Everyday Things in Premodern Japan: The Hidden Legacy of Material Culture [Reprint 2019 ed.] 9780520922679

Citation preview

Everyday Things in Premodern Japan

Everyday Things in Premodern Japan The Hidden Legacy of Material Culture

Susan B. Hanley

U N I V E R S I T Y OF C A L I F O R N I A PRESS Berkeley / Los Angeles / London

University of California Press, Ltd. London, England © 1997 by The Regents of the University of California First Paperback Printing 1999 Library of Congress Cataloging-in-Publication Data Hanley, Susan B., 1939Everyday things in premodern Japan : the hidden legacy of material culture / Susan B. Hanley. p. cm. Includes bibliographical references and index. ISBN 0-520-20470-0 (alk. paper) 0-520-21812-4 (pbk. : alk. paper) 1. Material culture—Japan. 2. Japan—Social life and customs. 3. Japan—History—Tokugawa period, 1600-1868. 4. Japan— History—Meiji period, 1868-1912. I. Title. GN635.J2H35 1997 3 06'.0952—dc20 96-33421 CIP

Printed in the United States of America 9 8 7 6 5 4 3 2 1 The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984. ©

To the memory of my mother, Frances Alden Hanley, who encouraged me to follow any career I wanted, and to my aunt, Susan McCallum Haskins, who has not only shown me what a woman can achieve in life, but who has fully supported my academic career from my student days and made my study ofJapan possible.

Contents LIST OF ILLUSTRATIONS

ix

PREFACE

xi

NOTE ON TRANSLITERATION 1 The Level of Physical Well-Being in Tokugawa Japan

XV

I

2 Housing and Furnishings

25

3 A Resource-Efficient Culture

51

4 A Healthful Lifestyle

77

5 Urban Sanitation and Physical Well-Being

104

6

Demographic Patterns and Well-Being 7 Stability in Transition: From the Tokugawa Period to the Meiji Period

129

155

8

Physical Well-Being: A Comparative Perspective

176

GLOSSARY

199

INDEX

205

Illustrations

Figures 1. Nineteenth-century formal room with shoin elements 2. Nineteenth-century house with kamado in the doma and living room with irori 3. Women wearing kosode

32 64 69

4. Wooden bath 5. Drainage in a town street 6. Nineteenth-century toilets

98 117 123

Tables 1. 2. 3. 4. 5.

National population data by region Life expectancy for sample areas in Tokugawa Japan Life expectancy estimates Percentage of women married, by age group

Percentage of households containing married women in the childbearing ages 6. Sex ratios of last-born children 7. Number of shops in Tokyo, 1881,1891, and 1900

132 134 136 142 144 147 165

Preface

I became interested in history as a child because I was fascinated with how people lived in past times. If the material conditions of life were different, how did this affect people's thought and behavior? When I began to study Japan in college in the early 1960s, two ideas were impressed upon me. The first was how different the Japanese were from Westerners, despite the fact that both Japan and Western nations experienced feudalism and industrialization. The second was that Japan had been a very poor nation in the nineteenth century, one far behind the West in its standard of living. Over the years I find that these ideas have not been dispelled, nor have they been connected to the study of either the material conditions or the culture in Japan in premodern times. By the 1960s scholars in Japan and the West, however, were beginning to question the extent of Japan's economic backwardness in the nineteenth century and to provide evidence of a growing economy during the entire span of the Tokugawa period. My first major research project was to pursue the puzzle of why, if the economy was growing, the Japanese were widely practicing abortion and infanticide as demographic scholars claimed. My eventual answer was that Japanese were controlling their population growth, not for reasons of abject poverty, but in an effort to maintain and improve their standard of living. They achieved this in part through drastic methods such as abortion and infanticide and, more importantly, through nearly universal social controls such as permitting only one son per household to marry, delaying marriage in periods of economic recession, and preventing the establishment in farm villages of new households without a specified amount XI

xii

PREFACE

of land sufficient to support a family. These findings were published in Economic and Demographic Change in Preindustrial Japan, 1600-1868 (Princeton University Press, 1977), coauthored by Kozo Yamamura. My demographic research led me to ponder the question of what life was really like in the Tokugawa period. How did people live who made the demographic decisions I could see in the village household registers I used as sources? If the economy was growing, how did life change in the eighteenth and nineteenth centuries? Was the material culture improving? If so, why do we think of life prior to the Meiji Restoration as backward compared to that of the West? In terms of daily life, how did Japan compare to the early industrializers, especially England and the United States? And what difference did the answers make? Daily life is difficult to study, as I quickly learned. Only rarely do diarists and historians chronicle ordinary, everyday life; it is unusual to discover the journal of a public official who wrote down everything he ate when traveling or the diary of an urban housewife who noted every time the family bath was heated. Because such finds are rare, sources from all over Japan must be used in studying daily life, and for that reason they do not form a good, solid body of evidence for any one region. But partly because the study of daily life proved so difficult, I found it particularly fascinating. Like every other author, I would like to imagine that I have produced a work all of my own, on my own. I was very sure when I started this study that the ideas were mine, growing out of my earlier research. And then I discovered that people all over the world were researching various aspects of daily life, from Fernand Braudel of the Annates school to the American archaeologist James Deetz, to mention two Westerners whose work I much admire, and numerous Japanese and other scholars mentioned here and in my notes. Research and concerns of one generation lead to the questions the following generation tries to answer, and, in many cases, each scholar independently comes to ask similar questions pondered by fellow scholars. So that although I like to think that this study is my very own, I would also like to fully and gratefully acknowledge all the help I have received from many sources. My American mentor who first suggested I study the Tokugawa population encouraged me to do research on material culture and asked me to write a chapter in The Cambridge History of Japan, volume 4, titled "Tokugawa Society: Material Culture, Standard of Living, and Life-Styles" (Cambridge University Press, 1991)- I am deeply grateful to Emeritus Professor John W. Hall of Yale Univer-

PREFACE

xiii

sity for his guidance, encouragement, and the example he set for me as a scholar. My first mentor in Japan was Professor Hayami Akira, under whom I studied historical demography and the reading of the shumonaratame-cho, the primary documents I used for studying Tokugawa population. I now realize how much he and his research group at Keio University influenced me to begin the study of seikatsu-shi, the history of everyday life. I owe a special debt to Professors Wakita Haruko and Wakita Osamu, who helped me in my research on material life from the outset. They went far beyond the role of advisors, not only guiding me to various sources and then helping me read them, but also taking me on numerous site visits throughout western Japan and putting me up in their home countless times. They were responsible for the initial version of this book being published by Chuokoronsha as Edo jidai no iscrn (The Tokugawa legacy) in 1990. I would like to thank Mr. Iwata Gyo, my editor at Chuokoronsha, for his help and his faith in my work. And above all, I am grateful to the translator of the Japanese version, Professor Sashi Akihiro, whose superb translation is largely responsible for Edo jidai no isan winning in 1990 the Joseph Roggendorf Prize given by Sophia University for a work furthering East-West understanding. This study would not have been possible without the kindness of hundreds of colleagues, friends, and acquaintances, only a few of whom I can name here. I am grateful to Professor Yasuba Yasukichi of Osaka University, who helped me formulate and temper my position on the premodern standard of living by engaging me in friendly academic arguments. Ms. Nakano Setsuko of Kanazawa University helped me puzzle my way through original sources of the Tokugawa period, primarily diaries not meant for others to read, much less a twentieth-century American. I would like to specially thank Professors Ishige Naomichi, Hirai Kiyoshi, Okawa Naomi, Kito Hiroshi, Kuwahara Minoru, and Shiraki Kosaburo, and Ms. Koizumi Kazuko. These and many others in Japan have guided me through historical sites and their own houses; have led me to historical sources I would not have found on my own; have explained menus, types of plants used for food, and historical artifacts; and provided over the years various kinds of help that they may not even remember but that have been crucial to my research. I cannot thank enough Mrs. Tanaka Ayako, who has helped me try to understand Japan ever since the week I first arrived in the summer of 1961.

xiv

PREFACE

I would like to acknowledge the following publications that have published various parts of this research as I completed it over the years: "A High Standard of Living in Nineteenth-Century Japan: Fact or Fantasy?" The Journal of Economic History 43, no. 1 (March 1983); "The Material Culture: Stability in Transition," in Japan in Transition from Tokugawa to Meiji, ed. Marius B. Jansen and Gilbert Rozman (Princeton: Princeton University Press, 1986); and "Urban Sanitation in Preindustrial Japan," The Journal ofInterdisciplinary History 18, no. 1 (summer 1987). Some of the tables in chapter 6 are from my Princeton book, Economic and Demographic Change, which is now out of print. The students in my graduate history class critiqued the manuscript, saving me from numerous inaccuracies and inconsistencies, and to all of them I am grateful: Sue Duvall, Jeanne Farrar, Stacey Holmes, Sue Sunmi Kim, Peter Lee, Ethan I. Segal, and John H. Sagers. Amy Williams Middleton kindly took time from her own work to do the illustrations. I am also grateful to two anonymous readers for the University of California Press for their suggestions, and it has been a pleasure to work with my editors, Sheila Levine, Scott Norton, and Carolyn Hill. Special thanks go to Martha Lane Walsh, managing editor of The Journal ofJapanese Studies, who from the beginning of this study has edited and critiqued my work, and to date she has received more argument than thanks from me for the improvements she has made. Finally, I must mention my husband, Kozo Yamamura, who has served as my severest critic, thereby slowing down my progress at many points, but who in the end has helped me produce much better work than I would have otherwise.

Note on Transliteration

Macrons have been omitted from common place names, recent reign names, and Japanese words that appear in English-language dictionaries (except in the Glossary). Japanese names appear in Japanese order, with family name first, with the exception that the names of authors of works in English follow the order given in the publication and the author's preference for use of macrons.

CHAPTER

ONE

The Level of Physical Well-Being in Tokugawa Japan

The Argument

How well did the Japanese live prior to their industrialization in the late nineteenth century? The most widely held view is that Japan was a poor, backward country with a low standard of living, and when it began to industrialize, it had a lower standard of living than did Western countries when they began this process. However, though this view is held by most historians and economists, this view is rarely shared by the scholars who study the lifestyles and material culture of the Japanese.1 A reassessment of how well Japanese lived by the time they began to modernize their economy and the implications for industrialization are the subjects of this study. During the period just prior to Japanese efforts to industrialize, the Tokugawa period (1600-1867), cities grew to include at least one in every ten Japanese, developing into cultural and economic centers from 1. As Kelley and Williamson say, "the model most often employed to analyze early Japan is one where labor-surplus conditions prevail. . . . [which] implies that none of the benefits of modernization accrue to workers in the small 'modern' sector. What is the historical evidence? Was extreme austerity the vehicle by which peasants and laborers in the 'traditional' sector contributed to Japanese progress?" Allen C. Kelley and Jeffrey G. Williamson, Lessonsfrom Japanese Development: An Analytical Economic History (Chicago: The University of Chicago Press, 1974), pp. 22-23.

2

P H Y S I C A L W E L L - B E I N G I N TOKUGAWA JAPAN

which ideas, knowledge, and products were diffused throughout the country. The first urban commoner culture in Japan, which flourished in the late seventeenth century and again in the early nineteenth century, was possible because of the rise in urban income, and the steady transmission of culture and goods throughout the countryside was a result of the rise in rural income. Largely a population of self-sufficient cultivators in the early-seventeenth century, the Japanese by the midnineteenth century were able to buy books, furniture, sweets, fresh fish, hair ornaments, and all sorts of small luxuries even in remote villages. As early as the beginning of the eighteenth century, members of the samurai elite were complaining that the commoners were living above themselves, imitating the lifestyles of their samurai betters. A dramatic increase in consumer goods among the general populace was also found in seventeenth-century England, just prior to industrialization. 2 In light of all that has been written on the positive developments during the Tokugawa period, it is puzzling that some economists consider the standard of living to have been low, particularly since they have found that other preconditions for industrialization existed prior to the Restoration—such as high agricultural productivity, commerce on a nationwide basis, and institutions and practices essential for industrialization (a monetized economy, banklike money-exchangers, and the like). If Japan had all of the other necessary preconditions for industrialization—except for the new technology, which it could borrow—then I find it unlikely that it had a very low standard of living, particularly in light of the evidence that life expectancy in Japan in the mid-nineteenth century was higher than that of the industrialized countries at the onset of their industrialization and was similar to Western life expectancy in the mid-nineteenth century. The problem is that determining the standard of living is difficult in most cases and virtually impossible when there are no reliable data or gross domestic product. Japan before the Meiji Restoration was an example of the latter, a country where many daily goods were not traded. Even had they been, it would have been difficult to compare their prices with prices in other countries because the material cultures were so different. It is essentially impossible to make a comparative study of Japan's living standard using prices and quantities of goods consumed. 2. Joan Thirsk, Economic Policy and Projects: The Development of a Consumer Society in Early Modern England (Oxford: Clarendon Press, 1978).

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

3

More significantly, economists may be using the wrong measure, or an insufficient one, in focusing on their concept of standard of living. For economists, a low standard of living signifies a low saving rate since the poor consume a very large proportion of their income. Savings are important because they enable the investment necessary to industrialize, and since the Japanese industrialized rapidly and without borrowing significant amounts of foreign capital, it would seem logical that they had the necessary standard of living.3 In general the rate at which families save is determined by how high their incomes are, and the standard of living as defined by economists is usually indicative of how well people lived or how much they were saving. But this does not mean that we know how healthy they were. Scholars of the West have argued for decades whether the Industrial Revolution brought with it an immediate rise in the standard of living. That is, did the Industrial Revolution improve the lives of the populace while the process was ongoing, or did the lives of urban factory workers deteriorate? Recent research reveals that the answer may not be simple. Costa and Steckel have found convincing evidence that in the United States in the nineteenth century, income increased prior to the Civil War but health declined, whereas improvements in health outpaced increases in income in the twentieth century.4 Exactly why health declined is open to speculation, but clearly some economists now realize that standard of living cannot be used as the sole measure for how well people live, even taking income distribution into consideration. However, income, or the amount of goods and services one can purchase, is not a sufficient measure for how healthy or productive people are. It does not tell us how well educated or alert the populace is, nor how capable people are of working long hours at various tasks that require both physical and mental stamina. In other words, it does not tell us how suited the population is for the various occupations involved in a modern industrial economy—from installing and running factory equipment to the tasks involved in setting up and maintaining the in3. For the importance of savings with regard to Japan's industrialization, see Henry Rosovsky, Capital Formation in Japan, 1868-1940 (New York: The Free Press of Glencoe, 1961). 4. Dora L. Costa and Richard H. Steckel, "Long-Term Trends in Health, Welfare, and Economic Growth in the United States," National Bureau of Economic Research, Cambridge, Mass., NBER Working Paper Series on Historical Factors in Long Run Growth, Historical Paper 76, November 1995. © Dora L. Costa and Richard H. Steckel.

4

P H Y S I C A L W E L L - B E I N G IN TOKUGAWA JAPAN

frastructure to support them: bureaucracies, transportation and communications systems, and the like. Badly malnourished and illiterate people will be effective at none of these jobs. Economists have become aware that in dealing with people as labor, they must consider quality as well as quantity. The kind of labor needed to run a modern factory, even in the nineteenth century, is very different from what was required to run a farm. The workers not only need a variety of skills but also the ability to work continuously for long hours and be mentally alert at the same time. As societies industrialize, they need not only a supply of labor for modern economic activities, they need to invest in their human capital. Forms of investment in human capital include schooling, medical care, and on-the-job training, 5 all of which improve the quality of labor either physically or in terms of people's skills. Examples do not exist of a population eking out a subsistence living in agriculture suddenly and directly establishing a modern industrial economy. The aim of this study is not to offer still another quantitative "guesstimate" of Japan's standard of living for the mid-nineteenth century, using questionable data and procedures debated even among economic historians, in order to estimate how much capital Japan had to invest in modern industry. Rather, it is to attempt to answer the question of how well the Japanese lived in a broader sense, including aspects other than income, so that we will have better knowledge of the quality of Japan's labor force at the onset and during the early years of its industrialization. Because other studies make clear that Japan had a sufficiently educated labor force by the 1870s,6 and one that became increasingly well educated during the Meiji period (1868-1912) and beyond, I attempt to assess the level of physical well-being, which is based on income but also evaluated on the basis of many other indicators and considerations. 5. Gary S. Becker, Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education, 2d ed. (New York: National Bureau of Economic Research, 1975), p. 9. 6. Economic historians are generally agreed that it is necessary to have a 4 0 percent literacy rate in order to successfully industrialize. Japan had roughly this rate by the Meiji Restoration, and the percentage rose quickly during the Meiji period. See R P. Dore, Education in Tokugawa Japan (Berkeley: University of California Press, 1965), and Susan B. Hanley, "La relación entre educación y crecimiento económico en Japón," in La maldición divina: Ignorancia y atraso económico en perspectiva histórica, ed. Clara Eugenia Núñez and Gabriel Tortella (Madrid: Alianza Universidad, 1993).

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

5

The level of physical well-being encompasses many of the elements that today are discussed under quality of life.7 However, quality of life is an amorphous concept that has taken on broad, philosophical connotations in recent academic discussion, including subjectively measured as well as objectively measured elements.8 In contrast, the general level of physical well-being, though a more awkward term, broadens the scope of inquiry and can still be objectively measured. Thus this concept has a double advantage: it gives not only a better measure of how well people live, but also it can be more readily evaluated than can the standard of living in premodern societies. For Tokugawa Japan, I base the evidence for the level of physical well-being on the material culture, which reveals the patterns and characteristics of consumption, which in turn determines the physical well-being of the population. By studying these aspects of premodern Japan, we will not only better answer the economists' questions, but also learn a great deal more about what life was like in the Tokugawa period. In this chapter I first present my argument why the standard of living alone is insufficient as an indicator of how well people live, insufficient for analyzing the preconditions of industrialization, and certainly insufficient for making cross-cultural comparisons. I present my alternative, the level of physical well-being, and show how it can be defined narrowly enough to make it an important measure. Then I explain my sources—the material culture—and how analyzing consumption enables us to better estimate both the standard of living and the level of physical well-being. I end the first half of the chapter with my hypothesis about the level of physical well-being in Tokugawa Japan and its significance. In the second half of the chapter, I present preliminary evidence for the Tokugawa period to support this argument and summarize the contents of the remaining chapters. 7. I am grateful to Ann Waswo of Oxford University for suggesting this term. Physical has been included here because the concept of well-being encompasses happiness as well as physical conditions, and I am making no attempts to measure subjective feelings toward life. 8. Amartya Sen initially discussed the standard of living as the concept was developed in economics, and later he added an analysis of the quality of life, which includes not only how much money people have, but also life expectancy, health care, education, the kind of labor they do, political and legal privileges, freedoms, family and other relations-an almost endless list that describes what makes people feel happy and fulfilled. See Amartya Sen et al., The Standard of Living (Cambridge: Cambridge University Press, 1987), and Martha Nussbaum and Amartya Sen, The Quality ofLife (Oxford: Clarendon Press, 1993). The list of items is from p. 1 of the latter.

6

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

The Argument and Its Significance Standard of living is a term used by economists to refer to the quantities of goods and services available for consumption; it is expressed in monetary terms and is a measure of per capita income. To calculate it for a nation, the gross domestic product (GDP),9 which is the total value of goods and services for a nation, is divided by the population. Thus per capita income is an average income derived from two aggregate statistics: total income and total population. Crude as it is, this measure is found useful in modern nations because it is statistically possible to obtain it from data normally collected and because it gives information needed for policy making as well as to make international comparisons of economic performance. Economists will readily admit that as an indicator, the standard of living is only as good as the data used to calculate it. It depends on an accurate count of the population. If there is a considerable amount of undercounting, the standard of living will appear higher than it actually is. But of more concern in most modern nations is the calculation of GDP. Even in today's advanced countries with all of their administrative capacities, some portion of income goes unreported, either because that income is not readily measurable or is impossible to measure or because it is underreported or not reported at all. For example, it is impossible to calculate implicit income since no one would think of reporting or even calculating the value of vegetables grown in a home garden or the value of exchanging a plumbing repair with a neighbor for homegrown apples. The value of unpaid housework and other services provided within the family are never calculated into GDP; only when a family has to pay for child care and housecleaning is the value of this work counted as income. And income is often underreported to avoid or evade taxes—people hide tips and cash payments, and firms "cook" their books in myriad ways. In short, every society has an underground economy, the income from which never reaches GDP figures. The size of the underground economy in the industrialized nations of today is far from negligible.

9. Gross domestic product (GDP) is the sum of all goods and services produced within a country, whereas gross national product also accounts for international trade and capital flows. Since the Tokugawa economy was largely a closed one, GDP is the most appropriate concept to use here.

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

7

If these are problems in calculating the GDP for an advanced industrial country, it is clear that it is impossible to derive reliable estimates for a preindustrial country that not only did not calculate its gross domestic product, but could not because the economy was not fully monetized and because the government lacked the political and administrative capabilities of modern governments. The Tokugawa government had difficulties in accurately assessing the rice yield, its staple crop, and did not even attempt to regularly tax commerce, initially for practical and philosophical reasons, and later because it was politically, and most likely administratively, incapable of doing so. Problems in calculating the standard of living are compounded when trying to make comparisons between countries, and even more so when completely different cultures and historical periods are involved. Even if accurate estimates of the standard of living could be obtained for Japan and, say, for England, either for the mid-nineteenth century or at comparable stages in their industrialization, how can a comparison be made? During the Tokugawa period, there was no exchange rate, which could provide a basis for a rough comparison, because Japan carried on no significant international trade; its economy was so closed that by the mid-nineteenth century the ratio of silver to gold was 5 or 6:1, compared to a 12 to 15:1 ratio in the West. Even today in countries with fairly similar consumption patterns, international exchange rates do not reflect accurately the differences in the standard of living among nations. To determine what income is worth in another country with another currency, one has to construct market baskets of goods and services and do elaborate estimations. A further problem with using per capita income as a measure of the standard of living is that it tells us nothing about the distribution of income. The Marxists are correct in criticizing neoclassical economics for often ignoring the importance of income distribution. For neoclassical economists, income distribution is determined by the contribution all economic actors (entrepreneur, provider of capital, and labor) make in producing the total value of goods and services they jointly produce, and thus these economists have little to say about how changing income distribution might benefit society as a whole. If an economy is growing, it is thought that the income generated from this growth will filter down to all income levels and everyone will benefit at least to some extent. Thus, growth is the central concern in neoclassical analysis, in contrast to distribution in Marxist analysis. This is to say that economists, eschewing normative judgments, are

8

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

preoccupied with growth and not the "fairness" of distribution within an economy, and as a result, they are more interested in the amount of savings (capital available) than the patterns and characteristics of consumption. Because the rich tend to save more than the poor, a skewed income distribution may help increase the amount of capital within an economy in contrast to the case in which income is more evenly distributed. Therefore it is not surprising to find many economists interested in the pattern of income distribution only in relation to capital formation. For such economists, income distribution and consumption (and the living standard that determines the level of income and consumption) are important only in determining savings and not for their effect on the level of physical well-being. But the Marxists falter on this as well, because their disaggregation is by class; their analysis falls short when applied to what seem clear and rigid class lines that nonetheless no longer reflect economic status, a situation that occurred during the latter half of the Tokugawa period. Though important, the standard of living is insufficient for measuring how well people live. It tells us only how much income people have on the average, and this includes only income that is known or can be estimated. It does not tell us what their nutritional level is, how sanitary or hygienic their living conditions are, the prevalence of disease, or how well they use their natural resources. The usual assumption is that the higher the standard of living, the more advanced the economy, and the better off people are. However, the correlation is not this simple. It is possible for a population to have had a higher level of physical wellbeing at an earlier historical period when the standard of living was lower, because by the later period the standard of living may have risen but the problems associated with early industrialization were encountered, such as pollution, crowded urban living quarters, and an unbalanced diet. It is even possible for the correlation in the change of standard of living and the level of physical well-being to be negative. A society can experience a rise in income and yet face an increased risk of disease or poorer sanitary conditions and, therefore, more unhealthful living conditions. In Japan in the late-nineteenth century, more people could afford to buy polished white rice, and they substituted this more expensive staple for the cheaper brown rice. Their incomes had risen, and therefore, by definition, so had their standard of living, but these same people now faced a vitamin B deficiency and were subject to beriberi. Even members of the imperial family are reported to have died of beri-

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

9

beri during the Meiji period, which meant that those with the highest standard of living faced malnutrition and disease that did not affect the lower-income Japanese eating mixed and semipolished grains. Similarly, in the United States after World War II many people could afford to eat expensive, well-marbled beef, and the incidence of heart disease rose. A high standard of living does not necessarily translate into a similarly high level of physical well-being. The preceding are not isolated examples. A long list could be drawn up of circumstances in which a rise in income subsequently led to a deterioration in physical well-being. When the water closet, or flush toilet, was invented in England, it seemed to raise the level of sanitation for those families who could afford to have the system installed. But they were consequently subjected to noxious gases until the system could be perfected, and everyone, rich and poor alike, suffered from having sewage flushed into the Thames, which served as London's water supply. Another example is the high fashion of the nineteenth century. Women who could afford the corsets and restrictive clothing may have been the envy of those who could not, but the difficulties they had in merely moving and breathing led to physical problems and many of the famous "women's complaints." In any society, people may be pulled to the cities by the attraction of higher wages, but those who earn more by doing so may be living in crowded, unsanitary, and unhealthful living conditions. Yet it is possible for a small rise in the standard of living to bring about a larger increase in the level of physical well-being, or to reallocate resources so that a higher level of physical well-being is achieved without any increase in the standard of living. Planting sweet potatoes in place of grain in part of a family's upland fields can lead to a more nutritious diet. When fuel is expensive, shifting from an open fireplace to an enclosed stove that requires less fuel will free income for other uses. The diffusion of new carpentry techniques may enable an even higher-quality house to be built of lower-quality, and hence lower-cost, materials. All of these occurred during the Tokugawa period. Economists might counter that these considerations of health and well-being are not what is important in the measure of the standard of living. If one is concerned with the questions of how and why a society industrializes, then what is important is attaining a certain level of income (and one that is preferably growing) that will enable this society to save and create wealth. The well-being of the population has only belatedly become a consideration of economists such as Robert W. Fo-

io

P H Y S I C A L W E L L - B E I N G IN TOKUGAWA JAPAN

gel and his group at the University of Chicago. When economists moved from a concern with quantifiable measures to quality, most did not go far enough, but limited their consideration to levels of skills.10 I argue that in order to industrialize, it is necessary to have not only a sufficient supply of labor, but also a labor force that is both healthy enough and educated enough to staff the new factories, deal with the new technology, serve in the administration, become bankers and communications experts, and handle all the matters relating to the creation of an industrial economy. Economists have too often talked of labor in terms of measurable units and considered Japan to have had a labor surplus into the early-twentieth century. However, if the number of workers available was the only thing that mattered, it would not have made sense for a dual structure of wages—in which higher wages went to the skilled and better-educated employees of large firms in contrast to less-skilled employees of smaller firms—to have developed in Japan during its industrialization. Since we cannot calculate the standard of living accurately and since the standard of living doesn't tell us how healthy the Japanese were or how well they lived, I propose a new measure: level of physical well-being. The level of physical well-being is defined here as the standard of living plus "quality factors" that can be positive or negative. Note that I am not rejecting the concept of standard of living. Rather I am incorporating it into a new measure, one that will give additional information and is easier to estimate. What I call quality factors are elements in an environment or society that are not captured when income is calculated but that must be considered when evaluating the level of physical well-being.11 Though there is some overlap between the term quality factor and the economist's externality, a quality factor is a measure broader than that of standard of living and, unlike externalities, some 10. When economists began to use what is known as the "growth equation" to analyze the contributions made by capital, labor, and technology toward the growth of industrial economies, they found that the contribution of labor was underestimated. This could be corrected only by "adjusting" the labor data to consider human capital—that is, the contribution resulting from education and skill levels. 11. In the 1985 Tanner Lectures, Amartya Sen questioned whether the concept of the standard of living could capture how well off people were when one person might have a slightly higher income than another, and therefore be able to buy and consume more food, but actually be in poorer health because of a higher metabolic rate and a parasitic disease. Sen, The Standard of Living, p. 15. The concept of the level of physical well-being attempts to make this distinction; the second person would be considered to have a higher level of physical well-being though a lower income and therefore lower standard of living.

