest adds up yearly and becomes [an enormous amount], as seen above. On the
other hand, if one manages two bales of rice by oneself, with the interest one will have one hundred and seventeen [fifty-five ]
bales of rice ten years later. Peasants should not be negligent in this.
27. Both people working in the mountains and those working at the seashore should apply their minds and work hard every
day without negligence. It may happen that one is not able to work because of bad weather, rain or wind, or long illness; one
therefore should not waste one's earnings.

28. Mountains and seashores are sparsely populated; therefore people can make a [good] living there. People in the mountains
bring down firewood and timber and sell citrus fruit. People at the seashore produce salt and catch fish and always make a
good living selling them. But they tend to spend their earnings then and there, thoughtlessly,

― 371 ―

without regard for the future. It is said that in time of famine they suffer much worse than villagers and that more of them die
of starvation. They should always keep in mind sufferings of times of famine.

29. When a peasant who is single cannot cultivate his fields because of the time he has taken off or because of illness, his
goningumi should help with cultivation and maintain the condition of the fields. Furthermore, if a single peasant has plowed
the paddies, picked up and prepared seedlings, and planned to bed them out the next day and is then appointed for corvée at
the intendant's office or for some other official duty, he will have to take time off from his work for three or five days. The rice
seedlings that he had already taken out of their beds for transplanting will turn bad. Other plants will also become knotty, and
because he missed the right time for planting, his rice will not grow well that year. He will have a poor harvest and then
become bankrupt. This is also true for crops in dry fields. If the opportune time for sowing or planting is missed, crops will not
grow well. The headman and goningumi head should take this into account, and if a peasant who lives alone is assigned an
official task, they should try to find a substitute for him from among a wealthier peasant's servants.

30. A husband and wife who live all by themselves and have a hard time making a living are the daily object of the villagers'
scorn. If they then start to raise their living, obtaining much rice and money, they will begin to be treated nicely in word and
deed. The headman, titled peasants, and then others will address them politely, upgrade their seating rank [at meetings], and
entertain them. In contrast, once a rich peasant fails in his business and becomes poor, he is disdained and not spoken to by
anyone, including his family, relatives, headman, and goningumi heads. Therefore, one should behave well and see to it that
one makes a good living.

31. If there is one person in the village who works hard at cultivation, behaves well, and is wealthy, all will learn from him and
apply themselves better. If there is one village like this in a county, the entire county will become wealthy. Consequently, a
province becomes rich, and then the neighboring provinces will be influenced by this. Intendants shift office all the time, but
peasants leave their fields to succeeding generations and make a good living. If peasants work hard and attain wealth, is that
not their success and merit? On the other hand, if there is one lawbreaker in a village, the entire village is affected and
peasants and servants will engage in endless quarrels. If one person breaks the

― 372 ―

law and is taken to court, high and low will have to work harder, and the village will have heavy expenses to bear. Everyone
should take care not to get involved in any violation or crime. Headmen should keep this in mind and inculcate the peasants.
The peasants should entertain good relations with neighboring villages and not engage in suits with people from other
domains.

32. Even a stranger of unknown provenance can be kind to us; [therefore,] it is not filial piety if one is devoted to good parents.
True filial piety is to show great devotion to parents who are unreasonable, immoral, and selfish (murihido[*] nite wagamama).
Filial piety means to be well-behaved, have a good reputation, and as brothers have good relations . The older brother takes care
of his younger one, who is obedient to him. Moreover, one can make one's parents happy if one does not go buy and drink
much sake, is not quarrelsome but obedient from the bottom of one's heart. If one observes this, gods and Buddhas will bestow
their blessings, the Way will be observed, fields will be cultivated, harvests plentiful. On the other hand, no matter how
devoted one may be toward one's parents, it is difficult to be filial when in misery. Therefore, one should try hard to be
successful. Poverty may cause illness, warp one's mind, make one steal or break the law, and one will wind up tied up,
arrested, put into a prisoner's sedan basket, receive the death penalty, and be crucified, and so on. How painful this must be
for the parents! Moreover, one's wife, children, brothers, and lineage will feel shame and grief. One should ponder this day
and night, behave well, and do nothing unbecoming.

33. If, no matter how hard one works or how much gold, silver, rice, and money one has or how good a homestead, one
violates the shogunal law or does not observe the intendant's regulations, one will be arrested and sued. Also, one who drinks
much, is violent-tempered, quarrelsome, and selfish will go bankrupt. Be sure to be merciful and honest and not violate the
shogunal law. Headmen should stress this a great deal with the small peasants, and parents with their children.

34. Village bridges and roads must always be kept in good condition. If they are bad, travelers will be inconvenienced, people
and horses from the village will be troubled, and this will result in mishaps. Headmen and goningumi heads should always be
concerned with this and use slack times or holidays to fix them.

35. If we are right-minded, no one treats us badly. If someone treats us badly, that is because our heart is not in the right place.
The

― 373 ―

same is true for all relationships between intendants and peasants, masters and servants, parents and children, husbands and
wives, fellow peasants, and headmen and the peasants. If the headman and the peasants respect the intendants, are well-
behaved, do not violate the shogunal law, and apply themselves to everything, the intendants will not harbor any bad feelings.
Keep this respectfully in mind every day and do not oppose others. In all these matters, take good care and apply yourself
earnestly. When one's living improves and one comes to have rice, silver, and grains, one can afford to build a house and buy
clothes and food as often as one wants. Even though one has lots of rice, gold, and grains, the intendants will not take it away.
Since this is a time of great peace in the realm, no one will forcibly take it away. One can, therefore, live comfortably for
generations, and even during famine will one be able to take care of wife, children, and servants. Nobody lives better in peace
than a peasant who has paid his taxes. You should understand this well, teach it to your children and grandchildren, and make
a good living.