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

n

quality factors can be evaluated by the market. Similarly, there is an overlap between quality factors and nonmarket transactions (goods and services not traded, thus not included as a part of measured GDP per capita). Because some quality factors can be measured, though not in terms of income, I have not used the word qualitative, which implies something that cannot be quantified. One can measure a negative quality factor such as a polluted water supply, or a positive one such as fast-running rivers with little contamination. Other examples of quality factors are the quality and level of nutrition, incidence of disease, level of general health, number of children per family, the percentage of dependent persons (too young or old to work), the size and quality of housing, the kind of heat available, and the many other aspects of life that affect our physical well-being. These quality factors are affected by and related to income, but in determining the level of physical well-being what matters is the net effect of the positive and negative factors. If the positive quality factors in a society outweigh negative ones, then the level of physical well-being will be higher than the standard of living, and if the reverse is true, then the standard of living will be higher than the level of physical well-being. I believe the former to be true for the Tokugawa period. My argument is as follows: As income rises over time, so does consumption, but not to the extent that income does because as incomes rise, people tend to save a larger percentage of their incomes. However, if resources are used efficiently and housing and sanitation facilities are constructed so that people are less likely to come into contact with disease-bearing bacteria, and if the diet becomes well balanced, then the increase in the level of physical well-being may be greater than the increase in income over time. And if in another society, as incomes rise, people switch to a more unhealthful diet, then the rate of growth in the level of physical well-being may be less than that of the growth in income. This can occur when people are already getting sufficient protein and fat in their diet but continue to substitute meat and dairy products for vegetables, creating an unhealthy nutritional balance. Also, the concept of well-being of a people presumes that the majority are better off, since it refers to the well-being of an entire society, whereas a society may experience a rise in the standard of living because a few magnates or plantation owners become extraordinarily rich. In the second case, the standard of living as a statistic will show an increase, but the people as a whole will not be better off. We are led astray when we think,

12

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

as we often do, in terms of the mean, ignoring the measures of dispersion, that is, distribution of income. The evidence used in this study is the material culture of the Tokugawa period. Material culture is here defined as the physical objects that people use or consume in their everyday lives, most of which are either made or else natural objects put to specific use by people. 12 1 concentrate on what are considered the basics: food, clothing, and shelter, and concomitant aspects such as hygiene and sanitation. The artifacts of daily life reveal use of resources, the level of technology, how people cooked, what kind of houses they lived in, and levels of comfort, sanitation, and health—in short, how people lived. This is not primarily a quantitative study; for the most part the information that exists does not lend itself to quantification. However, this kind of evidence is well suited to making comparisons, either over time or across cultures. The diets in two countries may be very different, but analysis of content reveals nutritional similarities and differences, and when there is sufficient information, it is possible to estimate caloric intake. But when quantitative evidence, such as population statistics, is available, health can be evaluated by comparing life expectancy and the pattern of mortality between two periods or two societies. The material culture yields evidence on each independent variable in the equation determining the level of physical well-being: the standard of living and both positive and negative quality factors. For the economist, the standard of living is equal to income, and income is equal to consumption plus saving, so by studying material goods we are looking at the consumption and saving side of the equation. The material culture, being made up of the goods people consume, represents their consumption in terms of material objects, but it can also give a fairly accurate indication of the services they receive as well because, for example, large houses and complicated kitchen equipment may require servants. And because a house is the most expensive item owned by a family, it is equivalent to much of a family's wealth and is usually in 12. Although material culture can be defined as "the study through artifacts of the beliefs . . . of a particular community," as by Jules David Prown in "Mind in Matter: An Introduction to Material Culture Theory and Method," in Material Life in America, 1600-1860, ed. Robert Blair St. George (Boston: Northeastern University Press, 1988), p. 18, the term is so widely and generally used that I prefer it to Braudel's "material life." The problem I have with the term that Braudel has made famous is that it is so allencompassing. Nor do I have the vision or model of structure that Braudel has. See Fernand Braudel, The Structures of Everyday Life, vol. 1 of Civilization and Capitalism, Fifteenth to Eighteenth Century (New York: Harper and Row, 1981).

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

13

keeping with its total wealth. Therefore, housing that is owned represents past saving. By looking at the consumption and saving side of the equation instead of income, we do not have to estimate unreported income, implicit income, and services supplied outside the market or resolve the numerous problems that plague those trying to calculate the standard of living through income. Whereas the standard of living can be calculated using either income or consumption and saving, the level of physical well-being can be assessed only through the kinds of evidence used to analyze consumption and saving, that is, through the material culture. The amount of income and savings used to pay for a house reveals only information about cost, whereas data on housing itself tell us how much space people had, the quality of the building, what resources were used, and what facilities were for cooking, heating, and sanitation, as well as how expensive it was. I am not arguing that this kind of evidence should be used in place of good modern quantitative data but that it is better than making "guesstimates" based on numerous and possibly unwarranted assumptions. In addition, it gives us information pertinent to economists who should be considering but have ignored this kind of data in their quest for quantification. Based on an analysis of the material culture, I argue that the standard of living and the level of physical well-being continued to rise throughout the Tokugawa period, during the seventeenth through nineteenth centuries, though not at constant rates and not at necessarily the same rates. This rise was the result of the growth of the economy; the diffusion of technology, significant new foodstuffs, and various improvements in clothing and housing; stabilization of population growth in the early-eighteenth century; and the installation and regulation of water supply and waste removal systems by government in urban centers. Because income rose more quickly for the commoners than the samurai, income came to be more evenly distributed over time. By the eve of the Meiji Restoration, the Japanese standard of living was at a similar level with that of England on the eve of its industrialization a century earlier. Although the average income of the Japanese in the 1860s was not quite as high as that of the English in the eighteenth century, I argue that the level of physical well-being was at least as high as that in England in the nineteenth century, at a time when Japan had not yet begun to industrialize and England was already an industrial nation, and certainly as high as when England began to industrialize.

14

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

Because of Japan's isolation during the Tokugawa period, it developed a material culture and lifestyle that were in many ways distinctive from those of any other country and most certainly from that of the Western industrializing countries. I believe that the preference of the Japanese for their traditional lifestyle throughout industrialization, enhanced in part by government regulations to limit imports of consumer goods, enabled both political and economic change to proceed more smoothly than would otherwise have been the case. The Japanese had a stability in their daily lives that helped them through the turbulent Meiji transition years when government, social systems, the economy, and much of public life seemed to undergo continuous change. However, the material culture of the early Meiji period not only provided for an adequate level of physical well-being, but the type of housing and some of the material infrastructure, such as disposal of sewage, were still in use a century later. The level of physical well-being was high enough that Japan had a relatively healthy labor force and one sufficiently literate upon which to build an industrial nation. The preference for the established lifestyle meant that the country could use its resources and scarce foreign currency for the purpose of building a modern economy, and at the same time this preference kept native industry vibrant. And the social and economic developments that had tended toward the leveling of incomes, at least between class lines if not among them, led to a more homogeneous material culture and lifestyle than one would ordinarily expect in a country with such theoretically rigid class lines. This in turn led to a social mobility that aided in the creation of a modern state and economy.

Preliminary Evidence The preliminary evidence offered here first covers the standard of living, both because this is an important aspect of the level of physical well-being and because it is not treated separately elsewhere in the book. Scholars have long agreed that the seventeenth century in Japan was one of unprecedented economic growth prior to industrialization. When peace prevailed and economic growth continued following the establishment of a stable regime over Japan by the Tokugawa house in the early seventeenth century, the innovations in technology, agriculture, and systems of human organization—which developed

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

15

during the prior century of civil strife—were diffused throughout the country. New engineering techniques were used to complete major irrigation, flood control, and land reclamation projects, adding large amounts of arable land. New "contracts" worked out between the rulers and cultivators resulted in increasing control by the latter over their land and crops, and this, combined with the gradual freeing of persons in indentured servitude and serflike statuses, gave cultivators increased incentive to raise productivity and output. In addition, the absence of war, major famine, or destabilizing political crisis during the seventeenth century led to a concentration on economic growth almost certainly unparalleled in previous centuries.13 The growth in the agricultural sector made it possible to feed an increasing rural population and at the same time support a rapidly growing urban population. Government measures to control the samurai encouraged urban growth, so that by 1700 the urban population is estimated to have been at least 10 percent of the total population.14 This provided the basis for the increasing commercialization of the Japanese economy. Urban affluence led to a cultural flowering centered in Kyoto and Osaka at the end of the seventeenth century. Although Marxist historians have argued that both the economy and population stagnated during the eighteenth and nineteenth centuries, most scholars studying the Tokugawa period now find convincing evidence of continued economic growth, though at a slower rate than in the seventeenth century. If the economy was growing, and at a rate faster than the population, then by definition the standard of living rose, meaning that on the average people had more income. This is particularly true for the eighteenth and nineteenth centuries when the national population surveys 13. See Kozo Yamamura, "Returns on Unification: Economic Growth in Japan, 15501650," in Japan before Tokugawa, ed. John Whitney Hall, Nagahara Keiji, and Kozo Yamamura (Princeton: Princeton University Press, 1981), pp. 327-72; Nagahara Keiji and Kozo Yamamura, "Shaping the Process of Unification: Technological Progress in Sixteenth- and Seventeenth-Century Japan, The Journal ofJapanese Studies i+, no. 1 (winter 1988): 77109; Thomas C. Smith, The Agrarian Origins of Modern Japan (Stanford: Stanford University Press, 1959); and Susan B. Hanley and Kozo Yamamura, Economic and Demographic Change in Preindustrial Japan, 1600-1868 (Princeton: Princeton University Press, 1977). 14. These measures included the sankin kotai system of alternate attendance of the daimyo in Edo and limiting each daimyo to only one castle for his domain. Gilbert Rozman, Urban Networks in Ch'ing China and Tokugawa Japan (Princeton: Princeton University Press, 1973), p. 6, estimates Japan to have been 10 percent urban, whereas Sekiyama Naotaro, Kinsei Nihon no jinko kozo (Tokyo: Yoshikawa Kobunkan, 1958), p. 239, considers 12 percent of the population to have been urban.

i6

PHYSICAL W E L L - B E I N G IN TOKUGAWA JAPAN

undertaken at the direction of the Bakufu show a commoner population fluctuating around twenty-six million, from 1721 when the survey was first undertaken, to the last in 1846. 15 If during the second half of the Tokugawa period the population was not rising, then even a very small rate of economic growth would still have led to a rise in the standard of living. And the information we have on both income and consumption attests to its continued rise. Though the data that exist are local or regional and limited to certain periods, taken together they provide strong evidence for a rising real income.16 Hayami Akira, who examined the wages of certain employees, the hokonin, in Kanto villages, concluded that from the early-seventeenth century to the early-nineteenth wages rose more than the prices of goods. 17 Sano Yóko found that the real wages of carpenters in Edo rose between 1830 and 1894, despite the various fluctuations to be expected from such events as the multiyear crop failures in the 1830s and the rebuilding of Edo in the late 1850s following the Ansei earthquake.18 Saito Osamu found that the rise in wage levels with the growth of commerce and manufacture led to a decline in the effective rents on land.19 This meant that real income rose for wage workers and cultivators alike, and evidence shows that this was not just in the most advanced areas such as the Kinai, but also in the most backward, such as Tohoku in the north. People living in the Tokugawa period were well aware of rising incomes; a headman in what is now Akita Prefecture wrote in 1825: "Every family in the village without exception works at sericulture in the intervals of farming, thereby earning more income." 20 The evidence that exists on the consumption side of the equation provides equally strong support for a rising standard of living throughout the Tokugawa period. Some of the most illustrative evidence on

15. See Sekiyama, Kinsci Nthon, pp. 137-59, for the Bakufu compiled figures. 16. I cannot repeat here all of the evidence. See Hanley and Yamamura, Economic and Demographic Change, especially chapters 4 through 7. 17- Hayami Akira, Nihon keizai-shi e no shikaku (Tokyo: Tóyó Keizai Shimposha, 1968), pp. 165-66. 18. Yohko Sano, The Changes in Real Wages of Construction Workers in Tokyo, 1S301894, English Series No. 4 , Institute of Management and Labour Studies, Keio University, Tokyo, 1963. 19. Saito Osamu, "Tokugawa koki kara Taisho zenki ni itaru nógyó chingin no chókiteki süsei," Shakai keizai shigaku 39, no. 2 (June 1973): 170-89. 20. Thomas C. Smith, "Farm Family By-Employments in Preindustrial Japan," The Journal of Economic History 29, no. 4 (December 1969): 709.

P H Y S I C A L W E L L - B E I N G IN TOKUGAWA JAPAN

17

the changes in what was bought and used in daily life comes from the domain of Okayama, one of the more economically advanced areas in western Japan. During the sixteenth century and well into the seventeenth, the rural population bought what it needed from peddlers who hawked their wares from village to village. In 1655 the domain government prohibited these sales on the basis that cultivators were spending too much money on articles they did not need. Total prohibition proved unsatisfactory, however, and so peddling was licensed and the number of goods limited—in 1666 to only eleven items: fishing nets, dried fish, salt, dried seaweed, tea, rapeseed oil, kindling, wooden water dippers, basket tops, oars, and farm tools. As demand—and undoubtedly open violation of the decree—grew, additional goods were added to the permitted list. By 1705, there were thirty-one, including pottery, cotton, pans, rice pots, straw mats, paper, fans, and rulers.21 By the mid-eighteenth century, demand in Okayama had increased to such an extent that major changes took place in rural marketing. Whereas the early peddlers had been based in the castle town, the surrounding towns had now grown to the point that they not only supplied the needs of the rural villages but even began to get a foothold in the markets of the castle town. By the late-eighteenth century, villages' demand for goods was sufficient to support shops of their own. In 1813, one shop in the village of Oi sold ink, paper, writing brushes, pots, needles, pipes, tobacco and pouches, teapots, various containers and dishes, vinegar, soy sauce, bean paste, salt, noodles, kelp, sake, cakes, tea and teacups, senbei (rice crackers), grain, oil, candles, hair oil, hair strings and hairpins, cotton cloth, towels, tabi (socks), footgear including zori, geta (wooden clogs), and waraji (straw sandals), funeral requisites, and "other everyday necessities." Other shops in the village sold various kinds of food and farm necessities, such as tools and fertilizers.22 Evidence of the rising standard of living for a large majority of people is also evidence of a more equal distribution of income. Samurai were increasingly unable to offer competitive wages to their employees, merchants complained that high wages alone were insufficient to retain staff, and the larger landowners in the countryside complained about being unable to obtain daily farm laborers because other options were 21. Ando Seiichi, Kinsei zaikata shójjyó no kenkyü (Tokyo: Yoshikawa Kobunkan, 1958), pp. 125-28. 22. Ibid., p. 95-

18

P H Y S I C A L W E L L - B E I N G IN TOKUGAWA JAPAN

more attractive for workers. Kozo Yamamura has shown that the samurai serving the Bakufu were able to maintain a steady real income during the Tokugawa period, largely through the aid of various Bakufu measures, but the samurai serving the daimyo by and large became much worse off over time as domain finances weakened and the daimyo "borrowed" from their retainers' stipends. However, neither group enjoyed a rise in real income, and because many samurai households were not able to afford to pay the rising wages servants could command, at the very least they faced a lowered quality of life. By the end of the Tokugawa period many samurai and commoners alike were working in the same lower-echelon staff positions within the daimyo's castle, and a large proportion of samurai of low status were engaged in cottage industry and other kinds of employment.23 While the income of the samurai—who constituted no more than 7 to 10 percent of the population—either held steady or deteriorated, most of the rest of the Japanese enjoyed a rising standard of living.24 That most Japanese shared in the growing total output is evidenced both by the widespread sale of daily goods over time and theoretically by what is known as the wealth effect. It is obvious that a shop could not have been established in the rural villages had there not been a demand by most of the residents for the goods sold on a scale to make the shop profitable. A major study of the food industry in modern Asia reveals that "trigger points" in income—that is, a leap in the average income per person—occur before people switch to the consumption of a new type of good. This is obviously more pertinent for expensive consumer durables than for an inexpensive product such as candy, but it can be traced for basic food products.25 If the Japanese economy continued to grow throughout the Tokugawa period, giving rise to an increased standard of living for a population that was stable or only very slowly growing from the eighteenth century on, why are the more reliable estimates of Japan's per capita income on the eve of industrialization so low? Minami Ryoshin's estimates for Japan's per capita GNP at the onset of industrialization (1886) 23. Kozo Yamamura, A Study of Samurai Income and Entrepreneurship (Cambridge, Mass.: Harvard University Press, 1974), pp. 131-33. 24. The proportion of samurai in the population differed by domain, ranging perhaps as high as one-quarter in rare cases, but the usual estimate lies somewhere between 7 and 10 percent. Shinmi Kichiji, Kakyit shizoku no kenkyii (Tokyo: Maruzen, 1965), pp. 63-64. 25. The Economist, December 4 , 1993, p. 15.

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

19

are only 60 percent of England's at a similar stage of development (1765-1785).26 According to Angus Maddison's estimates, Japan's per capita GNP in 1870 was just one quarter that of the United Kingdom's in the same year and 36 percent that of the U.S. 27 Nishikawa Shunsaku is extremely critical of the efforts to estimate per capita income for Japan for the nineteenth century even using the concept of purchasing power parity (PPP), a method adopted by economists making international comparisons of living standards.28 Although using the PPP method is superior to trying to estimate per capita income using the exchange rate alone, extrapolating backward using PPP is seriously flawed because the base used is twentieth-century prices and mix of goods. Even applied to present-day economies, the real per capita income deviates from the estimate to the extent that there are nontraded goods in the international market. And this method does not attempt to deal with cultural preferences or changing tastes within a culture. I remain extremely skeptical about all of the estimates made. First, as already noted, Japan had an economy that was essentially closed to the world market until the late-nineteenth century, certainly in terms of goods in daily use. Second, the goods used on a daily basis were very different from those in the West and also quite different from those in use by the 1960s and 1970s. Third, many goods and services were not traded and therefore cannot be factored into an estimate. This being the case, such calculations are an interesting exercise, but what they can tell us is far from adequate in indicating how well people lived. This is especially the case because the estimates vary so widely depending on the method used that one can find credible whichever one wants to believe.29 An example of a tenant farmer's budget for the 1840s shows how

26. Estimated using Kuznets's method. Minami Ryoshin, The Economic Development of Japan: A Quantitative Study (Houndmills: Macmillan Press, 1986), p. 13. 27. Angus Maddison, Economic Growth in Japan and the USSR (London: George Allen and Unwin, 1969), p. xvi. 28. Nishikawa Shunsaku, Nihon keizai no seicho-shi (Tokyo: Toyo Keizai Shimposha, 1985), pp- 217-21. 29. If forced to choose among the estimates of Japan's per capita GNP at the onset of industrialization compared to that of other nations at the same stage, I would agree with Yasukichi Yasuba that "all in all, the Japanese standard of living may not have been much below the English standard of living before industrialization." "Standard of Living in Japan before Industrialization: From What Level Did Japan Begin? A Comment," The Journal of Economic History 46, no. 1 (March 1986): 224.

20

P H Y S I C A L W E L L - B E I N G IN TOKUGAWA JAPAN

complex it is to analyze even one person's income and expenses.30 Although a tenant, this farmer was running a large operation for Tokugawa times, and this is undoubtedly the reason for his keeping detailed records. Of the rice this farmer grew, part was consumed, part was sold, and part was used as a means of exchange. Cash income was denominated in gold, silver, and copper, and the relative values fluctuated from year to year and season to season.31 His business expenditures in his budget were mixed with those for the household, since both were considered a unit. This farmer's main crops were rice, barley, and daikon (large, white radishes).32 The yield from his one hectare of paddy was approximately 102 bushels of rice, of which a quarter was paid to his landlord as rent, and another quarter paid the land tax. From a halfhectare of dry fields he harvested twenty-five thousand daikon, which he sold, making considerable profit even after paying for fertilizers and transportation to market. If the farmer's cash income is converted into gold, it amounts to about twelve ryd, which is roughly half the income of a carpenter in Kyoto for whom we also have income and budget figures. Of the cash income, the farmer spent two ryd on salt, tea, and oil; one for firewood; two for farm implements and household furnishings; one and a fifth for clothing; two for social obligations such as festivals and funerals; one and a fifth for daily labor; and one for unexpected social obligations. Very little was spent on food because the family grew its own vegetables and these were not entered into the farm books. There were no cash expenditures for rent. From this information we know a lot about how this farmer lived, but we do not know his real income. We know that although he grew his own food, he was heavily involved in the market economy. Although a tenant farmer, he had sufficient cash income to spend one quarter of it for social obligations, and he hired labor. It is highly likely that he lived on a vegetarian diet, obtaining protein from soybeans and grains, since he is not reported as spending income on fish and he had to use cash to obtain sufficient firewood, which means he did not have suf30. Kyuso zappitsu, cited in Ono Takeo, Edo bukka jiten (Tokyo: Tembosha, 1980), pp. 132-36. 31. See E. S. Crawcour and Kozo Yamamura, "The Tokugawa Monetary System: 1787-1868," Economic Development and Cultural Change 18, no. 4 , pt. 1 (July 1970). 32. For the information on this farmer's budget presented in this paragraph and the following one, see Ono, Edo bukka jiten, pp. 132-36.

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

21

ficient access to woodlands to obtain protein in any other form. By the 1840s he was buying clothing and spending cash on farm equipment and household furnishings as well. This was no subsistence farmer! Income and expenditure summaries do not really tell us much about how well people lived. We have to turn to the material culture, where there is an abundance of information, though most of it is not quantitative. Evidence on how people lived can be found in house plans and surveys of housing, the physical artifacts of daily life, diaries and journals, and various government documents. Travelers' letters and journals can be very useful, as they often record details of daily life that are not noted by the inhabitants of a place because they seem too ordinary and unnoteworthy. All of these kinds of evidence must be used with caution, just as with any kind of historical evidence, as they may be biased, inaccurate, or idiosyncratic. But they offer a wealth of information with which to supplement the information we now have on income and the standard of living, and they will lead to more accurate international comparisons. Various kinds of evidence on housing indicate that it clearly improved during the course of the Tokugawa period, as is discussed in detail in chapter 2. Houses were better built, and the scattered evidence we have indicates that for the most part they became larger. One of the key indicators that people built better housing is that more Japanese were able to install flooring, instead of leaving floors of pounded earth, and by the mid-nineteenth century, many villagers had tatami (rushmat flooring) and nearly all had at least wooden floors in their houses. Irregularly shaped, rather primitively built one- or two-room houses were no longer the norm, although they did not disappear. The evidence reveals that not only could people afford to spend more on daily consumption goods that became increasingly available over time, but whether village heads or tenant farmers, they also saved and built better houses. From the style of housing, clothing, food, and forms of luxury, we know that the Tokugawa material culture was resource-efficient. The new styles of housing that originated with the samurai and were copied by commoners made housing more healthful and comfortable. At the same time they used fewer nonrenewable resources, such as huge logs, and more that were readily renewable, such as rush-matted flooring and boards and posts sawn from secondary growth. Clothing took a minimum of material, certainly compared to Western dress, used an entire

22

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

length of cloth with no waste, and could readily be adapted for anyone without resewing a garment. Japanese also used far more wild plants and life from the sea as foodstuffs than did people in Europe. And luxury was found not in abundance, as in the West, but in the use of a few flowers or display of a single art object. All of this is elaborated in chapter 3. The result was that the population achieved a higher level of physical well-being using fewer resources and with the need for less income than was required in the West for the same level of physical well-being. The Japanese had a relatively high level of physical well-being because their lifestyle was comparatively healthful. Though few ate meat, and fish was not a daily part of the diet for most, they obtained protein from plant sources, notably soybeans and grains. Most people could not afford to eat much polished rice, and so the rice and other grains they did eat had more of the nutrients left in them than a staple of white rice would have provided. Grain was supplemented with sweet potatoes and a variety of vegetables and would seem to be a diet preferable to that of workers,and their families in English mill towns, one that consisted largely of bread and margarine. The few existing figures that enable estimation of calories indicate that Japanese in the mid-nineteenth century had sufficient nutrition for their body stature, although height is partially determined by nutrition. But the Japanese lifestyle was also healthful because it was hygienic, certainly compared to either Europe or the U.S. in the mid-nineteenth century. Bathing was a regular part of life by this time, people customarily drank their water boiled in the form of tea, and they carefully collected their bodily wastes to be used as fertilizer. In the cities, the water supply systems and waste removal were regulated by government, and people had little incentive to dump night soil where it would cause contamination, because they could sell it to farmers for cash. The healthful and hygienic lifestyle of the Japanese is discussed in chapters 4 and 5. How healthy the Japanese were in comparison to the people in already industrialized England or industrializing America might remain a matter open to argument were it not for the fact that demographic records exist that enable us to compare mortality. On the basis of the demographic records, it can be argued that the Japanese prior to industrialization lived as long as did the English a century after they began to industrialize. These same records indicate that the rise in physical wellbeing during the Tokugawa period was not coincidental; people were clearly acting to form their families in a way that would maintain or enhance their standard of living. The evidence is provided in chapter 6.

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

23

The major reason for wanting to ascertain the standard of living and the level of physical well-being of the Japanese prior to industrialization is to compare it with life in Japan during and after industrialization and with life in other countries that industrialized in the nineteenth century. The final two chapters, 7 and 8, make these comparisons. Chapter 7 argues that despite the fanfare over importations from abroad, the lifestyle and material culture remained similar from the late Tokugawa through the Meiji period. The major changes during these years were in the continued diffusion of what had been aspects of the samurai culture during the Tokugawa period, the most important example of which is housing. This is important because it means that the Tokugawa lifestyle provided a good basis for continued use during industrialization and was capable of absorbing change. Chapter 8 compares the level of physical well-being during the Tokugawa and Meiji periods with that in the West during the same centuries and at the onset of industrialization for each. The conclusion reached in both comparisons is that the Japanese level of physical well-being was higher than we have been led to think it was, whereas the level of physical well-being for the majority in the West was lower than would be expected by a comparison with Japan. When one compares the material culture, level of physical well-being, and use of resources of Japan with other industrialized nations at the onset of their industrialization, there is no surprise that the Japanese were able to industrialize as readily as they did. Nor, given the historical path that Japan followed, should one be surprised at the lifestyle and material culture in Japan in the post-World War II years, when the Japanese adopted many Western aspects of life but also retained traditional ways. And many of the problems the Japanese face today in terms of lifestyle and material culture, such as the lack of modern sewer systems and inadequate housing, are the results of the Tokugawa legacy, sometimes, ironically, because the premodern systems were so good that modernizing was not a necessity. These too are traced in chapter 8. The physical aspects of life form the background for everything we do. Our environment, both natural and artificial, is so much a part of everyday life that few people think about it in terms of how it molds our lifestyles, affects our social relations, determines how we allocate our time, forms our view of the world, and influences our behavior. From everyday lives come many of the determinants of history, and when these are ignored by historians, whatever they write will be distorted, or, at best, one-sided. For how people cope with their physi-

24

PHYSICAL WELL-BEING IN TOKUGAWA JAPAN

cal environment has determined the outcome of battles, what kind of economy is developed, and whether a people is decimated by famine and epidemics or prospers and undergoes economic or population growth. Learning more about the physical aspects of life in the Tokugawa and Meiji periods will enable us to better understand Japanese history as a whole. Few historians even note the crop failures of the last years of the Tokugawa Bakufu or the earthquake in 1855 that destroyed at least fourteen thousand buildings in Edo. No one I know of has analyzed how traditional lifestyles contributed to Japan's national welfare in the Meiji period or how it relates to Japan's social and economic development in the twentieth century. This study does not claim to find conclusive answers to the larger questions of Japanese history, but I hope it will provide useful and significant clues for discovering answers to the questions raised in this chapter, while telling the interesting story of how Japanese once lived and how their traditional material culture and lifestyles developed.

CHAPTER

TWO

Housing and Furnishings

The oldest house in Japan that can be dated exactly was built in 1607, and there are only a very few others in existence from the seventeenth century.1 Even in districts where we would expect to find older houses, there remain no more than one or two built before 1700. In a country that had a population of from fifteen to eighteen million in 1600, why are there so few houses that are three and four centuries old, and none any older than that? Medieval houses can be found all over Europe, and though some of these are made of stone, huge multistoried apartment buildings with timber frames are still occupied in many towns and cities in Germany and other countries. The traditional answer has been that because of Japan's rich vegetation, houses were built of renewable resources and tensile materials, such as wood, bamboo, and thatch, and that in a country subject to earthquakes, fires, and the ravages of a damp climate, houses were not built to last. It has even been suggested that the Japanese think of their houses as they do furniture: consumer durables rather than permanent assets.2 However, many wooden temples built centuries earlier, such as the Horyu-ji whose buildings date from the early eighth century and are among the oldest in the world, remain in good condition. 1. Okawa Naomi, "Minka," in Kinsei II/Kindai, vol. 5 of Nthon no kenchiku, ed. Ito Nobuo et al. (Tokyo: Daiichi Hoki, 1976), p. 80. 2. Fujimori Terunobu, "Traditional Houses and the Japanese View of Life," Japan Echo 17, no. 4 (winter 1990): 72. "Zaisan de wa naku shohizai de aru" are Fujimori's words in the original Japanese version, "Dento kaoku saiken," Kokusai koryu, no. 53 (1990): 79. 25

26

HOUSING AND FURNISHINGS

The real reason lies not in the nature of the Japanese climate or building materials but rather in the transformation of Japan's economy, society, and technology that occurred during the sixteenth and seventeenth centuries. Without national statistics, it can be difficult to document changes in national and personal wealth, economic growth, and the standard of living, but one very good measure is housing. Housing is the most expensive item most people own or use; it can be modified only at great expense of time, labor, and capital; and more reliable information exists on housing than the other elements of the material culture of the Tokugawa period. Thus housing provides the best evidence for the increase in wealth during the Tokugawa period and, as a corollary, the rise in the level of physical well-being. The seventeenth century is the first in which houses were built to last from one generation to the next and of a quality still appreciated and usable in the late-twentieth century. Changes in building techniques combined with a rise in wealth led to this construction of permanent houses for people who were not in the governing elite. As housing improved, people began to acquire more furnishings and other goods, the most significant of which was bedding. Much of the evidence for this acquisition of belongings comes from information on how people began to store their possessions. This chapter examines housing and furnishings as indicators of a rise in personal wealth and the level of physical well-being for Japanese of various classes from the seventeenth through the nineteenth centuries. During the first century of the Tokugawa period, major changes were seen in nearly all facets of daily life. Peace, combined with new agricultural techniques and the growth of commerce, resulted in economic growth and a rise in the population. The relative value of land rose while that of labor fell.3 In an effort to make the swelling number of cultivators bear some of the risks of farming, indentured servants were made into tenant cultivators and given small plots of land to farm. These people were now free to marry and have families. Not only was there a proliferation of small houses, especially in the warmer western regions of Japan, but the population increased at possibly an unprecedented rate throughout the rest of the century.4 Trade flourished, cities 3. Kozo Yamamura, "Pre-Industrial Landholding Patterns in Japan and England," in Japan: A Comparative View, ed. Albert M. Craig (Princeton: Princeton University Press, 1979), p. 285. 4. Susan B. Hanley and Kozo Yamamura, Economic and Demographic Change in Preindustrial Japan, 1600-1868 (Princeton: Princeton University Press, 1977), pp. 43-45.

HOUSING AND FURNISHINGS

27

were built, and a whole new urban and artisan-merchant class culture grew from the new mass of commoners clustered in the cities with wealth they had never before experienced.5 Most of these new ways of life had their origins in the fifteenth and sixteenth centuries, partly in the new material culture of the elites in the Muromachi Bakufu and partly in the technologies, inventions, and materiel developed for the waging of war and growth of military, political, and economic power by the Sengoku daimyo and the unifiers. Ironically, warfare is often the stimulus for developments that benefit humanity—witness the boon from the invention of the tin can for food preservation to feed Napoleon's armies—and this was no less true for Japan in the sixteenth century than for other societies. But the material culture of the Muromachi elites and the benefits of Sengoku developments arising from the needs of the warring daimyo benefited only the few until the peace and the growing economy of the Tokugawa period enabled these to be diffused among the common people, the vast majority of Japanese. Though economic historians continue to debate the level of development of the economy, the rate of economic growth, and the amount of economic development in the Tokugawa period, 6 researchers into topics pertaining to Japan's material culture find evidence demonstrating that life in Japan changed because of an increase in wealth from the seventeenth century on.7 Because Japan was not at the standard of living Europeans were in the mid-nineteenth .century and, more importantly, because Japan had not started to industrialize, we tend to undervalue the level of living that Japan reached during the Tokugawa period. Commoners' houses, called minka, provide evidence for an important historical trend in the rise in the level of wealth and standard of 5. For works providing evidence for this in English based on Japanese sources, see Thomas C. Smith, The Agrarian Origins of Modern Japan (Stanford: Stanford University Press, 1959); articles by E. S. Crawcour, John W. Hall, and Thomas C. Smith in Studies in the Institutional History of Early Modern Japan, ed. John W. Hall and Marius B. Jansen (Princeton: Princeton University Press, 1968); and Kozo Yamamura, "Returns on Unification: Economic Growth in Japan, 1550-1650," in Japan before Tokugawa, ed. John Whitney Hall, Nagahara Keiji, and Kozo Yamamura (Princeton: Princeton University Press, 1981). For the most comprehensive source in Japanese, see the eight-volume series Nihon keizai-shi, edited by Umemura Mataji et al. (Tokyo: Iwanami Shoten, 1988-1990). 6. For the basic argument and works in English see chapter 1 of Hanley and Yamamura, Economic and Demographic Change. 7. Miyazawa Satoshi, "Kinsei minka no chiiki-teki tokushoku," in Kenchiku, ed. Amakasu Takeshi et al., vol. 7 of Koza Nihongijutsu no shakaishi (Tokyo: Nihon Hyoronsha, 1983), p. 15+; Okawa, "Minka," p. 81.

28

HOUSING A N D FURNISHINGS

living for the general population. Minka were built in styles and technologies developed in the fifteenth and sixteenth centuries, which were diffused throughout the countryside as families increased their wealth and could afford the building materials, the tools, and the specialized carpenters to construct larger, better-quality dwellings. The oldest minka may well have been atypical at the time they were built, but they were forerunners of the main style of housing in the Tokugawa period and of the construction techniques used in later centuries, and thus they were not only built to last but people treated them so that they did. What may have been the most important new method in house construction was the use of foundation stones. Earlier Japanese houses relied on posts for support, not walls, and until the seventeenth century, posts were set directly on the ground, where they would rot in the damp climate. The innovation was to rest them on foundation stones, preventing direct contact with the earth but at the same time keeping the framework flexible since the posts could move during a moderate earthquake without breaking. Though this seems like a logical and very simple step, it required suitably shaped stones, which became widely available only after the technology used to build the great castles of the sixteenth century was developed. Other innovations in house construction during the seventeenth century were the use of thicker and higher-quality materials, braces placed between posts, and walls covered with mud plaster or boards. Increasingly doors were made of w o o d rather than split and woven bamboo, and they were built to slide laterally in frames rather than to lift upward. Wooden floors gradually became more common than earthen. 8 In short, houses became sturdier and were not only of a higher quality but also more comfortable because people sat on floors and not on the earth. These innovations increased the level of physical wellbeing, but they also represented a rise in the standard of living, since the new or improved materials and techniques either cost more or required new tools or both. Houses built from the seventeenth century on were more symmet-

8. Numerous descriptions of the new methods of construction exist in Japanese, including volumes 4 and 5 of Ito, Nihon no kenchiku; Hirai Kiyoshi, Nihon jutaku no rekishi (Tokyo: N H K Books 209, 1986); Shiragi Kosaburo, Sumai no rekishi (Tokyo: Sogensha, 1978); Kuwahara Minoru, Jukyo no rekishi (Tokyo: Gendai Kogakusha, 1979). In English see Teiji Itoh, Traditional Domestic Architecture of Japan (New York: Weatherhill/Heibonsha, 1972) and Chuji Kawashima, Minka: Traditional Houses of Rural Japan (Tokyo: Kodansha International, 1986).

HOUSING A N D FURNISHINGS

29

rical in shape than their predecessors. A few houses built in the seventeenth or early-eighteenth century were built on sites of former houses, a fact that became apparent when the houses were renovated or moved from their sites. The earlier house was constructed with the posts set directly on the ground and was usually smaller than the later building. But what is surprising is the uneven shape of the earlier dwelling, judged from the location of the posts. Prior to the development of the Tokugawa minka, houses with posts set at uneven intervals with a resulting irregular layout seem to have been common. The later houses, however, all have right-angle corners, equal distances between the posts, and the various features we associate with the minka and other traditional kinds of Japanese housing.9 The seventeenth-century minka are clearly the products of an advanced technology that swept through the country in that century but was developed during the Sengoku period. Historians have tended to ignore until recently the technological progress that went along with the economic development that made the politico-economic system of the sixteenth century possible.10 In order to build their castles, mine the ore needed for weapons, and dam rivers to create irrigation systems to supply their fiefs, the Sengoku daimyo supported and actively encouraged innovation in technology. The advances in mining and in stonecutting that enabled the construction of stone castles in the sixteenth century made it possible for the local elites, and not just the rulers, to use foundation stones for their houses in the seventeenth century. The developments in transportation, sleds, flat-bottomed boats for heavy cargo, and techniques in carrying rocks made it possible for foundation stones to be used widely. Among the most significant developments in technology in the Sengoku period that applied to the construction of wooden buildings were a new construction method called kiwari, the importation of large saws (oga or ojjabiki) from China, and the innovation of a plane made of an iron blade mounted on a large wooden base (daikanna). H Although the kiwari method was extremely complex, put simply, the diameter of 9. See the examples from Kyoto and Kanagawa shown in Miyazawa, "Kinsei minka no chiiki-teki tokushoku," p. 167. 10. The eight-volume series Koza Nihon gijutsu no shakai-shi, Nagahara Keiji and Yamaguchi Keiji, general editors (Tokyo: Nihon Hyoronsha, 1983-1985), marked a new appreciation for the history of Japanese technology. n. The technological developments of the sixteenth and seventeenth centuries are summarized in Nagahara Keiji and Kozo Yamamura, "Shaping the Process of Unification: Technological Progress in Sixteenth- and Seventeenth-Century Japan," The Journal of Japanese Studies 14, no. 1 (winter 1988): 77-109.

3o

HOUSING AND

FURNISHINGS

the main post was the basic value for the calculation of the other major portions of the structure. This method of construction required a considerable amount of calculation, familiarity with the lumber to be used, and the technology to cut and finish lumber. A guide to these techniques, developed in the fifteenth and sixteenth centuries, was first published in 1608,12 which meant both that the method had evolved to the point it could be set down and that there was a demand for a guide so that the method could be more easily disseminated. These methods of construction enabled the development of a new style of housing for the samurai elites and, for commoners, the development of houses requiring exact measurements and straight boards and posts. There is evidence that these techniques were widely used in some locales in the construction of housing in the sixteenth century. The castle town of Ichijodani in Echizen (Fukui Prefecture) was burned down by Nobunaga in 1573. The area was subsequently turned into fields, but when excavated after World War II, it was discovered that most houses had stone foundations under the main posts, which were square pillars, and the spacing of the posts was regular. No tile was used, but there were numerous metal fittings, such as for sliding doors, nails of various sizes, locks, latches, and door stops. The room decorations that survived revealed excellent craftsmanship.13 In contrast, a Bakufu survey of housing in villages—most of which would have been in central Japan—under its direct administration revealed that as late as 1644 many houses still had no stone bases under the posts.14 The diffusion of the new kinds of housing construction generally followed the progress of economic development within Japan. Thus, the minka built using these new methods were to be found in the Kinai in the mid-seventeenth century, and in the Chubu, Kanto, Chugoku, and Shikoku regions in the latter half of the century. From the lateseventeenth century into the early-eighteenth, minka began to appear in both the Tohoku region and Kyushu.15 The minka took on characteristics of their regions, as the result of climatic differences, differences

12. This treatise, entitled "Shomyo," was written by Heinai Yoshimasa and his son Masanobu. See Nagahara and Yamamura, "Shaping the Process of Unification, pp. 9 0 - 9 1 . 13. Asakura-shi Iseki Chosa Kenkyujo,

Ichijodani

(Fukui: Asakura-shi Iseki Chosa

Kenkyujo, 1981), especially pp. 33 and 4 7 .

Tochi seido-shi 1, ed. Kitajima Taiket Nihon-shi so (Tokyo: Yamakawa Shuppansha, 1975),

1 4 . Oishi Shinzaburo, "Kinsei shakai no seiritsu," in Masamoto, which is vol. 7 of

PP- 34-35. 15. Miyazawa, "Kinsei minka no chiiki-teki tokushoku," p. 154.

HOUSING A N D F U R N I S H I N G S

31

in the building materials available, and cultural variations. However, there was a surprising uniformity to the types of minka built in the various regions from the seventeenth century on, no doubt because the new tools and carpentry techniques were to a great extent imported from the continent and subsequently diffused throughout Japan rather than indigenously generated. However, architectural historians who emphasize regional differences in roof styles, floor plans, and beamwork often overlook or undervalue the common points of this housing found in far-flung parts of the country. By the late-sixteenth century, the floor plans of houses indicate the house was divided into two major sections: the living space and the work space. The work space, even in the grandest houses of the elite, usually had an earthen floor—the doma,}6 The living space originally was in the same room as the work space; only as houses grew large enough to make a distinction was there a demarcation between the two sections in the dwellings of the common people. The very poorest lived on earthen floors, separating the living section from the work area with a low divider and putting hulls and other materials, covered by straw mats, into the living section to make it comfortable for sitting and sleeping. With the passage of time, more and more people added flooring, first wooden boards or slats made of split bamboo, and as they could afford it, tatami. In the early Tokugawa period, while the well-to-do commoners were building minka, the samurai were constructing houses based on the shoin style of the late Muromachi period—though both styles were made possible by the same technological developments of the sixteenth century. The shoin style was developed by the warrior elites and temples from the palace-style architecture of the Heian and Kamakura periods.17 In addition to the shoin, the built-in desk from which this style took its name, buildings in the shoin style included tokonoma (an alcove for displaying art objects), chigaidana (staggered shelves for decorative purposes),£[enkan (a formal entry), tatami (rush-mat flooring), fusuma

16. In mountain villages where wood was plentiful, there tended not to be doma in houses; almost none are to be found in Shirakawa-mura, for example. Okawa, "Minka," p. 85. 17. See Hirai, Nihon jutaku no rekishi, and in English, see Hashimoto Sumio, Architecture in the Shoin Style: Japanese Feudal Residences, trans, and adapted by H. Mack Horton (Tokyo: Kodansha International, 1981), and Ito Teiji, "The Development of Shoin Style Architecture," in Japan in the Muromachi Age, ed. John Whitney Hall and Toyoda Takeshi (Berkeley: University of California Press, 1977).

32

HOUSING AND FURNISHINGS

Nineteenth-century formal room with shoin elements

(sliding wall panels), and shoji (sliding doors). Gardens were an integral feature of the shoin style, and so there was usually an engawa (narrow wooden veranda) combined with shoji and amnio (sliding wooden outer doors) to enable rooms to be opened up to the garden on fine days but protected in the rain. The shoin style fulfilled the samurai need for formal meeting and reception rooms and their desire to live in the increased comfort this style provided. It came to encompass both the formal and the domestic architecture of the ruling class in the Tokugawa period with its widespread adoption during the first half of the seventeenth century. Residences were more often patterned after the less rigid style of the teahouses, termed sukiya, but they still adhered to the principles of the shoin style, usually containing all of its major features. This style of housing was in marked contrast to the early minka, which were well enclosed to protect the inhabitants from the elements as well as possibly from wild animals and human marauders. It made more use of boards instead of whole logs for posts and beams, and thus used less wood. Also, the use of ceilings, sliding doors, and built-in decorative elements required precision carpentry. The design of shoin architecture also meant that while skilled labor was necessary to build it, it required fewer hard-to-replace natural resources, such as old-growth trees. Just as the minka were larger than their predecessors, so too were

HOUSING A N D F U R N I S H I N G S

33

the houses of the ruling class larger in the Tokugawa period than in earlier periods. By the seventeenth century, rooms in the houses of the elite were connected so that houses were all part of one floor plan inside, rather than a series of rooms connected by corridors, as in the earlier shinden style. Large houses under one roof were made possible by the innovations in construction, but I also suspect that incentives for this development were efficiency in space, building materials, and possibly heating. Evidence for this comes from the fact that after the Meireki fire of 1657 in Edo, daimyo residences were rebuilt in the form of large and small shoin rooms set next to each other instead of in the previously prevalent style of a reception room and a shoin under two separate roofs. Furthermore, buildings more commonly contained a single row of formal reception rooms rather than the former double row of rooms.18 This new style of architecture could readily be adopted and adapted by samurai of all status and income classes, who were now gathered in cites and castle towns and could not readily build in the earlier shinden style, which required vast amounts of space that only the highest court nobles had available.19 We know that samurai were building in the new sukiya style not only from extant houses throughout Japan, but also from at least one survey of samurai housing made in the domain of Tsugaru (Aomori Prefecture) in the mid-eighteenth century. A total of 1,189 houses in the castle town of Hirosaki were included in the survey, though only 797 have complete descriptions including a plan of the lot, the floor plan, and the names of all of the rooms. Also noted were the location of gates, privies, the well, fences, what the house fittings were—shoji, tatami, the various decorative elements, whether there was a built-in sink, and a number of other details. Of the 797 houses for which we have details, 80 percent had a zashiki (the most formal guest room); 66 percent had a hiroma (guest room); and 53 percent had a genkan. Only 18 percent of the houses had no formal rooms at all, and these must have belonged to the lowest-status samurai who had neither the need nor the funds for receiving formal guests. The larger the house, the more formal rooms it had, which is logical because status and income were directly correlated among the samurai. Although the shoin style had been formalized over a century earlier, in this regional capital, the only decorative ele18. Hirai, Nihon jiitaku no rekishi, p. 162. 19. For a detailed study on samurai housing, see Sato Takumi, Kinsei bushi jiitaku (Tokyo: Sobunsha, 1979).

34

HOUSING A N D F U R N I S H I N G S

ment found in most guest rooms was a tokonoma. By this time, however, tatami was clearly standard in all samurai houses in this domain.20 The problem of space in premodern towns and cities in which the major method of transportation was by foot led to the development of compact housing for urban commoners as well as for samurai. Medieval scrolls reveal small shops and houses lined up next to the road without any space between the houses or between the road and the house,21 and this tradition was carried on into the Tokugawa period. Since frontage on main streets was of prime importance for merchants, the typical pattern for urban housing was blocks of buildings, with the merchants on long narrow lots facing a major thoroughfare, and laborers and the like renting small, usually one-room, apartments in long buildings called nagaya, which are the Japanese equivalent of tenements. The merchant houses, whether large or small, typically had a shop (mise) in the front where trade or business was conducted, with family (and employee) quarters behind the shop, and storage areas at the rear. The Kyoto pattern was for the earthen-floored kitchen area to be located in a side passage leading to the rear, called a toriniwa, but in the city of Edo there was typically no such passageway. In these buildings too the methods of construction and the design elements were those used in minka, and increasingly as townspeople imitated the samurai, those used in the sukiya as well. There were a number of major differences among cities in terms of housing, including the layout of residential areas, average size, and standard size of tatami but all houses had common elements.22 Thus, by the Genroku period of prosperity at the end of the seven20. Habuka Hisao, "Horeki-ki ni okeru Tsugaru-han Hirosaki joka no buke jutaku no omotemuki ni tsuite," Seikatsu bunka-shi, no. 12 (1987). 21. See Fujishima Gaijiro, Nihon no kenchiku (Tokyo: Shibundo, 1966), p. 114, and Wakita Haruko with Susan B. Hanley, "Dimensions of Development: Cities in Fifteenthand Sixteenth-Century Japan," in Hall, Nagahara, and Yamamura, Japan before Tokugawa. 22. Among the hundreds of books on the configuration of Tokugawa towns and on townhouses, see Nishiyama Uzo, supervising editor for the Kanko Shigen Hogo Zaidan, Rekishi-teki machinami jiten (Tokyo: Kashiwa Shobo, 1981); Koizumi Hiroshi, Edo 0 horu (Tokyo: Kashiwa Shobo, 1983); Jinnai Hidenobu and Itakura Fumio et al., Tokyo no machi 0 yomu (Tokyo: Sagami Shobo, 1981); and CDI (Communication Design Institute), ed., Kyoto shomin seikatsu-shi (Tokyo: Kagoshima Shuppankai, 1973). For one view of differences between Osaka and Edo housing, see Miyamoto Mataji, "Kyoju kara mita Osaka to Edo," in Osaka no kenkyu, ed. Miyamoto Mataji (Osaka: Seibundo, 1967). The best way to see what housing in Edo was like in the mid-nineteenth century is to visit the Fukagawa Edo Museum in Tokyo.

HOUSING A N D F U R N I S H I N G S

35

teenth century (1688-1704), there were at least three distinguishable trends in housing. First, the houses of commoners were becoming permanent, well-built structures made to precise measurements and with a quantum jump in quality resulting from the new tools and techniques made available throughout most of the country. Second, the samurai were increasingly building in the sukiya style, the epitome of which is the Katsura Detached Palace built in the early years of the Tokugawa Bakufu. This style signaled a new, open, and light architecture that in the late-nineteenth century formed the model for modern Japanese house styles. Third, urban dwellings became compact. Even the mansions for the well-to-do were under one roof rather than spread out as in earlier periods. Most commoner houses were long and narrow because of the high cost of frontage, and lower-status samurai lived in similar housing. In an effort to build larger houses on small lots, second and third stories were added when permitted, and even when not, ingenious methods were used to create more than one story of usable space. Despite these trends, many Japanese still lived in simple houses or huts similar to those of the medieval period, but there was a distinct trend for better-quality housing in all classes of society, made possible by the developments in architecture and technology in the late medieval period and by the growing economy and rising living standard of the seventeenth century. Coincidentally, it was also during the seventeenth century that farm housing in Japan began to differ most clearly from that found in Europe and England.23 Primarily constructed of wood, houses in medieval England and houses in Japan through the Tokugawa period were divided into two sections, living and service areas, with a hearth in the center of the living space. In both types of housing, smoke from the fire drifted up to the roof and out through the thatch or holes in the gable ends. Because houses were built for protection from the elements and wild beasts, and given the lack of glass or other appropriate materials to cover windows, houses had only a few small apertures to let in light and air and thus tended to be dark inside. The floor was usually at the ground level and earthen, or it might be raised off the ground, a few feet in the living area in the case of Japan, and in England, three feet if it covered a byre or storage area sunk into the ground under the 23. M. W. Barley, The English Farmhouse and Cottage (London: Routledge and Kegan Paul, 1961), pp. 3-20.