― 374 ―

Appendix 5
Regulations for Outcastes in Various Jurisdictions in Shinano
I. Komoro Domain, Kita-Saku District (1738)

1. Obviously, kawata should be courteous and avoid any insolence toward servants in warrior households, but they also
should not enter commoners' houses or shops; in matters of fires, quarrels, and disputes they should follow instructions.

2. They should not hassle for bargains in the towns or countryside, and they should not frequent teahouses or places that sell
sake or be self-indulgent (wagamama).

3. They should not wear short swords all the time, hut only, following directions, when they are on duty.

4. They cannot enter the castle to sell sandals or enter homes.

5. They cannot act selfishly (wagamama) when they venture into the countryside; when women sell needles in summer and
fall, they should do so properly and not hassle people.

Add : It is strictly forbidden for them to go to the towns or countryside and pressure people into lending them money.

6. They should settle quarrels and disputes and apprehend pickpockets at festivals, exhibits, rituals, shows, and so on, but
should not pass as ordinary commoners (heijin) and mingle with crowds of onlookers; we have heard that they behave in a
self-indulgent and rude manner, and therefore they should not mingle with commoners except in their official capacities;
women and children at shows should stand beside the eta and not hinder the heijin....

Ozaki, Shinshu[*]hisabetsu buraku , 73, 244, 322-24, 43-44, 325-26.

― 375 ―

II. Okudono Domain, Kita-Saku District (1738)

1. It is forbidden from now on for eta to wear sandals (zori[*] ) when going to the village; they need to wear straw sandals
(waraji ).

Add : They cannot enter the homes of peasants.

2. They should not be discourteous toward villagers in their dannaba or when going to villages beyond it.

3. As eta, they cannot wear short swords.

Add : They need permission from the village headman for trips, even day trips without overnight stays.

4. It is strictly forbidden for them to go to plays or entertainments.

5. They should not quarrel or fight at festivals, markets, or shopping districts.

6. When they are requested to patrol plays and entertainment in villages of a dannaba in a different domain, the requests and
directives of local headmen should be followed.

7. They should not temporarily put up people who have been banished, have absconded, or are of unknown provenance; this
obviously pertains to persons from other domains, but it also pertains to persons from the villages in the dannaba....

III. Chiisagata District: The Villages of Tatsunokuchi, Koshigoe, Nakamaruko, Iinuma, Mitakedo[*]
(1891)[1]

1. House or storehouse rentals, smoke outlets, and billboards are forbidden.

2. Wearing of short swords on private occasions is forbidden.

3. Umbrellas, geta, and leather-soled sandals cannot he used.

4. Children's forelocks should be kept short.

5. Haori coats or parasols cannot be used.

6. No silk garments of any kind can be used.

7. When going to other villages, you [cannot] look like titled peasants.

8. When meeting murakata [probably "villagers" rather than "officials"], you cannot extend greetings.

[1] The first two villages were part of a bakufu fief from 1704; the other three, part of Iwamurata domain from 1711. Today all
five are part of the town of Maruko, south of Ueda.

― 376 ―

9. When interacting with murakata, you cannot proceed further than the extended eave and enter their homes.

IV. Chiisagata District, Netsu Hatamoto Fief: Netsu Village, (1838)[2]

1. As decreed in past years, you should always be polite, and never rude, toward titled peasants.

2. You should not be rude toward titled peasants on your way to work or to the mountain forests.

3. Never engage in extortion or be rude in your dannaba.

4. When going to festivals, obviously in your own locale but also in others, men, women, and children should dress according
to their status without standing out.

5. Modesty is required in hairdos for women and children.

6. No entertainment or musical accomplishments are allowed.

7. New-style houses or multiple front entrances are forbidden.

8. No cutting of standing trees is allowed.

9. No building of new houses, not even one, is allowed.

10. Backroads are to be used for going to the plains and fields.

11. No taking of more than one sheaf is allowed; in other domains, however, you can do as you please.

12. When leaving the village, no umbrellas, geta, leather-soled sandals, lining, or tabi are allowed, only straw sandals.

13. No streamers or decorations are allowed at the Boys' Festival on the fifth of the fifth month.

14. When going to other places, you cannot look like titled peasants, but should set yourself apart from them; this applies to
both men and women.

15. Clothes with family crests are absolutely forbidden.

16. Momohiki (drawers) and fine patterns are not allowed for titled peasants and certainly not for eta.

17. As ruled before, obviously in the dannaba but also elsewhere, neither old nor young, neither men, women, nor children,
can proceed beyond the place under the protruding eaves into the homes....

[2] These rules were issued in the context of a reform that stressed frugality, as explained in the preamble (not translated). As
Ozaki remarks (327), this is really a transference, even in the language used, of warrior prescriptions for proper behavior of
peasants toward them to the manner kawata should adopt toward peasants.

― 377 ―

List of Characters
Note: The characters for names of authors appearing in the bibliography are not included .

Aida

Ainosho[*]

Aizu

Akashina

Aki

Akita

Ako[*]

akunin