36

HOUSING A N D FURNISHINGS

house. In both parts of the world, people had lived in sunken housing in earlier times. Thus, a medieval Japanese would have understood the use of a farmhouse in medieval England and the principle of having the hearth in the center of the building. Even the small size of ordinary houses in medieval England—one room for the poorest people, two for most of the rest—and the scarcity of furniture and other goods would have seemed natural. Although Japanese housing improved over time, it could be argued that the principles changed less than in the West. People continued to sit on the floor, even when they could afford to install first wooden floors and later tatami, instead of switching to chairs. And the fireplace, when present, continued to be centered within the house, with the smoke still vented through the gable ends of the roof. In Europe, however, the major innovation was "the containment of fire and smoke in a fireplace and chimney, which permitted a second floor to be installed above the great hall and partitioned for use as bedrooms."24 All of this is oversimplified, but we tend to emphasize the differences in lifestyles in Japan and the West in the nineteenth and early-twentieth centuries, and we do not compare them in earlier times when there was no contact between the two parts of the world but when life was, in fact, much more similar. On the one hand, there is a tendency to think of the Tokugawa period as one monolithic era of two and a half centuries because the same government remained in place and because the major social and economic institutions set up in the seventeenth century could still be found in the nineteenth. On the other hand, Marxist economic historians have divided the period into two: rapid growth of both the economy and population in the seventeenth century followed by stagnation in both for the next century and a half. However, viewed from the vantage point of the material culture, the Japanese continued to grow more prosperous in the second half of the Tokugawa period, though the growth and rate of change were almost certainly not as rapid as in the first half. This last view is confirmed by the research of economic historians using modern economic analysis.25 24. Richard L. Bushman, The Refinement of America: Persons, Houses, Cities (New York: Random House, 1992), p. 109. 25. In the past few decades numerous studies confirming this view have been published. Representative is Iwanami Shoten's Nihon keizai-shi series. See volume 1, Keizai shakai no seiritsu, 17-18 seiki, ed. Hayami Akira and Miyamoto Matao (Tokyo: Iwanami Shoten, 1988).

HOUSING A N D F U R N I S H I N G S

37

Two houses in Hida (now part of Gifu Prefecture) exemplify the changes and improvements in housing between the seventeenth and eighteenth centuries. The house of the Nokubi family is typical of houses built from the end of the sixteenth century into the beginning of the seventeenth and has a very closed-in feeling. It has few windows except for small holes about one foot square. The house is large for its period and has two irori or fireplaces on the earthen floored area. The sleeping room and guest rooms have wooden floors; there are no tatami in the house. In contrast, the house of the Wakayama family, built in 1751, has a ground floor constructed by carpenters; the rest of the house was built by villagers. The house has several rooms with tatami flooring, and most of the rest are of wood. It is of a grand style even for minka, four stories high to enable the family to raise silkworms in the upper stories, and with a bath inside the house and a special room for the Buddhist altar several inches higher than other rooms. Even if the families that built these two houses are not comparable in terms of wealth, clearly both the style and quality of housing changed in the century and a half between their construction.26 Housing continued to improve during the second half of the Tokugawa period. From the second half of the eighteenth century into the first half of the nineteenth, there was remarkable development in the minka in every area of the country.27 Sumptuary regulations of the Bakufu and various daimyo with regard to the size, style, and decoration of the houses of the commoners were increasingly ignored as people who could afford larger houses or rooms in the shoin style built them. Japanese retained the open fireplace, or replaced it with stoves for cooking but not heating, and increasingly installed tatami for flooring, while for the most part eschewing large items of furniture. Houses diverged further and further from the European pattern in terms of use, though from the outside minka looked even more like the farmhouses of Switzerland, Germany, and other parts of Europe than they had a century or two earlier when functionally they were in fact more alike.28 Houses were certainly growing larger, and though our information comes from scattered sources, our knowledge of floor plans indicates 26. Both houses are today located in Hida Minzoku-mura in Takayama, Gifu Prefecture, and are described in Hida Minzoku-mura, ed., Hida no minka (Takayama: Hida Minzoku-mura, 1987), pp. +5-46 and 3-8, respectively. 27. Ókawa, "Minka," p. 81. 28. Bruno Taut, Houses and People of Japan (London: John Gifford, 1937), pp. 105-7.

38

HOUSING A N D FURNISHINGS

an increasing number of rooms and more total floor space. By the late Tokugawa period, the average house size was larger than in the seventeenth century, though the trend was not one of steady, gradual increase everywhere. What happened in the village of Kosugaya in Owari Province (Aichi Prefecture) during the seventeenth and eighteenth centuries may in many ways be typical.29 Here the number of houses increased between 1646 and 1684 from fifty-six to sixty at the same time that average house size grew from 7 to 8.3 tsubo.30 But between 1684 and 1734, average house size fell to 7.4 tsubo, while the number of houses in the village rose to seventy. Of five years for which data have been analyzed, this is the only year in which average family size was over five people. It is likely that conditions in 1734 were not indicative of a drop in the standard of living, but rather that the drop in the number of live-in servants or employees, the smaller average house size, and the increase in the number of households all resulted from the trend toward letting servants become tenants and establish their own homes. This trend produced in an increase in houses, a rise in the birthrate, and the crowding seen only in this one segment of the Tokugawa period. Kosugaya is a very small and isolated sample, but it indicates that house size varied greatly by economic position and that the average house was very small by modern standards. Other evidence that seventeenth-century houses were often small comes from restrictions on building sizes. In 1656, in a newly reclaimed area in Musashino Province (near Edo), the magistrate in charge of the area issued specifications on the kind of housing newcomers could build. The size of house permitted depended on the size of the family. A couple could build a house of only seven tsubo, whereas a family with four to five members was allowed a house of eleven tsubo?1 Though it is difficult to draw conclusions from our information on nineteenth-century housing, it is clear that houses not only contained more rooms and living space, but the number of members per family decreased. In Okayama, the average household size fell from seven 29. Hayami Akira, "Kinsei Chita chiho no jukyo to kazoku keitai" (paper presented at the fifty-second annual meeting of the Shakai Keizai Shigakkai, 1983), pp. 7 - 9 . 30. One tsubo is 3.954 square yards. A tatami mat is roughly one yard by two yards in size, the exact size depending on the region of Japan, and therefore a tsubo is two tatami mats in size. 31. Kimura Motoi, "Nomin seikatsu no shoso," Seikatsu shi 2 (Tokyo: Yamakawa Shuppansha, 1965), pp. 207-8.

HOUSING A N D FURNISHINGS

39

people in the early-eighteenth century to five by the turn of the nineteenth century. The most important reason for this drop was a decrease in the average number of children people had, down to three or four.32 But because of the changes in family size over the life cycle of each family, it would have been impossible to regulate house size by the number of members for more than the year in which the house was built, and it is unlikely that this was attempted frequently. The Bakufu issued regulations on houses, exhorting commoners to make theirs simple in style and conform to their status. In contrast to the general prohibitions of the Bakufu, those of the various domains tended to be more specific, though with large variations by domain. In the domain of Aizu (Fukushima Prefecture), the size of the dwelling (as measured by spans) was determined by the assessed output of the farmland owned. In the Murakami domain in Echigo (Niigata Prefecture), a 1740 regulation prohibited farmers from building wings onto their houses. Though farmers needed a large doma in which to do farmrelated work, it wasn't considered necessary for them to have a zashiki, and so these were often forbidden for anyone not a village headman. The easiest way to get around the regulations was to build a house that looked like a working farmhouse from the road, and inside install a room for guests in the sukiya or a modified samurai style.33 Commoners proved to be clever at circumventing regulations. The distance between the main posts holding up the roof of a house was usually limited by statute,34 but farmers could build enormous houses just by incorporating into the house the space that was technically under the overhang of the roof. Since most roofs were high, whole rooms were added in this way, while keeping the size of the house within the letter of the law. City regulations often decreed that houses be no more than one and a half stories high, but townspeople circumvented this clause by building houses that were one and one-half stories at the front with a steep rising roof that slanted up from the street front. In the 32. Susan B. Hanley, "Family and Fertility in Four Tokugawa Villages," in Family and Population in East Asian History, ed. Susan B. Hanley and Arthur P. Wolf (Stanford: Stanford University Press, 1985). 33. Shiraki Kosaburo, Nihon no minka: sono keisei to densbd (privately published by the author in 198+), pp. + - 5 . 34. The motives behind these statutes range from a desire to protect scarce resources, such as old-growth forests, to a desire to maintain status differences between the classes and maximize tax revenues by making the lower classes focus on production rather than consumption.

4o

HOUSING AND FURNISHINGS

middle and to the rear of these houses, rooms on the second floor were of full height. Houses of this type can still be found in the city of Kanazawa, the post town of Kiso Narai on the Nakasendo, throughout the Kyoto area, and in countless other towns and cities in Japan.35 The authorities must certainly have been aware of mass flouting of regulations, but the existence even today of so many Tokugawa houses that did not meet the statutes at the time they were built is evidence that the growing commerce and rising incomes led the officials to ignore anything but the most flagrant violations. Some of the best information we have on changes in housing over the last two centuries of the Tokugawa period is from a Bakufu survey on houses in Tochigi Prefecture. When the shogun paid a visit to Nikko, the large retinue that accompanied him had to be housed along the way, and in order to determine which villages might be suitable for putting up the shogun's men, surveys were carried out on the village housing. Surveys remain for thirteen villages, mostly around Utsunomiya, but two from near Imaichi. The surveys contain the actual floorplans of the houses, with notations as to size, use of rooms, and flooring.36 The Tochigi data clearly demonstrate how dangerous it is to draw historical conclusions, from the models put forth by architects and others to depict historical changes in housing, from the samples of houses that end up in house museums, or even from samples from just one village. These surveys indicate that one cannot safely extrapolate regional trends from a sample of thirteen villages. One village, Tanaka, had a higher percentage of one-room houses in 1823 than did two out of the three village samples for 1712, over a century earlier. There were still residences composed of two separate buildings—with the work space and kitchen under a separate roof—in Nozawa village in 1823, though these were by no means as prevalent as in the three early-eighteenth-century samples. Though building a residence as two independent buildings was once quite common throughout western Japan and 35. For examples in the city of Kanazawa, see Kanazawa-shi Kyoiku Iinkai, ed., Kyuto no kuruwa (Kanazawa: Kanazawa-shi Kyoiku Iinkai, 1975), especially pp. 86 and 90; Toyoda Takeshi, editorial supervisor, Kanazawa zu-bydbu (Tokyo: Bun-ichi Sogo Shuppan, 1977), especially pp. 84 and 143; and Kanazawa-shi Kyoiku Iinkai, ed., Kanazawa no rekishi-teki kenchiku, Kanazawa-shi bunkazai kiyo, no. 57 (1986): 51-54. For examples in Kiso Narai, see Narai Kokuritsu Bunkazai Kenkyujo, Kiso Narai: Machinami chosa hokoku (Nagano: Nagano-ken Kiso-gun Narakawa-mura, 1976). 36. Tochigi-ken Kyoiku Iinkai Jimu-kyoku Bunka-ka, Tochigi-ken no minka (Tochigiken Kyoiku Iinkai, 1982).

HOUSING A N D F U R N I S H I N G S

41

the Pacific side of Honshu, it eventually was retained only in the remotest areas of the country.37 Despite these caveats, it is possible to posit certain trends in housing over the course of these two centuries in Tochigi Prefecture. First, on the average, houses grew larger over time. In the three earliest samples, there were one-room houses and numerous two-room houses in every village. Between 12 and 47 percent of the housing was in these two categories. However, in the later samples, only in Tanaka in 1823 were there any one-room houses, and in three of the villages, there were no one-, two-, or three-room houses by 1823; all had at least three rooms plus doma. The average house size in these villages was twenty tsubo, larger than the seventeenth-century examples, though still small by most standards. If population trends were the same in this region as in most of Japan, families were growing smaller while houses were growing larger. A second trend was for households to be housed under one roof rather than two. In the early-eighteenth-century surveys of Tochigi Prefecture, the larger houses were under two roofs, usually with the doma separately housed. In none of these samples was there a four-room plan under one roof. In the samples after 1750, however, only Nozawa in 1823 had any houses with the doma under a separate roof. Houses in the cold and snowy regions of Japan tended to be large, with the house and barn often under one roof or continuous roofing, and houses in the warm parts of western Japan often had a separate work and cooking area later in the Tokugawa period. But in the mid-portions of the country, the shift was from small houses of one or two rooms for those with lower incomes and large houses under two roofs for the well-to-do to larger houses for all, with all rooms under one roof. As the standard of living in the countryside improved, four-room houses (four equal-sized rooms in the shape of a square) with the doma at one side were diffused throughout the country.38 Although the size of most commoner houses was small by modern standards, there is little reason to believe that Japanese houses were smaller than housing for farm laborers and urban workers in England in the same period or the homes of colonial settlers in the seventeenth century and the log cabins of the eighteenth and nineteenth centuries 37. Kawashima, Minka, pp. 165-66. 38. Okawa, "Minka," p. 101.

42

HOUSING AND

FURNISHINGS

in the U.S. The dwelling of an American pioneer was typically a oneroom cabin, or at most two rooms with a loft. In England, rural laborers in the eighteenth century lived in one-room cottages, some with leantos attached. By the nineteenth century, many English workers could expect to live in semidetached houses with two rooms on the ground floor and two above. An ideal gardener's cottage consisted of two rooms, with total dimensions of 22 by 9-75 feet, but many low-income people still lived in one room with mud floors in the mid-nineteenth century.39 Not only did the Japanese build more expensive and durable houses during the Tokugawa period than previously, but the quality in terms of the level of well-being improved as well. Flooring is one of the best indicators of this change in quality. Again, there are few quantitative studies, but one that does exist is a detailed Bakufu survey made in Bitchu Province in 1865 for the purpose of determining houses suitable for quartering soldiers. Of eighty-six houses surveyed, thirty-three were found suitable for quartering soldiers. Those considered unsuitable had no tatami, only straw mats spread on the floor; undoubtedly most were too small as well. By the Bakumatsu years, every household in this village that owned land yielding two or more koku of grain (ten or more bushels) had tatami in its house; even some farmers with less land had tatami, but those with no land of their own had none. Thus, all but the lowest-income families had six or more tatami mats in their houses, and it is clear that this kind of flooring had become standard in farm houses.40 What is significant in terms of the standard of living is that increasing numbers of people could afford this relatively expensive new housing that required not only more costly materials but carpentry work of a higher quality than was needed to merely build a crude dwelling to provide shelter. Commoners were not only constructing their houses to be more comfortable and durable during the Tokugawa period, they were also imitating the styles of the samurai elite. A document from Kyoto dated 1729 noted that thirty years earlier people had earthen floors in their houses and spread mats of straw or other materials for sleeping and sitting, but nowadays they were not only using wooden

39. John Woodforde, The Truth about Cottages (London: Routledge and Kegan Paul, 1969), pp. 15, 35. + 0 . Fujisawa Shin, "Bakumatsu-ki noson ni okeru kaiso-betsu jutaku kozo ni tsuite," Okayama Daigaku Kyoiku Gakubu kenkyii shiiroku, no. 21 (1966).

HOUSING A N D F U R N I S H I N G S

43

floors, but putting in ceilings and tatami in imitation of the samurai.41 The highest authorities and Confucian scholars alike complained about commoners mimicking their betters—by equipping their houses "with shoji, covered ceilings, karakami partitions, tatami mats and mosquito nets." 42 The numerous regulations against building houses in the samurai style found throughout Japan from the mid-Tokugawa period on are testimony to the fact that many commoners were building houses beyond their station in life. Just as commoners and samurai were building new, more expensive types of housing in the Tokugawa period, they were buying more consumer goods of all kinds. Because the style in nineteenth-century Japan was not wall-to-wall furniture as it was in Victorian England and because traditional Japanese interior decor emphasizes built-in decoration and a minimum of furniture, we can forget just how many things traditional Japanese houses did contain by the nineteenth century and how few Western ones did in the seventeenth and eighteenth centuries.43 Evidence for a proliferation of goods during the Tokugawa period in the homes of commoners, and not just the very rich ones, comes from the development of methods of storage: tansu (chests) for clothes, storehouses for goods of all sorts, and oshiire (built-in cupboards for bedding). The need for containers in which to store household items and the fact that these became ordinary items found in the homes of all but the lowest-income people by the end of the nineteenth century indicates that the average Japanese owned an increasing number of items that had to be stored. This need and the number of objects developed for storage provide good evidence for a rise in personal wealth and the standard of living, and thus considerable detail will be given here. 41. Miyamoto Tsuneichi, "Minshu seikatsu yoshiki no hensen," in Jwanami Koza Nihon rekishi, vol. 23, ed. Ienaga Saburo et al. (Tokyo: Iwanami Shoten, 1972), p. 127. 42. J. R. McEwan, The Political Writings of Ogyii Sorai (Cambridge: Cambridge University Press, 1962), p. 44. In 1788, the high-ranking Bakufu official Matsudaira Sadanobu issued a decree banning luxuries for peasants who had become forgetful of their status. Kodama Kota, Kinsei nomin seikatsu-shi (Tokyo: Yoshikawa Kobunkan, 1957), p. 185. 43. For example, a survey of probate inventories in Wethersfield, Connecticut, for the seventeenth and eighteenth centuries reveals a clear increase in furniture. In the 1640s, the average house did not have a single chair in it, but by the 1790s, the average house contained nearly sixteen chairs. Kevin M. Sweeney, "Furniture and the Domestic Environment in Wethersfield, Connecticut, 1639-1800," The Connecticut Antiquarian 36 (1984): 10-39.

44

HOUSING A N D F U R N I S H I N G S

Tansu developed toward the end of the seventeenth century and were popularized in the eighteenth.44 This development coincides with the first major peak of economic prosperity and indicates that a growing number of people owned more clothing than they could wear at one time, dishes and serving utensils that needed to be stored when not in use, and numerous small items, such as writing utensils, business papers, sewing equipment, cosmetics, and the like. There were also storage trunks, trunks meant for carrying goods from place to place, trunks to fit on ships, and others. This does not mean that there were not storage containers in previous times—chests known as hitsu had been used for centuries—but the development of huge wheeled tcmsu or large tansu designed to fit under a stairway indicated a new level of wealth among farmers and townspeople. In fact, people had so many items to store that numerous varieties of storage chests developed, all with different names: sbo-dansu, cha-dansu, temoto-dansu, ishd-kascme-dansu, kusuridansu, cho-dansu, oshiire-ishd-dcmsu, and so on.45 One of the most popular were huge chests on four wheels, known as nagamochi kuruma, which were kept near the entrance of a house so that when fire threatened, the valuables could be wheeled away. The Meireki fire that burned in Edo for three days in 1657 was observed by a member of the Dutch East India Company, Zacharias Waganaer. So many people were trying to flee with these huge chests on wheels that they clogged the streets, and Waganaer and the members of his mission ended up having to clamber over chests and roofs in order to escape the fire.46 When commoners began to own so many goods that a few chests would not hold them all, they began to build storehouses. These were particularly necessary in cities where fire often destroyed whole sections of town; a storehouse properly constructed might survive the fire and protect the valuables of a household. Storehouses were nothing new; the Shosoin is probably the most famous, but even in the Nara period 208 storehouses were listed in Suruga alone. By the sixteenth century the rich all owned storehouses, but so did many other people. These 44. Koizumi Kazuko, Kajju to shitsunai isho no bunka-shi (Tokyo: Hosei Daigaku Shuppankyoku, 1979). 45. These are, respectively, chests for storing written materials, chests for tea ceremony utensils, small chests for samurai women, stacked chests for clothing, medicine chests, account chests, and clothing chests to store in the closet. 46. Ty and Kiyoko Heineken, Tansu: Traditional Japanese Cabinetry (New York: Weatherhill, 1981), pp. 21-23.

HOUSING A N D F U R N I S H I N G S

45

were used not only to store documents for monasteries and businesses such as pawnbrokers and wholesalers, but they held decorative seasonal goods for individual households. Itoh argues that it was the development of the storehouses that made the Japanese-style house possible: houses in which interiors are kept nearly bare and furniture and other furnishings are brought in as required, enabling the occupants to change the function of the various rooms as needed or desired.47 Clearly, to live this kind of lifestyle there must be someplace where goods are stored when not in use. There is considerable indirect evidence that the number of storehouses was increasing during the Tokugawa period, especially in the seventeenth century. Plasterers were among the lowliest laborers in 1500, but their status advanced as they played an important role in building castles and became indispensable in the Tokugawa period in building storehouses. In Edo, by around 1730, plasterers working on the new type of storehouses, which were much more fireproof than earlier types, were paid three times the fees of the carpenters. Storage in Edo along the banks of rivers and canals became such a problem by the mid-sixteenth century that the government encouraged the building of storehouses and issued blanket permission for their construction. They prevented timber and other goods stored in the open from becoming a fire hazard, and a line of storehouses in itself often acted as a firebreak. Not only are storehouses frequently mentioned in stories of the great fires in Edo in the seventeenth century, but they were even more frequently built thereafter in an attempt to deal with fire. One philosophy was that city buildings were going to burn down anyway; one might as well be resigned to this fact, build flimsy houses, and put all of one's money and valuables into a good storehouse and into a new stock of timber kept in the suburbs for rebuilding after the inevitable fires. Storehouses were not only built for protection against fire and theft and as a place to store inventory and articles not in use, but they were a major status symbol in the Tokugawa period. As Saikaku put it: "First, sacks of rice; second, a two-story house; third, a three-story storehouse."48

47. Teiji Itoh, Kura: Design and Tradition of the Japanese Storehouse, abridged ed. (Seattle: Madrona Publishers, 1980). See pp. 16, 27, 76, 82, 75, and 38 for material cited in this section. 48. Ibid., p. 38.

46

HOUSING A N D F U R N I S H I N G S

Japanese houses were traditionally built without interior storage cupboards, just as they were in the West. But whereas Westerners slept in beds and kept the bedding on them year round, the Japanese slept on the floor. Today bedding is put away in the daytime in the oshiire, but what did people do before these storage cupboards were built into their houses? The answer seems to be that they had little bedding to store. People slept with their garments spread on top of them, and so they wore their "bedclothes" in the daytime. When commoners began to partition their houses, they made one corner, that farthest from the door and with no window, into a combination storage room and sleeping room, usually called the nando. Strange—and cold—as it seems to us today, it appears that most people didn't use what we think of as bedclothes until well into the Tokugawa period. The nando had a wooden floor, was surrounded by walls on three sides, and had a high sill on the fourth where there was a door. The room was usually the smallest in the house, and a number of people crowded into such circumstances at night would have meant that body warmth went a long way toward keeping them comfortable. People put straw and other field wastes—in some places, seaweed—on the boards for warmth and comfort, and when it was really cold, people in the Toyama-Niigata area in 1835 slept in their clothes by the fire or slept in straw bags, one couple to a bag.49 When cotton began to be widely used in the seventeenth century, more references to bedding appeared.50 With the introduction of proper bedding for those who could afford it, there arose a need for a place to store it in the daytime. Much of the early bedding was for guests and important persons who would not have slept in a crowded nando, and guest rooms or sitting rooms were used for general purposes during the day.51 It was not practical to lug bedding out to a storehouse, and hence storage areas were needed within the house itself. Thus, houses of the well-to-do gradually began to include oshiire during the latter half of the Tokugawa period. The first of these were chests built into the wall of the nando or sleeping room; later these storage areas

49. Ogawa Koyo, Shinjo to shingu (Tokyo: Yuzankaku, 1973), PP- 170-71. 50. The terms yogi (bedclothes) and futon (quilt) didn't become general terms until the first half of the seventeenth century. 51. Although Westerners had a tradition of putting beds in parlors and rooms for general use, the Japanese had neither a tradition of beds nor would this have been practical with their small houses and use of tatami. The introduction of beds in contemporary Japanese homes has taken place only with the adoption of Western-style rooms.

HOUSING A N D F U R N I S H I N G S

47

were incorporated into the floor plan of the house at the time it was built. Normally they are the size of one tatami mat, with a shelf in the middle to divide it into half, and entered through fusuma, which means that it can be difficult to tell whether a cupboard or another room lies beyond the sliding doors. Although oshiire are a part of what we now consider traditional housing, they were in fact a relatively late development that came only when Japanese had thick quilts and other items to store. An economic historian interested in quantitative data in order to precisely assess the changes in the standard of living from the earlyseventeenth century to the mid-nineteenth would be most unsatisfied with the evidence presented here. But taken as a whole, this evidence, partial and anecdotal as most of it is, substantiates the hypothesis that the Japanese increased their personal wealth, and with it their standard of living, over the course of the Tokugawa period. This, of course, corroborates the economic historians who on much better evidence have effectively argued that the Japanese economy grew, not only during the seventeenth century, but in the eighteenth and nineteenth as well. What is surprising is that specialists in Japanese architecture and furnishings, none of whom is trained in economic history and whose work lies outside the debate on whether and how much the Tokugawa economy grew, are convinced that the changes they observed during the Tokugawa period were caused by the growth of the economy and a corresponding rise in personal wealth. Miyazawa Satoshi stated that the minka began to appear from the end of the sixteenth century into the seventeenth because of changes in lifestyles accompanying the rise in productivity in the farm villages, the diffusion of architectural techniques, and the establishment of the ie or Japanese family system. The development of minka in each region of Japan followed the economic progress of that region, and thus the first minka were seen in the most advanced area around Kyoto and Osaka.52 Okawa Naomi saw a remarkable development of the minka in size and construction methods during the three hundred years from the mid-sixteenth to the mid-nineteenth centuries, taking place against "a background of an accumulation of wealth and improvements in life for the masses."53 Shiragi Kosaburo believed the biggest reason for the es52. Miyazawa, "Kinsei minka no chiiki-teki tokushoku," p. 154. 53. Ókawa, "Minka," p. 81.

48

HOUSING A N D F U R N I S H I N G S

tablishment of a new living environment in the developed regions around Osaka from the middle of the Tokugawa period on was that the strength of the economy had advanced beyond both an exchange and barter system and a feudal system, and this enabled the well-to-do merchants to flout the restrictions against building in the sukiya style. According to him, the minka have to be understood in terms of the economic strength of the commoners. 54 Koizumi Kazuko, an expert on Japanese furniture and furnishings, saw a pattern that whatever belonged to the privileged classes of one period in Japanese history becomes sought after and finally attained by the ascendant class in the next. . . . By the end of the eighteenth century . . . the merchant townsfolk were enjoying a life of comfort such as they had never known. . . . The custom of making bridal furnishings even passed down to ordinary commoners, especially in western Japan. . . . These furnishings ceased to be the showpieces of upper-class weddings, but now met the everyday needs of a lifetime, from marriage on. Typical items included chests, trunks, hampers, mirrors and mirror stands, clothes racks, cosmetic cases, and sewing boxes.55 In short, during the Tokugawa period the balance shifted from the preponderance of income and wealth being held by the samurai and a relatively small number of others to a more even distribution, with commoners in the large urban centers and larger landowners in the countryside possessing wealth and incomes equal to or greater than those of many samurai. 56 Though many Japanese were poor by nearly any standard at the end of the Tokugawa period, even an overview of the developments in the material culture offers good evidence that not only was the wealth of Japan as a whole increasing, but it was increasingly being shared by the commoners. In looking at the consumption side of the equation, we see that the quality of housing rose in terms of size, materials, carpentry, and furnishings, and this means that the standard of living rose because people had more money to spend on housing. As the evidence on housing and furnishings makes clear, not only were Japanese building better-quality houses with more furnishings and other goods, but the changes made life in general more healthful. 54. Shiraki, Sumai no rekisbi, pp. 9-10. 55. Kazuko Koizumi, Traditional Japanese Furniture (Tokyo: Kodansha International, 1986), p p . 1 6 4 , 1 7 2 , 173-

56. Kozo Yamamura, A Study of Samurai Income and Entrepreneurship (Cambridge, Mass.: Harvard University Press, 1974), pp. 120-33.

HOUSING AND FURNISHINGS

49

Raised flooring put people above the damp, cold ground and allowed air to circulate under the floor. Flooring of split bamboo and boards laid with spaces between them was cooler and more comfortable in the hot, muggy climate of western Japan in the summer. The new techniques of construction opened up the houses to the sun in the winter and fresh air year round, highly desirable especially in the milder and more humid regions of Japan. Instead of closed houses with few openings other than the door, the residential style adapted from the shoin style substituted sliding doors for immovable walls. As Morse put it, "the whole side of a house may be flung open to sunlight and air."57 The trend toward raised floors and the new style of open houses did not benefit everyone. Only the well-to-do could afford houses with shoji, engawa., fusuma, and the other features of the sukiya style that created the open, airy houses Morse refers to. But whereas the earthen floor was the norm in Japan in the early-seventeenth century, by the nineteenth, it was not. In the village of Nishikawa in Bitchu-no-kuni (Okayama Prefecture), all of the houses had flooring by the Bakumatsu period, though the farmers with the least amount of land could still not afford tatami.58 The lower-income families had sharply slanted roofs without eaves and consequently had plastered walls with no openings in the rear. Such houses were ill-suited for circulation of air and sunlight because the backside without the door was virtually sealed. The higher the income group, the larger the rear door, the more rooms the house had, and the more healthful the environment. Similarly, in the Tochigi sample of thirteen villages dating from 1712 to 1842, as houses grew larger over time, they tended to have not only flooring but tatami, as did virtually all of the houses with four or more rooms. By the time Edward Morse visited Japan in the 1870s, he expected all houses of the middle classes—that is, houses belonging neither to the rich nor poor— to have tatami in all but the kitchen and bathroom. The increase in furnishings is an indication not just of a rise in the standard of living but in the level of physical well-being. This is especially true for bedding, which meant that people slept warmly in winter. Those who were ill could be better cared for and did not have to huddle 57. EdwardS. Morse, Japanese Homes and Their Surroundings (1886; reprint, Rutland, Vt.: Charles E. Tuttle, 1972), p. 7. This was also noted by others, including Henry T . Finck, Lotos-Time in Japan (New York: Charles Scribner's Sons, 1 9 0 4 ) , p. 45. 58. Fujisawa, "Bakumatsu-ki noson ni okeru kaiso-betsu jutaku kozo ni tsuite," pp. 37-55-

5o

HOUSING AND FURNISHINGS

next to others to obtain minimal warmth, thus spreading their diseases. Bedding not only made life a good deal more comfortable but on the margin must have extended life expectancy as well. As the Tochigi and other samples show, the improvements in the standard of living and level of physical well-being did not benefit everyone or occur at the same pace over time, but a comparison of housing and furnishings in the seventeenth century with the mid-nineteenth clearly shows that on average people became wealthier and the material culture more healthful.

CHAPTER

THREE

A Resource-Efficient Culture

The Westerners who visited Japan shortly after it was opened in the nineteenth century found a lifestyle totally different from the one they were accustomed to. Some, like Edward Morse, were charmed by it; others found it extremely uncomfortable. Hugh Wilkinson, who visited Nara in the 1880s, found his hotel "only a thin framework house of wood, with sliding screens of thinnest wood and paper for walls. . . . The furnishing of the place could certainly have cost but little; nor can it be said to be in accordance with our ideas of comfort. There are no chairs, or tables; neither beds, washstands, chests of drawers, basins, nor looking-glasses encumbered the rooms.'" Morse commented: "An American finds it difficult indeed to consider such a structure as a dwelling, when so many features are absent that go to make up a dwelling at home,—no doors or windows such as had been familiar with; no attic or cellar; no chimneys, and within no fire-place, and of course no customary mantle; no permanently enclosed rooms; and as for furniture, no beds or tables, chairs or similar articles,—at least, so it appears at first sight." 2 Nor did Westerners find the food to their taste. A meal that the shogun sent to Lord Elgin's mission to Japan included what to Laurence Oliphant "seemed to be pickled slugs." Oliphant tasted everything sent 1. H u g h Wilkinson, Sunny Lands and. Seas: A Voyage in the S.S. "Ceylon" (London: J. Murray, 1883) as q u o t e d in H u g h Cortazzi, Victorians in Japan (London: T h e Athlone Press, 1987), p. 249. 2. Edward S. Morse, Japanese Homes and Their Surroundings (1886; reprint, Rudand, Vt: Charles E. T u t d e , 1972), p. 7. 5i

52

A RESOURCE-EFFICIENT CULTURE

in, "an experience from which I would recommend any future visitor to Japan to abstain."3 There are also humorous and complaining accounts about problems with baths, bedding, footgear, and just about every other aspect of Japanese life. The Japanese were well aware of how different life was in Japan from the West and the problems Westerners encountered. Inoue Jukichi, writing in 1910, noted that "our rooms look very bare to foreigners and appear to lack comfort to those who have lived in European apartments; but from the Japanese point of view, rooms furnished in the approved European style suffer from excess of furniture and partake too much of the nature of a curiosity shop or a museum." 4 What occurred in the second half of the nineteenth century was a clash of cultures, two different traditions whose material cultures were further apart than at any other time from the medieval period to the present. Western Europeans and Americans were undergoing industrialization, and the ability to produce goods cheaply for mass markets had created the fashion for houses filled with furniture, objects of art, and other furnishings so that no spot remained unadorned. Dress was elaborate, with yards of cloth in women's skirts, and the food, at least in England and the U.S., centered around a diet of meat and was heavy in starches, root vegetables, and dairy products. In strong contrast, Japan during this same period was "closed" to influences from other countries and imported neither culture nor industrial technology. Instead, everyday life became thoroughly Japanese in flavor, even those aspects that had originally been imported from China hundreds of years earlier.5 While the Japanese had gone in their own direction in near total isolation, the West had substituted industrially produced objects of daily life for the handcrafted objects of premodern times. Because the Western material culture was the more technologically advanced, it was implicitly, if not explicitly, held up as the standard by which Westerners and Japanese alike judged Japanese culture in the late-nineteenth century. And because Japanese material cul3. Laurence Oliphant, Narrative of the Earl of Elgin's Mission to China and Japan in the Tears ¡8S7, '58, '59, vol. 2 (Edinburgh: W. Blackwood, i860) as quoted in Cortazzi, Victorians in Japan, p. 254. 4. Jukichi Inoue, Home Life in Tokyo (1910; reprint, London: KPT Limited, 1985), pp. 41-42.

5. A number of scholars have argued this. See, for example, Tatsumura Ken, Nihon no kimono (Chuko Shinsho 120) (Tokyo: Chuokoronsha, 1966), p. 77, and Kato Hidetoshi, "The Significance of the Period of National Seclusion Reconsidered," The Journal ofJapanese Studies 7, no.i (winter I98I):85-IO9.

A RESOURCE-EFFICIENT CULTURE

53

ture was so very different from Western and followed principles that in many ways were the opposite, it was found wanting, even by those Westerners who found Japan charming. However, if we look at Japanese material culture from viewpoints other than that of Western technology, it does not seem so inferior. In the nineteenth century, as well as today, many Westerners admired the artistic qualities of Japanese houses, though they found sitting on the floor uncomfortable. They could appreciate the aesthetic quality of the banquets even if most of the food wasn't to their taste. Setting aesthetics aside, even in standard of living and quality of life, Japan did not fare badly by the middle of the nineteenth century if our measure is not the quantity of goods and durability of structures but the quality of life the material culture provided. What the Japanese managed to achieve was a material culture that provided for their physical well-being but used resources economically. In any premodern culture, where transportation facilities are limited and foreign imports nearly nonexistent for commoner use, the people have to create a material culture from resources at hand. Though this was true for Japan until the late-nineteenth century, it was particularly true in the Tokugawa period because the trade that had started to occur in the sixteenth century was largely cut off through the closing of the country in the seventeenth. People in the Tokugawa period knew they had to rely on only the plants and animals available in their islands because they were not allowed to import, with minor exceptions, nor were they allowed to leave the country. Or as one Japanese has put it, they thought of their islands as people think of the earth today, their total environment within which they had to live/' By making optimal use of resources in the long run rather than maximum use in the short run, the Japanese were able to maintain a dense population on a relatively small amount of land on which to live and feed themselves. The Japanese had made good use of their resources for more than a millennium, and by using them wisely and having the good fortune not to be devastated by a disease such as the plague, by the mid-eighteenth century the country had a population of nearly thirty million on about 370,000 square kilometers of land, approximately one-fifth of which was the island of Hokkaido, largely unpopulated at this time. In contrast, France with 534,000 square kilometers to support approximately twenty-four and a half million people, and 6. Irie Takanori, Nibonga tsukuru shin-bummei (Tokyo: Kodansha, 1992), p. 150.

54

A RESOURCE-EFFICIENT CULTURE

England with an estimated population of nearly six million on just over 130,000 square kilometers had almost double the land per capita that Japan had (forty-six and forty-four persons per square kilometer respectively, compared to about eighty for Japan, or one hundred or more with Hokkaido excluded).7 Furthermore, only about 15 percent of Japan is considered arable, in contrast to England, about 50 percent of which is suitable for fields or pastures. But if much of the land in Japan is too steeply sloped for cultivation, the climate and soils compensate. There were numerous renewable and plentiful resources with which the Japanese built their houses: wood, reed, straw, bamboo, bark, and other natural materials. These materials were not only readily accessible, but lightweight and therefore safer in earthquakes than stone or brick dwellings. The basic building material was wood, which was used for the posts and beams—the structure of the house. It could also be used for the walls, ceiling, roof, flooring, and finishing and decorative elements, but so could a number of other materials, which were all more readily replenishable than wood. Farmers grew rush for tatami and reed for thatch and used the waste from fields, such as straw and hulls, for wall material and cushioning for straw mats to lay on the floor. Clay and plaster were frequently used for walls. Thus, people could afford to rebuild their houses every generation or so, and in a damp climate such as Japan's with houses built of wood and other natural materials, such rebuilding was desirable and often necessary. The gradual scarcity of good wood for beams and posts was certainly a major factor in the switch from building commoner houses using enormous logs of primeval timber to the lighter frame houses of the nineteenth century. By the mid- to late-seventeenth century, there was already a growing scarcity of good quality lumber as the forests were depleted in building castles, mansions for high-ranking officials, large farmhouses for well-to-do commoners, and other large buildings in demand as a result of the growing economic prosperity and the requirements of the ruling class. Japan's woodlands were also threatened by the increasing demand for fuel. To counter the growing problem of scarcity of good timber, various levels of government, both domain and the Bakufu, enacted regulations against overcutting of the forests, 7. The population estimates for France and England are from Michael Anderson, Population Change in North-western Europe, 17S0-18S0 (Basingstoke: Macmillan Education Ltd., 1988), p. 23. For Japan's, see chapter 3 of Susan B. Hanley and Kozo Yamamura, Economic and Demographic Change in Preindustrial Japan, 1600-1868 (Princeton: Princeton University Press, 1977).

A RESOURCE-EFFICIENT CULTURE

55

and as a result, there were few mountains stripped bare of trees by timber cutting. Other measures were the creation of tree plantations, closing forests to cutting, and an innovative measure known as wariyama designed to give incentive to the people to care for their trees. Wariyama allocated woodlands to certain households for specific, usually long, periods of time, which encouraged those in charge to protect the woods in order to use them and pass these economic benefits on to their descendants.8 Though Japan had seemed headed for the ecological disaster of deforestation during its rapid economic and demographic growth during the seventeenth century, the countermeasures put into effect during the seventeenth and early-eighteenth centuries were effective in preventing this by converging government interests with private interests.9 One result was the construction of fewer large minka than would otherwise have been built, because the huge logs used whole for these houses became less available as this resource was depleted. However, the new techniques of construction using sawn lumber mean that a switch in building styles resulted in a rise in the level of physical well-being. Houses partitioned into enclosed rooms with ceilings were warmer, and they could be more easily decorated, resulting in more aesthetically refined styles as commoners adopted the elements of the shoin style of the samurai for their farmhouses. Although Western visitors found the lack of furniture, particularly chairs, uncomfortable, some could see the advantage. The paucity of large items of furniture meant that small rooms could accommodate more people than if there were chairs, rooms could be used flexibly so that sitting rooms could be turned into bedrooms at night, and houses could be furnished more cheaply than in the West.10 But the Westerners who commented on the emptiness of houses and the lack of furniture were looking at things from a Western perspective and missed seeing 8. The problems and countermeasures regarding the forests have been thoroughly documented by Conrad Totman in The Green Archipelago: Forestry in Preindustrial Japan (Berkeley: University of California Press, 1989). 9- At the same time that government tightened its regulation of forests in the seventeenth century, plantation culture was developed, various arrangements for multiple use were created, and forests were leased under agreements that prevented their devastation and encouraged their renewal. But typically, when use of woodland was divided between government and villagers, the rulers retained rights to the large timber and villagers to undergrowth for use as firewood and fertilizer. See part 2 of ibid, for detailed discussion of the variety of arrangements used. 10. Morse, Japanese Homes, p. 114.

56

A RESOURCE-EFFICIENT CULTURE

what was important in many cases. What the foreigners saw was the lack of their own tradition, rather than a different tradition that made life comfortable for those accustomed to it and was more economical and efficient given the climate and type of housing. Japanese houses were not empty in the way that Western houses were when all the furniture was removed. As Koizumi Kazuko has pointed out, what would be retained as movable furniture in other countries tended to become built in as part of the house in the Japanese tradition." Tatami is a clear example. Originally movable mats used for sleeping and sitting in the Heian period, they gradually came to be built in, first as covering for the part of the room on which people sat, and then the entire floor. This also happened to shelves and the desk in the shoin style. Moreover, folding screens used to partition off part of a room became sliding wall panels, and in the Tokugawa period chests or cupboards to store bedding became built-in closets. Even decorative elements were built in. Instead of having to decorate bare walls of odd lengths with windows and doors scattered here and there, design principles dictated where the tokonoma and other built-in elements would be, and the partitions and the posts formed the primary decorative elements in the house. Except for adding seasonal decorations in the tokonoma and a few cushions to sit on, a Japanese room was complete and ready to be lived in when the structure was finished. Traditional Japanese furnishings were not the heavy articles of furniture associated with Western houses, particularly in the nineteenth century—wardrobes, four-poster beds, dressers, sideboards, sofas, tables to seat the entire family, and the like. Rather, the furnishings were readily movable, so that they could be changed with the season or with the use. Today a traditional room can be easily transformed from a sitting room for two to a dining room for eight or a sleeping room for one. With the exception of tansu used for storage, furnishings tended to be small and light—instead of sofas and chairs, there were cushions to sit on; instead of large wooden beds, quilts that were folded and put away during the day became the custom; and sewing boxes, cosmetic stands, lights, and heating devices were all portable. Since people sat on the floor, it wasn't necessary for furnishings to have legs, and when they did, they needed to be only a few inches high. This in 11. Kazuko Koizumi, Traditional Japanese Furniture (Tokyo: Kodansha International Ltd., 1986), p. n.

A RESOURCE-EFFICIENT CULTURE

57

itself made it possible to have furnishings that could be stored away, moved from place to place, and made with efficiency of materials. Nor did the Japanese make much use of large tables during the Tokugawa period. There was a proliferation of small tables and serving trays the type and style of which differed by social class. Those who could afford a more gracious lifestyle than eating out of a single pot usually ate at these individual trays raised off the floor a few inches on short legs. Sometimes the individual trays had drawers under them so that after a meal each person wiped out the dishes and utensils on the tray and stored them away until the next meal. In households in which each person had an individual dining set, this custom was fairly sanitary, but among the lower classes—and in whorehouses as well—eating utensils were often shared and used in turn. In farmhouses with openhearth fireplaces, the family customarily sat around it and shared food from the pot over the fireplace. In a country short of trees and houses short of space, these customs that made rooms multifunctional were both resource- and space-efficient.12 The primary reason this style of house and furnishings developed was that the Japanese belonged to the tradition of cultures in the world that did not develop chairs. In fact, until modern times, chairs anywhere in the world seem to have been the exception rather than the rule. We think of them as a permanent feature of Western culture, but this is because we see pictures of the chairs of Egyptian kings and of wealthy Greeks and Romans; most people did without. If they sat on anything, they sat on stools, whatever served as beds, or benches. Much of what there was to sit on was not very comfortable. In the Middle Ages in Europe, chairs were not for comfort; they were instead symbols of authority. Thus, prior to the development of overstuffed furniture and chairs that mold to the body, there was no reason in terms of comfort to prefer sitting on a piece of furniture rather than on the floor itself, unless animals were kept in the house—and they were not in Japan.13 In medieval times Western and Japanese uses of the home were not very different in principle. Medieval houses in both parts of the world were sparsely furnished, and multiple activities took place in one large

12. Ishige Naomichi, "Shokutaku bunka-ron," Gendai Nihon ni okeru katei to sbokutaku, ed. Ishige Naomichi and Inoue Tadashi, a special issue of Kokuritsu minzokugaku hakubutsukan kenkyu hokoku, no. 16 (1991): 18-22 and the appended materials. 13. Witold Rybczynski, Home (New York: Viking Penguin, 1986).

58

A RESOURCE-EFFICIENT

CULTURE

room. Furniture in the West was movable and often collapsible, just as it was in Japan. In towns, houses were divided into work areas and living areas, and both were large open spaces. Though the clothing and other customs were very different, lifestyles were not. Witold Rybczynski argues that it was during the seventeenth and eighteenth centuries that ideas of comfort in the home began to appear in Europe, 14 and it was precisely during these centuries that the ordinary Japanese began to live in a new kind of house with tatami to provide comfort for resting, just as Europeans began to put padding on their chairs. The new styles of house design and living were ushered in both in Europe and Japan in periods of economic prosperity. What made them so fundamentally different was that in Europe, in adding to the comforts of home, the emphasis was on increasing the number of goods and furniture. With the Industrial Revolution, an increasing number of people were able to participate in this new level of comfort, and homes were crowded not only with beds, tables, and chairs, but knickknacks, shelves, table covers, and so on. In contrast, in Japan the seventeenth-century peak of prosperity was accompanied by rapid population growth and pressure on resources, leading to an affluence that not only made the most of scarce resources, but also made a virtue of necessity and continued a form of luxury in austerity. Though the advances in technology and the economy made the new styles of housing possible for many in the Tokugawa period, the styles themselves can be traced back to the Heian period. The shinden style of the Heian aristocracy is characterized by houses of "bare stagelike settings of wooden floors and columns." Screens, curtains, and tatami were moved around to form a room to suit the needs of the moment. The concept was traditionally called shitsumi, referring to "the act of providing and arranging articles so as to create a room for some given purpose or activity." Koizumi argues that although the style of housing changed through the medieval period into the Tokugawa, from the shinden to the shoin, this practice of shitsumi continued and was what made traditional Japanese rooms so versatile in terms of function. 15 These two traditions, movable furniture and sitting on the floor, enabled the Japanese to develop lifestyles that even at their most luxurious were resource-saving. This tradition may have been at least partly in-

14. Ibid., pp. 90-107. 15. Koizumi, Traditional Japanese Furniture, p. n, and Koizumi Kazuko, Kajju (Tokyo: Kondo Shuppansha, 1980), pp. 18-21.

A RESOURCE-EFFICIENT CULTURE

59

fluenced by Zen Buddhism, but whatever the origin the result was a style of life for the elite that depleted resources less than the lifestyles of the wealthy elites in the West. In fact, it should be noted that even the word for luxury has different connotations in Japanese than it does in English. At the same time that zeitaku means "luxury," it connotes "waste." To be zeitaku means both "to indulge in luxury" and "to be extravagant," and today the word is used to mean "wasteful," whereas the English word luxury would never be used to refer to waste. Wood was needed for the frame of a house and for major decorative and structural elements such as posts, but in the Tokugawa period walls and flooring were decreasingly made of wood in favor of tatami, paper sliding doors, and plaster. Very little furniture was needed for function, decoration, or displays of wealth. Instead of displaying all of the artwork a family owned, one sample suitable for the season or a special occasion would be hung in the tokonoma, accompanied by one or two other objects such as a vase. The focus was then on the one item, which is what the family was proud to present to guests. Each item may have been luxurious, and this very luxury was all the more apparent for there being only that single object to admire, competing with nothing else. This austere form of luxury is apparent in many of the traditional art forms that became popular among those who could afford them during the Tokugawa period. One example of this austere form of luxury is the art of flower arranging. In Europe, the moreflowersthat are crammed into an arrangement, the more luxurious it is considered. Quantity is a major measure of quality. In contrast, in Japan, modern flower arranging (ikebana), considered to have originated in the mid-fifteenth century, uses only a few flowers to create a piece of art. This art form has elaborate traditions, but all give symbolic meaning to various elements of the arrangement, meaning that each element, each branch or flower, can be distinguished from another, in contrast to the large number of flowers that make up the Western bouquet. In economic terms, this means that ikebana requires far fewer flowers and branches for an arrangement than does a bouquet. The materials for an elaborate arrangement in Japan would look too sparse for anything other than a small homely bouquet in the West. As the art of flower arrangement filtered down to the commoners in the Tokugawa period when the merchants began to enhance their lifestyle with this art form, it was further simplified so that there were only three main branches to each arrangement rather than the seven or nine in the early styles. Whether one prefers the Western tra-

6o

A RESOURCE-EFFICIENT CULTURE

dition of abundant displays of flowers or the Japanese tradition favoring focus on the individual flowers and branches that make up the display is a matter of cultural background and taste. But it can objectively be said that the Japanese created an elaborate art form out of an artistic enhancement of a room, and they did so in a manner that was resourcesaving and economical. Tokugawa developments in comfort within the house were equally resource-saving, particularly with regard to the use of heat. Medieval homes, it is true, were inefficiently heated in both the West and Japan by open fireplaces in the center of the main room. Much of the heat went up to the roof and houses were smoky. Then fireplaces were developed in the West, which offered a greater degree of safety, permitted the development of ovens, and alleviated the problem of smoke, but they were not an efficient way to heat a room. The Japanese retained the open-hearth fireplace in the center of the room—the ¿ran—which permitted people to sit around it, warming a larger number simultaneously, while the heat and smoke dried and preserved items hung in the rafters and even protected the thatch roofs from rot and bugs. As people began to install wooden floors in their houses, the irori was moved to a room with a floor, so that people sat around it with more comfort, cooking, eating, doing various tasks, and socializing.16 Even as they developed new techniques to more efficiently provide heat, Westerners continued to follow the principle of heating space. They created more efficient fireplaces, later stoves, and finally furnaces that would heat the air and occupants within a room. This type of heating device, whether wood or coal is used, consumes much fuel, and with the rapid growth of population in Japan in the seventeenth century, fuel became a relatively scarce commodity, thus an expensive one that people could ill afford to use to heat the air. In order to save on fuel, the Japanese developed methods of providing heat using the principle of heating the body rather than the air in the room. They also made maximum use of the climate and weather to 16. Suzuki Yoshio and Takahashi Mayumi, "Nagashi, irori soshite kamado," Seikatsu bunka-shi, no. 8 (August 1985): 30-34. This information was gathered from a survey taken in 1955 of over three hundred minka that were studied to determine if they could be restored as national treasures. Most of the houses in this study were built between the seventeenth and nineteenth centuries, and building peaked in the eighteenth century. It is thought that in most cases the heating and cooking arrangements noted in the survey were the original, as it is usually possible to determine when renovations are made if the fireplace or stove has been moved. In 87 percent of the sample, the irori were located in wooden-floored rooms rather than in the doma.

A RESOURCE-EFFICIENT CULTURE

61

provide a comfortable atmosphere for people, rather than trying to keep it out. It has long been said that whereas enclosed European homes with thick stone walls are made for winter, open Japanese houses are made for summer, but this is only partially true. European rooms warmed only by space heaters or fireplaces are cold a short distance from the source of heat, and in summer the thick stone walls can make houses uncomfortably damp and cold even in pleasant weather; an enclosed house can be suffocatingly hot if breezes are kept out. In contrast, the type of Japanese housing that became increasingly popular during the Tokugawa period took full advantage of summer breezes and what sun there was in winter. And in inclement weather, rooms could be made into small cozy spaces by closing all the fusuma and shoji, and in the worst weather the wooden storm doors as well. To minimize the use of scarce and expensive fuel, methods were developed to warm the body without wasting fuel: the hibachi (an open charcoal burner), cmka (a container for hot charcoal used to heat the feet or bedding), and kotatsu (a small heater placed under a table that is covered by a quilt to keep in the heat). These heaters used charcoal and coals and thus were extremely fuel-efficient.17 Hibachi are extant that were made during the Nara period, but this form of heater, which really warms only the hands of those sitting around it, was limited to the rulers and the wealthy who could afford the metal necessary for a fireproof container. It was only during the Tokugawa period, when mining technology that was developed during the Sengoku period was combined with the prosperity of the economy, that ordinary folk were able to afford hibachi. By the end of the Tokugawa period, ceramics had developed to the point that hibachi no longer had to be metal, and this further popularized them. Hibachi could be placed anywhere, so guest rooms could be used year round; all one needed were a couple of pieces of charcoal. Portable heaters that used charcoal were especially convenient and could even be used under quilts and thus taken to bed at night. The heat lasts longer than a hot water bottle's and they do not leak, but they did have to be made of materials that would not burn. From the tradition of the irori and the anka developed various types of kotatsu. Whether the heater was a hearth sunk below floor level so 17. The growth in the charcoal industry in every region, technological advances in charcoal production, and the growth of the charcoal market are well documented. See Higuchi Kiyoyuki, Mokutan no bunka-shi (Tokyo: Higashi Shuppan Kabushiki Kaisha, 1962) and Nihon Mokutan-shi Hensan Iinkai, ed., Nihon mokutan-shi (Tokyo: Zenkoku Nenryo Kaikan, 1960).

62

A RESOURCE-EFFICIENT CULTURE

that one sat as at a Western table, or a portable heater set on the floor, the table combined with quilts raised indoor life to a new level of comfort in the Tokugawa period. Kotatsu made any quiet, indoor activity more comfortable and enabled hand work or study that would otherwise have been uncomfortable or impossible because of the cold. Movable hibachi and kotatsu meant that heaters could be moved to sunny spots of the house, enhancing comfort and enabling people to make maximum use of light. These devices, combined with the shoin-type of architecture, allowed large houses to be fully utilized year round and made possible a new lifestyle in which guests could be served away from family activities and in which family members could work or relax separate from one another. Here is an example of the development of resource-saving techniques that also made life more comfortable and gave people a higher level of physical well-being. Because of growing scarcity of fuel, especially on the plains where the largest concentrations of population were, people used ever-increasing ways of conserving fuel. One innovation that conserved two kinds of scarce resources was resting a pot on an iron ring in the irori so that it did not sit directly on the fire. Not only did this mean that earthenware pots instead of precious iron could be used for cooking, but that a small fire could be maintained under the ring instead of a large fire that would heat a hanging metal pot. Charcoal, which used fewer resources than directly burning wood and which could be more readily transported, was sufficient for this new method of cooking.18 How much the Japanese came to depend on charcoal is evident from the role it played in the economy. Although charcoal was used even in prehistoric times—there is evidence of its use in the Jomon period—by the Tokugawa period the quality of charcoal had improved remarkably; it had a higher carbon content, which meant that it burned longer at a higher heat.19 By the early-seventeenth century, every region in Japan was producing its own charcoal, and not only were there specialists who made it as an occupation, but ordinary villagers processed it during the winter. It is impossible to know the total output because so many people were producing it, but by the late Tokugawa period, as much as 2.38 18. For developments in the kitchen, see Ekuan Kenji, Daidokoro do£u no rekishi (Tokyo: Shibata Shoten, 1976) and Kanzaki Noritake, Daidokoro yogu wa kotaru (Tokyo: Chikuma Shobo, 1984). 19. Because Japanese houses were so much less air-tight than Western ones, there was not so much danger of asphyxiation, and when charoal was used in enclosed warmers, there was even less.

A RESOURCE-EFFICIENT CULTURE

63

million hyo (straw bags) were shipped annually to Edo.20 Some domains went into the charcoal business, others permitted some taxes to be remitted in charcoal, and it was usually taxed when sold at retail. Budgets of people in various classes reveal that charcoal was a major component in all of them, and firewood as well in most. A merchant household of ten spent twenty-four ryd on ground rent, thirty on food, and twelve on charcoal and firewood. A couple making about seven ryd spent one on these fuels. And a samurai whose real income was thirty ryd per year spent more than four on charcoal.21 People in the warmer and urban areas began to use the kamado. The kamado is a stove traditionally of stone, brick, or earth with openings at the top for the insertion of pots called kama. Since it was enclosed and directly heated only the pots, it was more fuel-efficient than the open hearth, but charcoal provided too weak a fire to cook rice, and so the kamado still required firewood. The innovation of the kamado tended to move the cooking area, a major source of fire danger, from the wooden-floored living area of the house to the earthen-floored doma. People in the warmer southwestern parts of Japan and fuel-short Kinai area around Kyoto and Osaka switched to using kamado, whereas minka with only irori were almost exclusively confined to the snowy regions of the north and the mountainous areas of central Japan.22 Not only could firewood be more readily obtained in these regions, but because of the long, cold, and snowy winters, the family needed the irori as a heat source. Since the minka that have survived are mostly houses of the well-to-do, many have both irori and kamado. The growing shortage of fuel and the innovations in stoves that accompanied it had side effects on the Japanese diet. Kamado, which usually had more than one place for a pot, could be used to cook more complicated meals than the stews that could be prepared over the irori. 20. The size of these straw bags differed by region and by period, but one hyo contained about two bushels, which is approximately nineteen gallons or seventy-two liters. 21. See Higuchi, Mokutan no bunka-shi, pp. 8-29, 89-125. To give some idea of the value of a ryd, the Bakufii made the koban, the one-ryd gold piece, the standard of value in Edo and its own lands. In 1700 it fixed the value at sixty momme of silver for daily use in Edo. By the nineteenth century, daily workers in agriculture had an annual income of about 360 momme, which would have been equal to six ryd. This income would have bought 5.5 koku of rice even if purchased in Osaka. See Hanley and Yamamura, Economic and Demographic Change, pp. 118-25. 22. Suzuki and Takahashi, "Nagashi, irori soshite kamado." Whereas 87 percent of the irori in the surveyed minka were on wooden-floored rooms, 87 percent of the kamado were found in the doma.

64

A RESOURCE-EFFICIENT CULTURE

Nineteenth-century house with kamado in the doma and living room with irori.

Rice or grain could be cooked as a separate dish from the soup and vegetables. Whereas this new cuisine was fuel-conserving, people with low incomes economized on fuel still further. Many families started to cook rice only once a day, and people who lived in one-room tenements in cities often did very little cooking at all. This gave rise to numerous

A RESOURCE-EFFICIENT C U L T U R E

65

shops specializing in take-out foods for consumption at home—the fast-food industry of the Tokugawa period. Not only were methods of cooking resource-conserving, so too did the very cuisine itself make maximum use of scarce resources. The land that could be given over to game had all but disappeared early in the Tokugawa period for everyone except the highest members of the ruling class. Buddhist strictures are often cited for the lack of animal protein in the diet in traditional Japan, but since the Chinese were also Buddhist and ate meat when possible, this argument is not very convincing. It seems more than coincidental that the consumption of meat declined as the population grew and the need to use all land for agriculture left little or none for grazing. Though meat from four-legged animals was proscribed by Buddhism, those who could afford to hunt ate wild birds, and outcast classes are known to have eaten animal flesh. This would lend credence to the argument that people gave up meat eating more because commons, woods, and open fields gave way to cropland than because Buddhism proscribed it. But a good number of Japanese did eat meat. For the most part they did not raise it, but obtained what wild birds and other animals they could, eating nearly every species they found. When the city of Edo was first built, the sudden influx of population caused a food shortage. As a result, samurai and townspeople alike ate dogs in the winter months; this was not famine food but a common practice in the region. Westerners in Japan in the late-sixteenth and seventeenth centuries noted that wild and tame animals and birds could easily be obtained in large quantities as food sources in the cities of Kyoto and Edo. 23 In these centuries and earlier there were no decrees issued by government prohibiting the eating of meat.24 Although a discussion of meat eating is not central to the argument of this chapter, that the Japanese ate meat goes against conventional wisdom that they did not and that the reasons were religious. All of the evidence substantiates the argument that the Japanese used all potential sources of food but did not sacrifice cropland, whose harvests directly fed the people, for pasture or fodder for animals to feed the few. However, in 1612, the Bakufu issued a decree banning the killing of cattle and also the sale of cattle that died naturally. This was not only

23. Harada Nobuo, Rekishi no naka no kome to niku (Tokyo: Heibonsha, 1993), P258. 24. Ibid., p. 92.

66

A RESOURCE-EFFICIENT CULTURE

a decree for Bakufu-held territories but was communicated to all of the daimyo as well. The concern of the Bakufu was not religious; rather it was to preserve draft animals in order to maintain and increase agricultural production. The need for such a decree belies the belief that Japanese didn't eat meat.25 A variety of evidence attests to the eating of meat in Japan during the Tokugawa period. The Meisan shojiki dmi, published in 1760, listed the kinds of meat sold in Edo's Kojimachi area. Included were boar, venison, fox, wolf, bear, badger, beaver, cat, and wild dog. Nor was meat eating limited to certain classes who did it clandestinely. The shogun was sent a gift of beef preserved in miso (bean paste) from the domain of Hikone (Shiga Prefecture), and the domain of Sendai (Miyagi Prefecture) had a well-known cook named Kikkawa Fusatsune who recommended making soup stock using a beef base. A number of books were published from the early-eighteenth century that gave advice on how to feed beef to the sick to improve their health and offered directions on how to prepare meat and game birds to make them "most deliciously." During the second half of the Tokugawa period, as wild game declined because of overhunting, chicken became more popular, as did their eggs. In 1875, a book called Manpo ryori himitsu bako described 103 different egg dishes, including one in which the white of the egg was cooked inside the yolk.26 Meat eating became more popular in Edo by the nineteenth century when shops selling meat proliferated. It was reported by Yamada Tose that even samurai bought meat. Most of what was sold was caught in the wild, and included boar, bear, wolf, badger, beaver, squirrel, and monkey. People who were opposed to the eating of meat decried the smell emanating from shops that sold it, and many may not have wanted to admit to the eating of meat. Euphemisms were used in talking about meat, with boar commonly called "mountain whale" and deer referred to as "maple."27 Teradako Seiken in his Edo hcmjoki, which was written about Edo's prosperity in the early 1830s, devoted one segment to the "mountain whale." Its opening sentence was "Scallions and meat—a 25. Kamo Giichi, Nihon chikusan-shi (Tokyo: Hosei Daigaku Shuppan-kyoku, 1976), pp. 189-90. 26. Matsushita Sachiko, "Edo jidai no ryori no zairyo to choriho," in Shoku to shoku kukan, ed. Nihon Seikatsu Bunka-shi Gakkai, vol. 5 of Seikatsu bunka-shi (Tokyo: Yuzankaku, 198+), pp. 4 4 - 4 6 . 27. Kamo, Nihon chikusan-shi, p. 209.

A RESOURCE-EFFICIENT CULTURE

67

perfect combination." This segment is a satirical description of meat eating in Edo. 28 The more adventuresome in the prosperous capital were not the only people eating meat. In the domain of Choshu, official records were kept in the nineteenth century of game sold in the domain, and the list included bear, wild boar, deer, rabbit, badger, monkey, pheasant, duck, dove, and other birds. Chicken and eggs were sold in seventeen villages.29 That the domain kept records is evidence that trade in game was a recognized and common activity. The value of this trade is no indication of how much game was actually consumed, since considerable amounts must have been consumed by the hunters and others and gone unrecorded. Meat eating was almost certainly more common, at least among the samurai, in western Japan than elsewhere. Japanese also raised pigs to sell to the Chinese and Koreans in Edo and Nagasaki. Pork was eaten by the Japanese in Satsuma, from which comes Satsuma-jiru, a traditional soup with meat in it, and in the 1850s it was widely known in Edo that employees of the Satsuma domain were raising pigs both to eat the meat and to sell it.30 Yet it is unlikely that the majority of Japanese were eating meat, and certainly not on a regular basis.31 Reports make it clear that selling and eating meat were not everyday activities, but rather something fairly unusual that occasioned comment. Game became ever more scarce as the population and the amount of land under cultivation increased. How much meat was eaten by whom and how often will likely remain unknown; what evidence does show is that Japanese did not ignore game as a food source and that there was such a demand that it was openly sold in many parts of the country. Nevertheless, for protein sources, most people relied on seafood and grains. In the Tokugawa period, the sea provided a never-ending source of protein, and the main limitation on its consumption was transporting it inland from the coast. Inland, the main source of protein was the soy bean, which has a much higher calorie output per acre than animal 28. Andrew Markus, "Meat and Potatoes: Two Selections from Edo HanjokiSinoJapanese Studies 4 , no. 2 (April 1992): 7 - 2 6 . The translation is Markus's on p. 12. 29. Goto Yoshiko, "Bocho fudo chushin-an kara mita Tempo-ki Choshu-han ni okeru omo na shokuyo jucho-rui ni tsuite," Yamaguchi Daigaku Kyoiku Gakubu kenkyii ronso 38, n o . 2 (1988): 1 9 - 2 9 . 30. Kamo,

Nihon chikman-shi,

31. Ibid., p. 195-

pp. 194, 210.

68

A RESOURCE-EFFICIENT

CULTURE

flesh, an inefficient source of protein in terms of the amount of land and grain needed to produce it. Thus tofu and other soybean products became a major source of both protein and calcium during the Tokugawa period. Compared to Europe, far greater use was made of wild plants. Ferns, burdock roots, and in time of famine various wild tubers and even bark were consumed. The sea provided not only protein sources, but also greens, as seaweed was dried and used widely in a variety of ways—from adding it to soup, brewing it as tea, and using it as medicine. But the staple of the diet was rice and other grains. Rice was the preferred staple, but not enough rice was produced to enable everyone to eat it as the staple food. Probably most people in cities were eating rice and most farmers were eating it mixed with other grains.32 People preferred rice because they liked its taste, but it could be argued that they came to prefer it, like many other peoples in Asia, because it was suited to a densely populated country. Rice has a higher yield to seed ratio than the staple cereals of northern Europe. It has been estimated that up to the seventeenth or eighteenth century in Europe, 20-25 percent of the yield had to be held over to use as seed the following year.33 Though this is only an estimate, and the estimate of the amount of rice crop that would have to be saved for seed varies widely with the conditions of agriculture, estimates for rice crops produced by premodern techniques are 1 - 2 percent. Rice produces very high yields, and efforts to increase the yield make the land more productive though they are labor-intensive. Thus they maximize use of land and allow population to increase, or where it is already dense, maintain it.34 The economy of resources can be found not only in housing and cuisine, but also in clothing. By the seventeenth century, the basic garment for formal and casual wear of all classes was the kosode, which fits the description of what Westerners envision when the word kimono is used. The kosode was so widely adopted by the eighteenth century that people were calling it kimono, which means "clothing." Originally an undergarment, it became the article of clothing worn immediately un32. Kito Hiroshi, "Edo jidai no kome shoku," Rekishi koron, no. 89 (April 1983); Shunsaku Nishikawa, "Grain Consumption: The Case of Choshu," in Japan in Transition, ed. Marius B. Jansen and Gilbert Rozman (Princeton: Princeton University Press, 1986). 33. B. H. Slicher van Bath, The Agrarian History of Western Europe A.D. 500-1850 (London: Edward Arnold, 1963), p. 382. 34. Francesca Bray, The Rice Economies (Oxford: Basil Blackwell, 1986), pp. 3, 15.

A RESOURCE-EFFICIENT CULTURE

69

W o m e n w e a r i n g kosode: l e f t , a m e r c h a n t ' s w i f e i n silk; r i g h t , a l a b o r e r ' s w i f e i n c o t t o n .

der outerwear such as raingear or formal outer garments worn for government ceremonies.35 Kimono are made from one long, rectangular length of cloth cut into eight pieces; the pattern is the same for everyone, male or female, child or adult. Every inch of cloth is used with no waste. No matter how efficiently an article of Western clothing is cut, in fitted clothes there will be pieces from which nothing can be made or at least that are waste from a particular garment. One exception might be women's skirts, but they require so much more material than a kimono that they are scarcely as economical. There are no buttons or hooks or any other gadgets needed to hold the kimono on. The obi (the band or sash tied 35. For a description in English of the evolution of the kosode, see Liza Crihfield Dalby, Kimono: Fashioning Culture (New Haven: Yale University Press, 1993), especially pp. 32-57.

7o

A RESOURCE-EFFICIENT CULTURE

around the waist) is sufficient. This means that when the garment is put on, it can be adjusted for how fat or thin the wearer is as well as for how tall. In contrast, Western clothes are made to fit one person. If that person grows or gains weight, the garment no longer fits, and likely as not, no one else will be the perfect fit for it. Often even hand-me-downs have to be resewn, and in a used garment such reworking will likely show. Not only did Japanese clothing use all of a length of cloth without waste, it used a lot less cloth, particularly clothing for women. The typical skirt worn by American women grew in circumference during the course of the nineteenth century, from 126 inches in the 1840s, to 170 in the 1850s, and to over 190 inches by the 1860s. Fashionable dresses could be over 250 inches around, requiring seven to eight yards in length for the skirt alone. A simple skirt for everyday wear consisted of six to eight yards of material gathered onto a waistband, and though easy to sew, if the women also had to spin and weave all of this material, making a dress consumed three to four weeks. By the 1870s, magazines such as Harper's were including foldout supplements of patterns with instructions, but these were extremely complex to follow and many a sad mistake must have resulted. Even the simple wrapper, worn around the house, as a maternity dress, or to receive guests, required ten yards of material twenty-eight inches in width.36 In contrast, the traditional kosode was made from a fixed length of cloth known as ittan or one tan, which was approximately fourteen inches in width and ten to twelve yards long. The current tan for cloth is 1,150 centimeters in length, or just over twelve and one-half yards, but the Tokugawa length was usually a bit shorter, varying by region.37 However, within each region the amount of cloth in a tan was fixed so that people could conveniently buy the amount of material they needed for one kosode. Thus, any kosode could be made for roughly half the amount of material needed for the simplest garment worn by an American woman in the nineteenth century. This saved not only resources but also precious time. The Japanese way of dressing was much more economical in a num36. Betty J. Mills, Calico Chronicle: Texas Women and Their Fashions, 1830-1910 (Lubbock: Texas Tech Press, 1985), pp. 19, 65, 68. 37. Any sewing class textbook contains this information. See, for example, Naitó Michiko, Shin-hifuku, rev. ed. (Tokyo: Gakushü Kenkyüsha, 1993) and Katayama Yoshiko et al., Koko hifuku, rev. ed. (Tokyo: Jikkyó Shuppan Kabushiki Kaisha, 1986).

A RESOURCE-EFFICIENT C U L T U R E

71

ber of other ways. Since the basic pattern was the same for all, women could readily learn how to cut and sew everyday garments for everyone in the family. The length of the kosode was based on the height of its wearer, but there is extra length that is folded over at the waist under the obi to make the garment the correct length. By adjusting the folds, a shorter or taller person can wear someone else's clothing without any alterations. Clothing for young children allowed room for growth, and there was neither the worry that the child wouldn't grow into it nor any taking down of hems. Japanese clothing was both time and resource-saving in other ways as well, since it was not ironed. Good silk garments were taken apart before laundering and laid flat on boards to dry, giving them a pressed look; everyday cotton items were hung on bamboo poles through the sleeves to let them dry straight. It is true that taking clothing apart for laundering and then sewing it up again is time-consuming, but in fuelshort Japan, this was more efficient than using heat for ironing, particularly since good outer garments were probably not washed any more frequently than was clothing in the West. One can debate whether Western or Japanese clothing was more time-efficient, but clearly Japanese clothing was more resource-efficient. The Japanese also invented a very useful, resource-efficient type of towel, known as the tenugui. Prior to the Tokugawa period, this was a strip of undyed linen called a tenogoi, but during the Tokugawa period cotton became the preferred fabric and gradually it became fancier, as did other items of everyday life such as underwear. Various colors, patterns, and dyes were used, and some tenugui were even tie-dyed. The width was set at approximately thirteen inches, but the length varied according to the pattern of the cloth as designs became popular, until the Bakumatsu period when the standard length was set at just over three feet.38 Since the tenugui was just a rectangular piece of cloth, it could be used for everything from a head covering or headband to a towel or a protective cover to keep dirt and flies off food. Used wet, it served as a washcloth, and when wrung out, it could be used to dry the body, particularly after a hot bath. It was small enough to tuck in practically anywhere, and thin enough to dry very quickly in the damp Japanese climate. When old, it could be used as a rag. The tenuguiwas so popular 38. Konno Nobuo,

Edo nofuro (Tokyo:

Shinchosha, 1989), pp. 69-70.

72

A RESOURCE-EFFICIENT CULTURE

that it was given as a gift on festive occasions. Though the Western bandanna was used in multiple ways, its use was limited by comparison with the tenugui. Not only did the Japanese use less material for their clothing and towels than did Westerners, they reused everything. After cotton came into widespread use, virtually all used clothing could be pressed into service for a variety of uses: diapers, rags, and other household items. Alice Mabel Bacon, who lived in Japan in the 1880s, noted that garments were "turned, dyed, and made over again and again, so long as there is a shred of the original material left to work upon."39 Although Westerners reused what they could, the fact that clothing was cut to fit a particular person meant that it had to be adapted to fit another, which might not be easily accomplished. Also, because it had plackets, narrow sleeves, button holes, fitted waists, and the like, there were a number of parts of the clothing plus a few scraps, such as hems, plackets, and seams, that could not be readily reused, as could all parts of Japanese clothing. Used clothing was also sold in Japan, just as it was in other parts of the world.40 By the mid-Tokugawa period there were many used clothing shops in Edo, and in Tomizawa-cho, there was a market for old clothes held every morning at dawn. There was such a brisk business in used clothing that people began to steal it to sell, and thus in 1724 the Bakufu tried to close this market, but the traders convinced them the trade was too important to discontinue. It was permitted thereafter but only during daylight hours. By the late-eighteenth century, wholesalers of used clothing emerged; they bought up old clothes in Edo and sold them throughout the Kanto region and to people in Tohoku. Being able to purchase used cotton garments was particularly important in the northern regions that were not able to grow cotton. A brisk market grew up in the Kansai, and the volume of the trade in used clothing was huge; by 1841, there were four major used clothing wholesalers with annual sales in excess of fifty-five thousand ryo. The largest of these was big enough to alone do twenty thousand ryo of business per year.41 With the kimono sold in a standard size, it was easy to market used clothing, 39. Alice Mabel Bacon, Japanese Girls and Women, rev. ed. (Boston: Houghton, Mifflin and Company, 1902), p. 18. 4.0. See Donald Woodward, '"Swords into Ploughshares': Recycling in Pre-Industrial England," The Economic History Review, 2d ser., 38, no. 2 (1985): 175-91. +1. Ito Yoshiichi, Edo no Tumenoshima (Tokyo: Yoshikawa Kobunkan, 1982), pp. 2628.

A RESOURCE-EFFICIENT CULTURE

73

and a booming business it was. Clothing was also sold in the West, and the garments of the well-to-do were so valuable that there are stories of robbers stealing the clothes off children, even in Victorian times. The point here is that the Japanese were no less resourceful in recycling clothing than people were in other countries. Like clothing, footgear was standardized in Japan. The basic footwear was a sandal with thongs between the big toe and the second toe to hold it on. The poor and travelers w o r c g e t a (wooden clogs) or waraji (straw sandals), which were not only readily and rapidly produced from materials around the farm, but almost anyone could wear anyone else's. Thus, they could be produced without the regard for size and fit necessary for Western boots. Straw footgear wouldn't last long, but it was cheap, and one could readily buy or make more. Such footwear didn't provide the protection that boots did, but then, one didn't get bunions or corns from a bad fit.42 Since clothing was expensive, protective aprons became widespread in Japan, just as they were in the West. Often the working men wore little but a loincloth, and women worked stripped to the waist in summer. In the same centuries in the West, people covered up the entire body, particularly women, and being caught naked was something terribly shameful. Nakedness was not particularly shameful in Japan, much less so than in other countries in East Asia, such as Indonesia and Korea.43 Japan's hot, humid summer obviously played a role in how much clothing was needed, but the net effect was that the Japanese could get by with fewer clothes, and since what they had was so highly standardized, maximum use could be made of it. Even for the rich who wore elaborately woven and dyed materials, the standardization meant minimum waste. And the fact that the kosode was made to fit anyone meant that it was interchangeable among people and could be borrowed, handed down, or passed on after a person's death without any alterations. Whereas European clothing had little fit until the Middle Ages, Japanese clothing retained this feature until people began to wear Western dress.44 42. See Nagakura Saburo, Hida no mmgu (Takayama: Takayama-shi, 1981) for examples of the various kinds of footgear, rainwear, and other daily items that were made from straw and other renewable farm materials. 43. Nomura Masaichi, "Remodelling the Japanese Body," in Culture Embodied, ed. Michael Moerman and Masaichi Nomura, no. 27 of the Senri Ethnological Studies ( Osaka: National Museum of Ethnology, 1990), pp. 260-61. 4 4 . Nomura, "Remodelling the Japanese Body," p. 267.

74

A RESOURCE-EFFICIENT C U L T U R E

It cannot be argued that only the Japanese were frugal; certainly all premodern peoples living in societies in which goods were expensive were frugal, and many or most still are even in industrialized countries. In The American Frugal Housewife, written circa 1828, the author gives instructions on how to renew feathers and hair stuffed in bedding, how to use green tea to restore rusty silk, and how to use up hardened bread crusts, and exhorts people not to throw anything away "so long as it is possible to make any use of it, however trifling." The author, Mrs. Child, insists that all bits of string be saved, rags not be thrown out just because they were dirty, and soap suds be used to nurture young plants. Her examples might come as easily from the Japanese traditions of wisdom of frugality reflected in Saikaku's stories as those seen in Ben Franklin's.45 Both Japanese and American women in the Meiji years noted the efficiency of time of the Japanese lifestyle. Alice Mabel Bacon commented that "in a house with no furniture, no carpets, no bricabrac, no mirrors, pictures frames, or glasses to be cared for, no stoves or furnaces, no windows to wash, a large part of the cooking done outside, and no latest styles to be imitated in clothing, the amount of work to be done by women is considerably diminished."46 Tsuda Umeko, who shared a house with Bacon in Tokyo in 1888 and 1889, was of the same opinion: "A Japanese house has truly so much less to be done than a foreign one."47 Bacon did admit that there was nevertheless work to be done in the Japanese house since bedding and mosquito netting had to be taken out at night and stored in the daytime, and there was also cleaning that had to be done on a daily basis. Whereas examples of frugality and the wise use of resources can be found in every society, it is possible to find numerous examples when the Japanese—and other Asians—were thriftier or made better use of scarce resources than people in the West. One example is the Asian custom of making teacups without handles. It may seem uncomfortable and awkward to someone not used to them, but it cannot be denied 4 5. The American Frugal Housewife was written by Lydia Marie Child in her first year of marriage, which was in 1827. Published by Carter, Hendee, and Co. of Boston, it proved so popular that by 1832 it was in its twelfth edition, of which a facsimile was printed in 1980. For the examples, see pp. 1, 8,12, 13, 15, and 19. 46. Bacon, Japanese Girls and Women, pp. 17-18. 47. A letter dated February 11, 1883, in Yoshiko Furuki et al., eds., The Attic Letters: Ume Tsuda's Correspondence to Her American Mother (New York: Weatherhill, 1991), p. 40.

A RESOURCE-EFFICIENT C U L T U R E

75

that it is quicker and more efficient to make a round cup without bothering to attach a handle, which takes time, can be easily broken, and uses more resources. Similarly, chopsticks were easily made and did not require the use of precious metals. 48 Since food was served cut up so that it could be eaten with chopsticks, it wasn't necessary for everyone, or nearly everyone, to have a knife at table. One in the kitchen would suffice. It can even be argued that Japanese underwear was more resource-efficient than Western. Men wore loincloths made from a long strip of cloth, and women wrapped a simple garment called the koshimaki around them, once they started wearing undergarments in the Tokugawa period. It could be argued that Japanese followed these customs because their standard of living was lower than that in the West, but this would then be evidence for the Japanese following customs that gave them a similar or higher level of physical well-being on a lower standard of living. In summary, during the Tokugawa period when Japan's economy reached new heights of prosperity, giving rise for the first time to a true commoner culture, population increase added to an already high base population. This led the Japanese in the seventeenth century to follow frugal, resource-conserving trends found in the medieval period, rather than create a new culture based on plenty as occurred in Europe. Though the Japanese ate wild game, they did not rely on meat as their main source of protein, and so when they began to run out of game, they substituted sources that made more efficient use of land rather than devote it to pasture or fodder crops. The solution was resourceefficient because soybean products combined with grain—particularly rice—is a perfectly satisfactory way to obtain needed protein. One could argue that the Japanese lowered their standard of living because they made this substitution out of economic necessity, but this does not mean that they lowered their level of physical well-being. The Tokugawa solutions to limited resources enabled the Japanese to reach a high level of civilization using a minimum of resources, and wherever possible, natural, renewable materials. These solutions led to a society in which beauty and luxury were found in good design rather than in a vast number of objects on display. The best, rather than abundance, was highlighted. Even the rich followed these principles and kept 48. Chopsticks were of wood but made to last; even those without lacquer could be used for years. These were not the disposable chopsticks found in contemporary Japanese restaurants.

76

A RESOURCE-EFFICIENT C U L T U R E

the large quantities they owned in numerous storehouses, bringing out only a few objects at a time to be appreciated. The material culture of the Tokugawa period can trace its roots to medieval times, if not earlier. But what should be noted is that the direction the material culture took was not the ornateness of the Toshogu shrines or the Kano school of painting, but the simplicity of the shoin style and ink drawings. This is not to argue that the rulers and the wealthiest of merchants were not guilty of waste, but the general principles of using less energy, fewer resources, and reusing and recycling let many more people participate in a higher standard of living and culture than if the trend had been toward the wasteful use of scarce resources. By defining luxury in terms of austerity and spareness, many more people were able to have a high level of physical well-being.

CHAPTER

FOUR

A Healthful Lifestyle

Whereas concepts such as luxury, comfort, and quality of life are subjective and vary from culture to culture, it is possible to more objectively assess elements of material culture as they affect human beings. How material culture influences health is a factor that transcends culture to a great extent, though not completely. This chapter examines how the kind and quality of food, clothing, sleeping arrangements, and bathing affected the health of people in the Tokugawa period and how these compared with those in the West.

Food The aspect of the material culture that is easiest to measure is food, but for that very reason, it is also the most controversial aspect, given the paucity of data. With food, one can measure both the quantity and quality—calories and nutritional value. However, it is difficult enough to know who consumes what today and close to impossible to assess the diets of people who lived a century or more ago. To add to the challenge, people don't often leave written records about the ordinary activities of daily life, and certainly what is eaten falls into this category. Diaries that are very explicit about other aspects of life—particularly social—are nearly silent on matters of food. Furthermore, people are notoriously unreliable in reporting what they had to eat. Oral 77

78

A HEALTHFUL LIFESTYLE

histories reveal that people cannot remember what they regularly ate, and if one constructs a pattern based on interviews, it in no way tallies with gross figures on crops raised or foods sold. People can remember what they ate on a social occasion, but not what they had for their evening meal three days ago. Assessing the Tokugawa diet is made even more complex because of the ideological ramifications for historians of how much rice people consumed. Although rice was considered the preferred staple, scholars frequently point out two pieces of evidence to show that many or most Japanese were too poor to be able to eat it and had to make do with grains considered inferior. One oft-cited document is the famous edict of 1649 in which the Bakufu exhorted cultivators not to give rice to their families at harvest time but to save it for the future. Instead they were to eat vegetables, millet, and other coarse grains. Nor were they to buy sake, a wine made from rice.1 The second document, a report written in the Kyoho period (1716-1736), states that farmers living in the flatlands where rice was grown ate rice regularly in the form of zosui (a porridge), but those in the mountainous regions who had to purchase it could afford to eat it only on the first three days of the New Year.2 However, as Kito Hiroshi has pointed out,3 someone had to be eating all the rice that was produced, the output of which grew faster than the population during the eighteenth and nineteenth centuries.4 And if people were not eating rice, why was the government exhortation necessary? Clearly, rice was widely consumed during the Tokugawa period. It was the major crop in the early seventeenth century, so that it could conveniently be used by the government as the unit of measure for the size of domains, samurai stipends, and tax rates. More than two and a half centuries later, in 1874, rice constituted 63 percent of all farm products.5 One of the few estimates that exist on output and food consumption for any part of the Tokugawa period is for 1840 in Choshu. Based on output and population, rice constituted an estimated 53 percent of 1. Keian nofuregaki, described in Watanabe Minoru, Nihon shoku seikatsu shi (Tokyo: Yoshikawa Kobunkan, 1964), p. 2 4 4 . 2. Tanaka Kyugu in Minkcrn seiyd, cited in Kito Hiroshi, "Edo jidai no kome shoku," Rekishi koron, no. 89 (April 1983): 43. Tanaka, first a local and then a Bakufu administrator, published Minkan seiyd in 1721. An astute observer knowledgeable of conditions in the Kanto region, he is widely cited because of his insight and detail. 3. Kito, "Edo jidai no kome shoku," p. 43. 4. Kikuchi Toshio, Shinden kaihatsu (Tokyo: Shibundo, 1964), pp. 223-25. 5. "Meiji 7-nen fuken bussan-hyo," cited in Kito, "Edo jidai no kome shoku," p. 4 4 .

A H E A L T H F U L LIFESTYLE

79

the grains consumed in that year.6 But, how much rice the Japanese ate at any time during the Tokugawa period depended on social class, income, the region in which they lived, and harvest conditions. Rice was the preferred staple grain of all people who could afford it, regardless of class. One word for a meal in Japanese is the same as cooked rice {gohcm), and everything else is considered a side dish. Although bread has long been at the heart of the Western diet, so that people talk of "breaking bread together," it did not occupy the center of the culinary traditions the wealthy developed as rice did in Japan. For Asians, rice has a variety of advantages over other grains in terms of preparation and taste. It retains its shape when cooked in a large quantity of water, and it combines well with other foods. Thus, families in Europe whose main meal was a staple grain usually had porridge, and often porridge alone. Rarely was it combined with another food except milk. In contrast, in Japan, one-pot meals had one or more grains— usually including rice—as their base, and other foods, particularly vegetables, were added appropriately during the cooking process. As the economy grew, income rose, and just as farmers ate more rice over time, so did they eat better milled and more highly polished rice. In the seventeenth century, people ate rice that was partially polished; it appeared whitish, but part of the bran remained. Late in the century, a process was developed that completely removed the hull but left the bran. White, or highly polished, rice was considered the highest quality and tended to be eaten in cities. But those who could afford white rice often developed a vitamin B deficiency, and thus beriberi became known as the "Edo affliction." People who became ill while working as servants in Edo found that when they went back to the country, they improved, but it was not until the early-twentieth century that the cause of this disease was discovered.7 It would be unusual to find any premodern society that depended on a single crop for its staple. Not only would it make poor use of human and natural resources, but it would be dangerous because crop 6. Nishikawa Shunsaku, "Grain Consumption: The Case of Choshu," in Japan in Transition, ed. Marius B. Jansen and Gilbert Rozman (Princeton: Princeton University Press, 1986), table 4. 7. In the Meiji period, the Japanese navy reduced the number of cases of beriberi on long voyages by decreasing the amount of rice in the diet and increasing supplementary foods. This was prior to learning the precise cause of the disease, although they knew that it was a result of the diet and prevalent only in rice-eating cultures. See Louis Livingston Seaman, The Real Triumph of Japan: The Conquest of the Silent Foe (London: Sidney Appleton, 1906), especially chapter 13.

8o

A HEALTHFUL LIFESTYLE

failure would cause widespread starvation. M u c h of Europe was affected by potato blight in the nineteenth century, but only in Ireland did it cause widespread starvation because so many people relied almost entirely on the potato for their nourishment. Like most peoples, Tokugawa Japanese relied on a number of staple foods, o f which rice was the preferred, with barley and wheat next in preference. People still consumed the oldest cultivated grains in Japan, awct and kibi, both of which are kinds of millet, and hie, a grass that is no longer considered particularly desirable as a food. By the Tokugawa period soba (buckwheat) and morokoshi (great millet) were also eaten. 8 A n d people continued to eat nuts, roots, and various tubers that have been part o f the diet from prehistoric times. One of the best pieces of evidence as to what people were eating is a survey that a shogunal doctor, Niwa Seihaku, undertook in 1734 on a nationwide basis. The purpose was to find out what the taxpayers were eating. It was found that in Shinano (Nagano Prefecture), as many as twenty different kinds o f rice were planted to ensure the maximum yield depending on the growing conditions, and in addition, people grew a variety of grains, including buckwheat, and beans in upland fields. T o supplement their diet, they grew sesame, poppy seeds, a variety o f daikon, and other foods. The survey revealed that people also ate things we don't usually think of as foods, including grasshoppers, horse chestnuts, and wild plants. Evidence for Shinano indicates that the diet changed by the season, depending on how much work had to be done. More rice was eaten during times o f heavy agricultural work. However, based on the limited village samples in this study, there doesn't seem to have been much difference in Shinano in the diet based on status within the household—whether master or servant. 9 A number o f new foods were introduced into Japan in the sixteenth century and throughout the Tokugawa period, and foods that had been introduced in earlier centuries were more widely diffused. For the first time manuals were published and widely disseminated telling farmers how to best grow various crops. Miyazaki Antei's Nogyd zensbo (1697) gave instructions for planting carrots and other new food crops. These new foods, most of which originated in the Americas, included pota-

8. See the glossary for fuller definitions of these grains. 9. Tsukamoto Manabu, "Ikiru tame no chie," in Mum no seikatsu bunka, ed. Tsukamoto Manabu, vol. 8 of Nihon no ktnsei (Tokyo: Chtiokoronsha, 1992), pp. 288-92.

A HEALTHFUL LIFESTYLE

81

toes, green beans, corn, red peppers, pumpkins, watermelon, spinach, and very late in the Tokugawa period, peanuts.10 For the Japanese, the most important of the new foods was the potato, which arrived in both Asia and Europe from South America in the sixteenth century.11 The sweet potato is thought to have been introduced to Japan in 1605 and the white potato at about the same time. The reaction of the Japanese was the same as the Europeans': to grow potatoes for the pretty flowers and then to use the tubers for horse fodder. But by the second quarter of the seventeenth century, people in western Japan had begun to eat them. Although originally called bareisbo or "horse fodder" by the Japanese, white potatoes became valued as a versatile food. It is the sweet potato that is credited with reducing the deathrate from famine in Japan. In 1732, locusts caused a major crop failure in Kyushu, but the deathrate was low in both Satsuma and Nagasaki because people were not relying entirely upon grain; now they had sweet potatoes to fall back on. Sweet potatoes could be grown upland, in contrast to rice, and they produced more calories per acre than almost any other crop. The protection against famine afforded by this new crop came to the attention of various people around Japan. A magistrate from Bitchu (Okayama Prefecture) tried so hard to popularize them that he was given the nickname of ktrnsho daikan (sweet potato magistrate), and in 1734 seeds from the Shimazu house were experimentally planted in Kozuke (Gumma Prefecture) and Shimotsuke (Tochigi Prefecture). The Ka-nsho-ki (Notes on sweet potatoes), with directions for planting them, was so popular that it was reprinted in the eighteenth century and again during the nineteenth. Sweet potatoes may well be an important factor not only in the maintenance of a dense population in the eighteenth and nineteenth centuries, but also in explaining why the population in western Japan grew faster than that of the rest of the country, particularly in the four domains most instrumental in overthrowing the Bakufu—Satsuma, Choshu, Tosa, and Hizen. What people ate and how they prepared it depended to a large extent 10. Watanabe, Nihon shoku seikatsu-shi, p. 191. See Thomas C. Smith, The Agrarian Origins of Modern Japan (Stanford: Stanford University Press, 1959), pp. 8 8 - 9 2 for a description of the treatise Nqgyo zensho and its influence on Tokugawa agriculture. n . Information on the introduction of potatoes in Japan comes from Watanabe, Nihon shoku seikatsu-shi, pp. 2 4 6 - 4 9 , and Adachi Iwao, Nihon shokumotsu bunka no kigen (Tokyo: Jiyu Kokuminsha, 1981), pp. 257-59-

82

A HEALTHFUL LIFESTYLE

on the utensils and technology available.12 These varied by region and changed over the course of the Tokugawa period, and the changes brought about a transformation in the diet. Two basic methods of cooking were traditionally used: cooking in a pot over the irori and cooking on the kamado. In families depending on the irori for cooking, one-pot dishes were popular because various ingredients could be added at appropriate times and left to cook in the pot. This was very similar to the one-pot method used in England, in which porridge was made of oatmeal or bags of food were lowered into boiling water and then eaten separately. Families who could afford kamado could cook more complicated meals, including the rice or grain as a separate dish from the soup and vegetables. By the Tokugawa period, the kamado was widely used in towns and cities, particularly in the Kansai region where it was difficult and expensive to obtain fuel, and irori predominated in the cold regions of the north. 13 The present method of steaming polished rice is a relatively new technique for preparing this staple and is connected to the use of the kamado. The equipment and method gradually developed from the midTokugawa period but were perfected and diffused only a century ago. Originally two methods were used. One was the same as the present method in which exactly the right amount of water is used from the start and the rice is steamed until the liquid is completely absorbed. The second was to start with more water than was needed. The excess was removed during the cooking process, and then the rice was left to steam. The first method is the more difficult to use because the temperature must be gauged precisely, the rice must be cooked slowly at first, and after the midpoint, the heat must be raised but the top not opened until the cooking is completed and the rice sits for some time. The rice can be easily burned and ruined, and thus considerable time and skill is needed to prepare rice in this method, in contrast to boiling grain in a pot on the irori. Clearly only the elite and well-to-do had the resources, time, and skill to prepare rice using this method, and the growth of its popularity over time attests to a rise in the standard of living and almost certainly a rise in the level of nutrition. The development of Japanese cuisine that accompanied these changes in rice preparation methods occurred from the mid-Tokugawa period 12. For good descriptions of cooking utensils and methods, see Ekuan Kenji, Daidokoro dogu no rekishi (Tokyo: Shibata Shoten, 1976). 13. Ibid., pp. 71-72.

A HEALTHFUL LIFESTYLE

83

on. Whereas Kyoto and Osaka were known for their cuisine, food in Edo was not considered tasty, though it gradually improved over time, particularly after various seasonings and condiments such as soy sauce became widely used. By the late seventeenth century, ordinary people began to eat a third regular meal at about one o'clock called chujiki, which means "to eat in the middle." As cities grew, the first restaurants began to appear. Gradually a variety of shops specializing in one food or another sprang up, from tempura shops, noodle shops, unagiya (specializing in eels), to sweet shops and carts that sold simple foods and drink in the evening. The chaya or tea houses developed into restaurants, but even the poor ate out at noodle shops. Clearly people had more money to spend on food, and even people living in cramped, rented rooms who couldn't really cook could afford sufficient and tasty meals.14 By the nineteenth century, so many people ate out, regardless of class, that the government tried to limit it—unsuccessfully. By midcentury, there were more sushi and soba shops than any other kind of eating places in Edo. 15 In i860, a high-ranking samurai from Kii (Wakayama Prefecture) wrote that in every corner of Edo there were food shops; soba shops, those selling food by the dish, and those selling shiruko mochi (rice cakes in a sweet bean soup) were all over. The numerous travelers to Edo not only needed places to eat, but also brought with them knowledge of foods from other parts of the country and took home what they learned in Edo. Everyone writing on changes in eating out in Japan noted both the new level of wealth that permitted it and the beneficial results on the diet, such as the introduction and diffusion of new foods. 16 The publication of cookbooks is an indicator of the levels of wellbeing and wealth reached by many in the Tokugawa period. People were no longer concerned with getting sufficient food, but rather a large number of people were able to obtain numerous ingredients and wanted to know how to prepare them. Outside of China, Japan was the only country in Asia in this period in which cookbooks were published, the first of which can be considered the publication of Ryori monogatari 14. Ibid., p. 164. 15. Sushi is small pieces of uncooked fish served on bite:size balls of rice. The soba shops sold buckwheat noodles. 16. Mitamura Engyo, Edo no ishokujii (Tokyo: Seiabo, 1957), pp. 10-93: Watanabe, Nihon shoku seikatsu-shi, 2 1 4 - 2 4 .

84

A HEALTHFUL LIFESTYLE

(The story of cooking) in 1643.17 There were well over a hundred cookbooks in circulation by the late-Tokugawa period, and there is no way of knowing how many variations of these were in existence. The number of cookbooks published in each decade reflected the general state of the economy, with the peak in the early-nineteenth century, the BunkaBunsei period of prosperity.18 In the eighteenth century there was a shift in the focus of cookbooks from the practical to pure pleasure in reading about food. 19 Chefs for the rich experimented with exotic new spices such as cinnamon, which began to be introduced in the earlynineteenth century. Clearly many people had reached a level of income at which they could afford a varied diet and wished to experiment with food. The backbone of Japanese cuisine remained the tradition inherited from the medieval period, and it was on this tradition that Tokugawa developments were based.20 During the Muromachi period, the style of cuisine disseminated from the Zen temples became such a fundamental way of eating that this is considered the period in which Japanese cuisine developed. The temples introduced foods such as manjü, yokan, and tofu (steamed buns, a jelly made of bean paste, and bean curd, respectively). Tea gradually became a national drink,21 and with the development of markets and the salt industry, fish was preserved and sold throughout western Japan, especially in the major cities and towns. Japanese began to eat a larger variety of vegetables, and the areas surrounding cities in which vegetables were grown increased. Part of this demand was caused by the popularity of shojin rydri, the vegetarian cuisine popularized by the temples. In the Muromachi period, seasonings that are an indispensable part of traditional Japanese cuisine were developed, though they were very different then from their modern versions. Soy sauce (shoyu) was not a seasoning found in the mid-Muromachi period, but records of the midto late-sixteenth century indicate that it was in wide use by then. We 17. Hirata Marie, "Nihon ryóri no kokoro," in Edo rydri hyakusen, ed. Fukuoka Hiroshi and Shimazaki Tomiko (Tokyo: 2001-nensha, 1983), p. 224. 18. Harada Nobuo, "Edo kóki no ryóri hon," in Fukuoka and Shimazaki, Edo ryori hyakusen, p. 238. 19. Ibid., p. 232. 20. See Watanabe, Nihon shoku seikatsu-shi, for an overview. Watanabe characterizes the Kamakura and Muromachi periods as the time Japanese cuisine developed and the seventeenth and eighteenth centuries as the period in which it was perfected. 21. Tea was known in the Nara period and then was abandoned as a drink. People began drinking it again in the Kamakura period. Watanabe, Nihon shoku seikatsu-shi, p.126.

A HEALTHFUL LIFESTYLE

85

can only surmise how it was made, but since shoyu means "oil of hishio," which was a kind of miso, it started out as the drippings from miso when it was made. True soy sauce gradually developed as the methods of producing miso were improved. Other seasonings created from miso began to be used and sold on the market, and gradually miso itself came to be used directly as a seasoning. Sake, sugar, and vinegar were increasingly used in cooking, and seasonings such as wasabi (a kind of Japanese horseradish), mustard, ginger, and pepper came to be widely used. Of all of these, only pepper was imported; the Europeans brought it to Japan in the sixteenth century.22 By the late-Muromachi years, all of the major elements of what can be considered traditional Japanese cuisine were present, from the staple foods to seasonings and combinations of how foods were served. Changes that occurred in the Tokugawa period were primarily refinements in production and changes in fashion. Of course, as incomes rose during the Tokugawa period, many more people were able to partake regularly of this cuisine. And what is so important about the widespread use of miso and soy sauce is that, eaten together with rice, they formed a complete protein, although a diet high in miso and soy sauce has a high salt content. But people ate soy beans not just in the forms of miso and soy sauce, but also as tofu. Thus, at the same time that the Japanese became less able to rely on protein in the form of wild game,23 they found another source of protein that could be diffused more widely through the population than animal protein ever was. Because of regional differences in the diet, income differences, and lack of information, it is impossible to hypothesize what the typical diet was for the Japanese as a people during the Tokugawa period. Presumably many continued to eat the same foods, prepared in the same ways as in earlier centuries. For example, in the fifteenth century the people in the very far north of Honshu had as their staple hiegayu, a porridge known to have been eaten in the Tokugawa period. People who lived in the lowlands where rice could be easily grown ate it in porridge form; a porridge would be easy to cook in a pot hanging over the fire, and greens and other seasonal foods could be added. Rice was often the "glue" that held together the coarse grains and was added to create the desired consistency. Seasonal greens were often added to make a vegetable and grain stew. This type of dish might be eaten only once a day; 22. Ibid., pp. 149-50. 23. See chapter 3 for a discussion of meat eating.

86

A HEALTHFUL LIFESTYLE

at other meals there would be gruel, according to a description of the diet in Kawasaki along the Tokaidd. In Musashino, on the outskirts of Edo, the diet from the mid-Tokugawa period on was said to consist mostly of coarse grains, usually a mixture of three parts awa and seven parts barley.24 Another important food was an unleavened bread. Suzuki Makiyuki, who traveled in the Nagano area, wrote in 1827 that the staple, both morning and evening, was yakimochi made of hie, and up until postwar Japan the Nagano-Yamanashi regions were noted for serving yakimochi on festive occasions and to guests. Yakimochi was made of ground grain and was baked slowly for a long time in the hearth over a low heat. It is in the tradition of unleavened breads eaten as staple food through south and central Asia, and the tradition is still found today in such foods as senbei and roasted rice cakes eaten at the New Year. Another Tokugawa staple was suiton, a round dumpling similar to matzo balls, which was cooked in hot water. Suiton could be made of almost anything: buckwheat, maize, potatoes, or wheat.25 What we think of as the traditional Japanese diet of steamed rice accompanied by soup, one or more side dishes, and pickles, developed slowly, being dependent on the ability to obtain polished rice and the cooking equipment and skill to steam it. Even samurai families often had a daily diet of coarse grains or rice mixed with other grains with a side dish of fish or something extra for the master, but only soup, pickles, and possibly boiled vegetables for the rest of the family and the servants. Kagawa Shin'ichi, a low-level samurai from Okayama who kept a diary from the 1830s through the 1860s, wrote that when he was young his family typically ate barley mixed with rice, a vegetable or tofu, and miso soup. He had fish at most only four to five times per month. When he went to his grandmother's, he had¿¡ohan that had no other grain mixed with it; he remembered his grandmother's rice in later years because it was so delicious.26 Another samurai wrote in the early 1860s that his daily meals con24. Tanaka Kyugu, Minkan seiyo, cited by Kimura Motoi, "Nomin seikatsu no shoso, in Seikatsu-shi, vol. 2, ed. Morimatsu Yoshiaki, Hogetsu Keigo, and Kimura Motoi, which is vol. r6 of Taikei Nihon-shi sosho (Tokyo: Yamakawa Shuppansha, 1969), pp. 201-2. 25. Otsuka Tsutomu, "Nihon-shoku no seiritsu to sono ayumi—kin-gendai no chushin ni," Seikatsu bunka-shi, no. 9 (March 1986): 19. 26. The meal he described is known as icbiju issai—that is, one soup and one vegetable as side dishes—and is now known as the "traditional" Japanese family meal. Taniguchi Sumio, "Bakumatsu ni okeru Bizen-han kakyu heishi no seikatsu," Kibi chiho-shi 13 (December 1954).

A HEALTHFUL LIFESTYLE

87

sisted of soup, pickles, and chazuke (hot tea poured over rice), and he considered that he had eaten well when he also had tofu, a cooked vegetable, an egg, or some low-quality fish with his meal.27 Clearly what most samurai liked to eat at the end of the Tokugawa period would be considered simple fare indeed today. One of the high officials who visited the U.S. following the signing of the treaty with Perry left a record of his trip. He wrote on the return home that thirty days after leaving New York all of the miso and soy sauce had been used up and they subsisted on dried radishes (kiriboshi daikori) with flaked dried bonito for seasoning. Whenever the men gathered, they talked of food; what they most wanted to eat when they got home was a meal with miso soup and pickles. Otsuka Tsutomu argues that this meal did not become standard in Japan until the twentieth century.28 On the other hand, there are numerous accounts of meals at inns, feasts on special occasions, and the delicacies given to the elite and wealthy that lead to the conclusion that the diet for some was at the gourmet level at least occasionally. In Yonezawa in the mid-Tokugawa period, a group of men who formed the governing body of the village held a meeting after the fall harvest at which they ate the following foods: salted salmon, tuna, bean curd, dried bonito, squid, herring roe, and dried herring—all purchased in a nearby town—eggs, dried nameko (an edible fungus), sea bream, fried bean curd, ayu (sweetfish), horseradish, and the list goes on.29 Many of the items were certainly not part of the daily diet, and not in this combination, but they were all available, and farmers had the income to purchase them for special occasions. Sugar was a luxury item and purchased only in small quantities, but it is significant that even people in the northern, poorer sections of the country could buy it and did by the mid- to late-Tokugawa period.30 One lower official in the Bakufu, who was something of a gourmand, kept a travel diary in 1856 that listed the menus of the inns he stayed in while making an official tour to the north of Edo. 31 Meals given this

27. Watanabe Zenjiro, Kyodai toshi Edoga washoku 0 tsukutta (Tokyo: Noson Gyoson Bunka Kyokai, 1988), p. 85. 28. Otsuka, Nihon-shoku no seiritsu to sono ayumi, pp. 20-21. 29. Kimura, "Nomin seikatsu no shoso," p. 204. 30. Mori Kahei, Nikon hekichi no shiteki kenkyii, vol. 1 (Tokyo: Hosei University Press, 1969), pp. 524, 536-40, 572. 31. Hayami Akira, "Bakumatsu-ki 'Kimi nikki' ni miru tabiyado no shokuji," Rekishi koron, no. 73 (December 1981): 80-87.

88

A HEALTHFUL LIFESTYLE

official were similar to those provided at traditional inns today. If the amounts were also similar, the diet of the travelers would have been adequate, with the possible exception of vitamin A. But in the home people also ate sweet potatoes, pumpkin, and squash, plus numerous greens, which would have helped balance their diets. With few exceptions, food doesn't usually figure prominently, if at all, in the various diaries extant from the Tokugawa period. The usual topics are work and social relations, and food is mentioned only when it involves an occasion out of the ordinary or social relationships. When mention of food does occur, the ingredients and dishes sound familiar to Japanese today. A commoner named Umeda, who served as a lowerlevel secretary in agricultural administration in Kanazawa, kept a diary from 1864 to 1868. There was little mention of food, and then such references were to being invited out. On an official trip in 1865, he described the menu by relating what was on each plate. Lunch at a chaya consisted of one tokkuri (bottle of sake), a small karei (plaice or turbot), a dish of boiled greens, rice, and tea. Supper was boiled daikon with buri (yellowtail) for flavor, miso soup with daikon, two pieces of yellowtail flavored with shiso (beefsteak plant), itokonnyaku (noodles made from devil's tongue), and two prawns. All the meals on the trip were so described, but it is impossible to discern from the diary entries how unusual this kind of meal was for this man. However, the meals were probably more varied and better than he usually had or he wouldn't have noted them in his diary.32 Omine-san, a pawnbroker's wife, whose diaries are extant for 1791 and 1825, focused on her social relationships in her diary. She had servants to do much of the housework, including most of the cooking, and she appears not to have had a great interest in food. She noted when she had chazuke (which she politely referred to as ochazuke), presumably because she made that herself, but otherwise wrote about food only when she fed guests, received a gift certificate for food items, or when food figured in her relations with others.33 The people who had the learning and leisure to keep diaries were not of the class of people who went hungry or who themselves cooked. Although diarists did not go hungry in time of crop failure, they

32. Wakabayashi Kisaburo, ed., Umeda nikki (Kanazawa: Hokkoku Shuppansha, 1970), p. 84. 33. Nichiroku, vol. 3 of Wakayama shishi, comp. Wakayama Shishi Hensan Iinkai (Wakayama: Wakayama Shishi Hensan Iinkai, 1975-1992).

A H E A L T H F U L LIFESTYLE

89

sometimes left clues as to how hungry others were. Kawai Koume was the wife of a Confucian scholar teaching in the domain school of Wakayama, and her husband had been adopted into her family. She kept a diary for nearly forty years until 1876, which meant that it spanned the years of the Tempo famine. Her diary for the year 1837 covered the details of her daily life, focusing on who visited whom and what items were given and received in the household. Her lifestyle was clearly not upset by the unusual events of 1837, though she did note the price paid when her husband's stipend rice was sold and the fluctuations of price in this year. Mention was also made of the riots in Osaka led by Oshio Heihachiro but only in the form of what was heard in discussions with visitors. She noted that beggars were dying and that many residents in her town had been reduced to begging—which means there were people with money to beg from—but her family seems to have profited from the bad harvests as a result of the jump in rice prices, and she and her husband were able to purchase items they would not otherwise have been able to because of the rise in the price at which their stipend rice was sold. Although the family had a tenement they rented out, there was no mention of families there suffering or of problems in collecting the rent. Her diary gives a very different impression of the Tempo famine than do the official reports on those who suffered.34 Terashima Kurando, a mid-level samurai of Kanazawa-han, also sheds light on the food conditions during the crisis year of 1837. He was put under house arrest and sent to live in a "hut" on the Noto Peninsula in 1837 because he took too strong a position siding with farmers. The fact that he had two servants makes it clear that he continued to live the life of a samurai.35 He kept a diary, and perhaps because he had nothing to do, his diary contains detailed records of what he ate every day, what food farmers gave to him, what his family sent to him, and so on. During the summer of 1837 he ate rice three times a day, often two and a half or three bowls per meal, along with a different set of side dishes every meal, though some of his meals were rather simple: rice, umeboshi (pickled plum) with sugar, and greens preserved in miso for breakfast. He seems to have had fish at least once a day, and often more frequently. Most of his food was sent by his family, and he kept 34. Kawai Koume, Koume nikki, Toyo Bunko 256 (Tokyo: Heibonsha, 1974). 35. Wakabayashi Kisaburo, editorial supervisor of both Shima monojj atari—Terashima Kurando Noto-jima ryiikei nikki (Kanazawa: Hokkoku Shuppansha, 1982) and Zoku Shima monogatari— Terashima Kurando ryukeichi Noto-jima kara no tegami (Kanazawa: Hokkoku Shuppansha, 1985).

9o

A H E A L T H F U L LIFESTYLE

telling them in letters that he was not in want of anything. He said that he was living among villagers who were in the midst of a famine and who were eating things such as igo, which is brown rice stretched by adding chopped stalks and other hard parts of seaweed.36 Despite the famine, the local people were able to supply Kurando with gifts of seafood, edible bamboo, and local specialties, and though he held deep sympathy for the farmers, he didn't give them any of the sweets he was sent because they wouldn't appreciate them.37 Although Kurando was himself asking for special seasonings such as sesame, poppy seed, mustard, and pepper, by the eighth month of the year people in the countryside had no rice and were eating beans, daikon, greens, and various grasses. He repeated stories he had heard of a mother leaving her children to beg for food and returning a few days later to find them dead. However, townspeople were eating rice, though they were starting to stretch it by adding carrot tops cut fine or daikon38—which in fact may have made for better nutrition though fewer calories. Anecdotal evidence is difficult to use in evaluating nutrition for a population. However, we don't have to rely on such evidence alone for ascertaining nutritional levels for premodern Japan. Two major assessments both use agricultural output figures for regions, and thus what is obtained is an average that did not necessarily pertain to any one person or family. Such estimates also omit what people grew in their own gardens, which was never included in the figures, but the amount must have been considerable. Unfortunately, what data do exist only cover part of the nineteenth century, but they indicate nutritional levels just prior to the Meiji Restoration and in the early Meiji years. The most ambitious project was the study of the Hida area of Gifu from data for 1874. 39 The information includes the amounts of 168 foodstuffs produced by villages and the amount of food imported and exported from the area. By dividing the total amount of food retained in the region by the total population and by 365 days, one can obtain a rough estimate of nutrition available to the "average" person in 1874. 36. Wakabayashi, Zoku Shima monogatari, p. 31. 37. 38. 39. can be

His words were "miso mo kuso mo doyo." Ibid., p. 243. Ibid., pp. 31, 45-46. The implications of this study with regard to the diet for the people in this area found in Fujino Yoshiko, "Meiji shoki ni okeru sanson no shokuji to eiyo: 'Hida

go-fudoki' no bunseki o tsujite," Kokuritsu Minzokugaku Hakubatsukan kenkyii hokoku 7, no. 3 (September 1982): 632-54.

A HEALTHFUL LIFESTYLE

91

The results of the Hida study indicate a heavy dependency on rice and millet, which led to a deficiency of certain essential vitamins and minerals, notably vitamins A and C, calcium, and iron. The diet was somewhat lacking in protein and very high in salt content. This evaluation is borne out by the leading causes of death as analyzed from the records of a local temple. Among major causes of death were childbirth complications (in which calcium deficiencies can play a part), cerebral apoplexy (connected to high salt intake), and epidemics (whose incidence is worsened by a low level of nutrition). The Hida estimates are for a mountainous area, and the authors of the study admit that some items known to have been consumed were not included in the survey, such as sweets, eggs, seaweed, various mushrooms, and wild greens that people could gather from the mountainside. These must have raised the vitamin content of the diet. From the data available for Hida, the authors estimated that the average daily caloric intake was roughly 1,850 calories. A second study has been undertaken on the domain of Choshu.40 Nishikawa Shunsaku determined the caloric intake in the 1840s to have been 1,663 calories per day from staple foods, which included rice, barley, wheat, millet, buckwheat, soybeans, red beans, and sweet potatoes. In terms of nutrition, the Choshu samples show a heavy reliance for protein on grains and soybeans, but primarily on grains, from which 80 percent of the protein was obtained. This estimate is derived from official figures and, hence, is likely to be an underestimate. Moreover, it does not include vegetables, seafoods, or other items not in the preceding list. Although other goods may not have added many more calories, clearly the daily average must have been at least a few hundred calories higher. The figures for Yamaguchi Prefecture in 1887 indicate a caloric intake from the same staple foods of 1,902 per day. Thus the estimated number of calories consumed per day is very close for both of these regional studies. It is impossible to infer a national dietary pattern from a collection of diaries and travelers' accounts and the regional studies that exist. Thus it would be difficult to argue solely on the basis of fragments of dietary evidence that the Japanese were well nourished, or that they were better nourished than Europeans during the seventeenth and eighteenth centuries. The estimates for caloric intake from the Hida and Choshu studies seem very low by current standards, but they may well have been sufficient for the body stature of people at the time. 40. Nishikawa, "Grain Consumption," pp. 436-37.

92

A H E A L T H F U L LIFESTYLE

Japanese army recruits in the late-nineteenth century had an average height of 162 centimeters, which is just under 5'4". 4 1 Massimo LiviBacci has estimated that, assuming substantially lower height and body weight than is the case today, "a population which could rely on a normal consumption of 2,000 calories per head would have been, in centuries past, an adequately fed population, at least from the point of view of energy."42 Also, the very young and the elderly would have consumed less, leaving more calories to the adult males. Even in the 1990s and in one of the most overfed populations in the world, the U.S., the recommended number of calories per day for adult women is only 1,800 to 2,000, and for the elderly about 1,500. However, the Tokugawa diet should not be compared with modern diets, but rather with those of other countries prior to industrialization. Europeans too were heavily reliant upon grains for both calories and nutrients, and even in England in the late-nineteenth century, ordinary workers ate little meat, saving it for special occasions or the head of the household.43 The staple food for the rural poor in England for the sixteenth to eighteenth centuries was "bread, usually made from barley, though occasionally from wheat, oats, rye, or maslin (a mixture of wheat and rye)." Laborers also ate vegetables, milk, cheese, and lard; meat was a rarity.44 Thus, with the exception of dairy products, for which the Japanese substituted soybean products, the Japanese diet was probably similar in content to European diets, even though methods of preparation and seasoning were very different. In fact, there is the suspicion that it may have been better in Asia for many: an Englishman who visited China in the mid-nineteenth century found the diet of the laborers who gathered tea leaves more nutritious than that either of Scottish farm laborers or English sailors. He thought a Chinese used to a diet of rice, vegetables, and a small amount of protein "would starve" on the diet of porridge and milk at one meal, bread and beer at the next, or on a ship dinner of dry salt beef and biscuit.45 41. Nihon teikoku daiku tokei nenkan for 1890, cited in ibid., pp. 4 3 9 - 4 0 . 42. Massimo Livi-Bacci, Population and Nutrition: An Essay on European Demographic History (Cambridge: Cambridge University Press, 1991), p. 27. 43. S. Meacham, A Life Apart: The English Working Class 1890-1914 (London: Thames and H u d s o n , 1977). 4 4 . J. F. C. Harrison, The Common People of Great Britain: A History from the Norman Conquest to the Present (Bloomington: Indiana University Press, 1985). 45. Robert Fortune, A Resident among the Chinese: Inland, On the Coast, and at Sea (London: John Murray, 1857), p p . 4 2 - 4 3 .

A HEALTHFUL LIFESTYLE

93

One difficulty in assessing the Tokugawa diet comes from the fact that nutritionists have insufficient information on what the body needs and how it meets these needs from food. This lack of complete knowledge becomes clear when we look at the effects of the diet of the monks at Eihei-ji, a Zen temple in Fukui, who follow a traditional vegetarian diet of many centuries ago. The monks consume a total of 1,200 calories per day. Breakfast consists of a gruel of unpolished rice (¿¡emmai okayu), pickles, and a seasoning made of sesame and salt. Lunch is rice with barley in it, plus greens seasoned with miso. The evening meal is the same as lunch plus a boiled dish. No meat, fish, or dairy products are consumed. Since the daily intake of calories is only about half that of the ordinary man, why these monks remain healthy is a puzzle. When novices first enter the monastery, they lose weight for the first two months and their legs swell, but after three months, they begin to regain the lost weight and their skin gains a luster. Doctors who have studied these Eihei-ji monks are at a loss to explain this phenomenon and can only conclude that the monks reach a kind of homeostasis. Since the monks are eating a traditional Japanese diet, the health of the monks raises the possibility that Japanese living on a similar diet during the Tokugawa period may have been more healthy than people on other diets with more calories in other countries.46 A number of nutritional studies suggest that the Tokugawa diet may have been healthier than the one that evolved after the Meiji Restoration. We think of the "traditional" Japanese diet as white rice, tea, miso soup, pickles, with side dishes of vegetables, fish, and some meat, usually flavored with soy sauce. Not only is white rice low in vitamin B, but tea drinking has been shown to have "deleterious effects on the thiamin status of human subjects."47 Much is still unknown about calcium and the human body, but studies indicate that calcium loss is related more to protein intake than to calcium intake; for adults, the higher the intake of protein, the greater is the excretion of calcium from the body regardless of calcium intake. A "multitude of factors—protein, phos4 6 . Ishikawa Naofumi, "Eiyo 6-bumme fukkura tsuyatsuya shugyo-shoku," Asahi shim-bun, July 29,1990, p. 1 4 . 4 7 . This study was carried out in Thailand where beriberi was still a common nutritional disease in villagers. Vichai Tanphaichitr, "Thiamin," in Nutrition Reviews' Present Knowledge in Nutrition, 4th ed. (New York: The Nutrition Foundation, 1976), p.

146.

94

A H E A L T H F U L LIFESTYLE

phorus, fluoride and hormonal factors— . . . appear to influence the metabolism of calcium," and hence the incidence of osteoporosis.48 Miso, pickles, and soy sauce are all high in salt content, which contributes to high blood pressure and is significant in the incidence of strokes and heart disease. Thus, as the standard of living rose, the gradual change away from a diet of a porridge of whole grains cooked with vegetables and beverages of hot water and occasionally tea to what became considered the "traditional" Japanese diet could well have meant a decrease in quality in terms of the level of physical well-being. As incomes rose, some could afford a greater variety of foods and hence a better diet, whereas others, particularly in urban areas, may have had a more limited diet. Just as for Europe, we cannot yet make a full assessment. Chapter 6 returns to the issue of nutrition in analyzing life expectancy.

Clothing and Bedding

Improvements in nutrition were not the only reason for a more healthful lifestyle for commoners in the Tokugawa period. One of the most notable developments was the use of cotton as an everyday material for all social and economic classes. Cotton is warm in winter, cool in summer, and its softness and the fact it could be washed must certainly have lowered the rate of skin diseases and irritations throughout the country. It was ideal for use as undergarments and casual clothing. Clothing stuffed with cotton batting was warmer than the options the Japanese had hitherto. And with the diffusion of cotton came the development of bedding, which meant that people could be warm at night without huddling next to others in a tiny, enclosed space. Cotton was introduced into Japan in the late-fourteenth and earlyfifteenth centuries. It was first widely used for clothing in the uniforms of the soldiers fighting the Sengoku wars.49 The greater degree of comfort and durability were the immediate attractions of this new fiber imported from the continent, but of equal importance was the fact that 48. Hellen M. Linkswiler, "Calcium," in Nutrition Reviews' Present Knowledge in 'Nutrition, pp. 232-39. Quote is from p. 239. 49. Nagahara Keiji, Shin Momen izen no koto (Tokyo: Chuokoronsha, 1990), p. 72.

A H E A L T H F U L LIFESTYLE

95

cotton can be readily washed. Although silk remained the preferred material for clothes of the rich, they liked the launderability and comfort of cotton for everyday items such as nightwear and summer casual wear. Lower-income people were formerly clothed in a-sa (translated as hemp or ramie), bark, paper, and assorted other substances, none of which is very satisfactory as material for garments, being neither warm, comfortable, nor readily washable. Thus it was ordinary people who benefited the most from the introduction of cotton. How widespread the use of cotton had become by the early-Tokugawa period can be seen from information on the trade in cotton within Japan. Exactly when the Japanese first began to grow it themselves is not clear, but it was an important commercial product even in the first half of the Tokugawa period, according to Nagahara Keiji. Numerous records indicate that it was grown and sold in various provinces in Japan, particularly in the Kinai at first. The Nqgyd zensho written in 1697 listed the major production areas as Kawachi, Izumo, Settsu, Harima, and Bingo, all provinces in central Japan. In 1714, cotton was second in value only to rice of the goods shipped into Osaka. With the exception of rice, cotton and cotton thread became the most important commercial goods in Japan, more important than rapeseed, dried sardines, sake, lumber, or indigo.50 A regional example of the importance cotton occupied in the lives of the Tokugawa population comes from Okayama, where it developed into a major industry. In the district of Kojima in this domain, cotton was grown, processed, and woven on such a large scale that by 1823 village officials tried to suppress weaving because it paid so well that large landowners couldn't compete for agricultural laborers. By the 1840s, villagers in Fujito had an average of one loom per household; some had as many as three. From early in the nineteenth century, villagers were forming tonya (wholesale organizations) to market cotton in "distant places," and Kokura weave became nationally famous. 51 With the diffusion of cotton throughout Japan came the development of what we think of as bedding. The references we have to futon from the Kamakura and Muromachi periods clearly refer to zabuton (cushions to sit on). The first note of futon as bedding dates from 1592, and even then it is not clear whether the words referred to bedding or 50. Ibid., pp. 130-93. 51. Okayama shishi, Sangyd keizai hen (Okayama: Okayama Shiyakusho, 1966); Ono Takeo, Nihon kinsei kikin-shi (Tokyo: Gakugeisha, 1935).

96

A HEALTHFUL LIFESTYLE

merely nightwear. However, from the 1620s on, there are numerous references to futon, and it is during the seventeenth century when cotton was becoming more popular as material for clothing that its use in bedding increased.52 The spread of cotton was closely correlated with the use of the words yogi (nightwear) and futon (quilts), and this suggests a connected, simultaneous development.53 The type of bedding used in the nineteenth century suggests that bedding developed from nightwear. In the Tokugawa period, the top quilt was shaped like a kosode, with sleeves and a neck, and quilts in this style could be found in the Kansai even after World War II. Early references to bedding indicate a single quilt with a sleeve on one side; the sleeper wrapped up in it with the sleeve on top. Samurai were supposed to sleep with the right hand free, and therefore on top, and thus bedding in this style was considered wartime bedding. Ieyasu is said to have slept in such a quilt.54 Of course, the poor could not afford such extravagance, but for those who could, quilts meant that the family was warm at night and no longer had to huddle together in the nando for body warmth. Lives of the sick and aged may sometimes have been saved because of the additional warmth, and less huddling meant people spread fewer germs. Bedding let people sleep in various places throughout the house, freeing them from the nando, which when spread with straw and closed up became bug-infested and damp. However, when people merely substituted cotton bedding for straw in the nando and left it spread out day and night, it must have been as unhealthful as straw because cotton readily absorbs moisture. Nevertheless, even if many people were still sleeping in bedding stuffed with straw or other materials as late as the mid-nineteenth century, they were more likely than not wearing cotton. The timing of the diffusion of cotton in Europe paralleled that in Japan. It was introduced in the same centuries and subsequently became the most popular material, just as it was in Japan. The major differences were that northern European nations had to import raw cotton from warmer countries to the south, and yet it was these same northern nations—as well as America—that developed the technology for processing and manufacturing cotton cloth that Japan was subsequently to borrow. The intro-

52. Ogawa Koyo, Shinjo to shinj;u no rekishi, vol. 7 of Fiizoku bunka-shi sensho (Tokyo: Yuzankaku, 1973), pp. 146-70. 53. Ibid., p. 150. 54. Ibid., p. 170.

A HEALTHFUL LIFESTYLE

97

duction of cotton into Europe made possible the development of underwear that could be washed, raising the level of comfort and hygiene as it did in Japan. Cotton was not only softer to the touch but could be boiled and sterilized when laundered.

Bathing If the Europeans boiled their laundry, the Japanese began to "boil" themselves, or at least that's what it seemed to Western observers.55 Bathing became a regular part of life in Japan during the Tokugawa period, though precisely how it began or when it became prevalent is open to question. Literary sources from the Heian period indicate that the aristocracy enjoyed no such custom in the tenth and eleventh centuries, but by the eighteenth century there were so many references to bathing and so many regulations that it was by then a widespread custom. The first mention of a public bath was one in Gion in the early 1320s.56 Though we don't know when public baths became popular, Miura Joshin, writing about Edo in the early-seventeenth century, noted that there was a "public bath in every neighborhood," that they were very inexpensive, and that everyone as far as he knew went to them.57 By the Tokugawa period, a variety of furo (the term for any kind of bath) were in common use, ranging from steam baths to tubs in which the bather sat in hot water and soaked.58 There are numerous references to baths in literature: Kitahachi in Hizakurige went to a ¿¡oemon-buro (a large iron pot with wooden planks to sit on), mention of baths was made in haiku, and there were many references in the works of Saikaku. The most solid evidence on the prevalence of public baths was the regulation of sento, the public baths, particularly from the eighteenth century on. The need for such regulations arose not only out of the number 55. Edward S. Morse, Japanese Homes and Their Surroundings (1886; reprint, Rutland, Vt.: Charles E. Tuttle, 1972), p. 202. 56. Ochiai Shigeru, Arau fuzoku-shi (Tokyo: Miraisha, 1984), pp. 71-72. 57. Nakamura Kazuhiro, comp., Edo shiryo sosho—Keicho kemmon-shu (1614; reprint, Tokyo: Shinjinbutsu Oraisha, 1969), pp. 160-61. 58. It should be noted that for the most part baths in Japan were not individual tubs of water used by just one person; the hot water was communally used in public baths and sequentially in homes. Thus they were not as sanitary as individual baths but enabled more people to bathe more often.

98

A HEALTHFUL LIFESTYLE

Wooden bath: water in the tub was heated from underneath with fuel fed from the other side of the wall.

of public bathhouses, but also out of the fact that they became associated with various social problems and crime, just as they did in other societies such as ancient Rome. Regulations in Edo in the 1790s prohibited mixed-sex bathing, and thus bathhouses had to set different days in the month for bathing by women and by men. Regulations in Hakodate (Hokkaido) for 1854 required a guard to make sure people's clothes and other belongings were not stolen, and although mixed bathing was permitted, one person from the bathhouse guild was to ascertain that a proper moral atmosphere was maintained. In Hakodate, the public baths were to be open from 6 A.M. to 8 P.M. daily with only two annual closures.59 From the numerous references to bathing in laws, literature, woodblock prints, travelers' notes, information on architecture and cities, plus humorous anecdotes, it is evident that all classes were bathing regularly by the nineteenth century.60 Not only were people going to the public baths in large numbers 59. Tanigawa Ken'ichi ct al., eds., Nihon shomin seikatsu shiryo shiisei, vol. 15 (Tokyo: San'ichi Shobo, 1971), pp. 57, 823. 60. Konno Nobuo, Edo nofuro (Tokyo: Shinchosha, 1989).

A HEALTHFUL LIFESTYLE

99

from the middle of the Tokugawa period on, but they were installing baths in their homes as well. In her journal, the pawnbroker's wife, Omine, noted the days on which she took a bath at home. In 1791, she took a bath in the fifth month on the first, fourth, fourteenth, twentyfourth, twenty-sixth, and twenty-seventh and in the sixth month on the second, third, fourth, and eighth. In the ninth month of the same year, her diary noted that she bathed on the second, fourth, sixth, eighth, tenth, fourteenth, sixteenth, eighteenth, twenty-third, twentyfifth, twenty-seventh, and twenty-ninth, and she heated the water for the family bath on the twelfth, though it is not clear whether she bathed on that day or not.61 So many different kinds of baths were used in the Tokugawa period that it is not often possible to discern which kind is referred to in such references as Omine's to an "evening bath." Early baths seem to have been steam baths (mushiburo), and early bathhouses used a combination of hot water and steam, but by the late-nineteenth century, the most common type of bathtub was one in which the bather immersed in hot water that was heated directly in the tub. However, the mushiburo was also common, and in the domain of Satsuma even in the late Meiji period, public baths had two baths, one with hot water and one using steam.62 From all the references to bathing by travelers and diarists and in regulations, novels, and other materials, it is clear that bathing was very common by the late-Tokugawa period. In the city of Edo, approximately six hundred public baths are estimated to have been in operation by the early-nineteenth century, and these were extensively used and extremely crowded.63 Bathing was considered an everyday aspect of life, and travelers wrote about it not as an unusual phenomenon, but with interest in the diverse kinds of baths found in the various parts of the country they visited. Despite all the descriptions and drawings of baths, we know little about how they developed, how much of the population used them and how often, and which kind was used. Clearly the well-to-do were more likely to have furo or access to them than the poor, and people who lived near hot springs bathed much more frequently than those elsewhere. But if we compare the customs of bathing in Japan with those 61. See the diary entries in Nichiroku for the dates referred to in the text. 62. Tanigawa et al., Nihon shomin seikatsu shiryo shusei, vol. 12, p. 354. 63. Ochiai, Amu fuzoku-shi, p. 75.

A H E A L T H F U L LIFESTYLE

in the West at the same period, the importance of furo and the level of cleanliness in Japan becomes evident. Whereas regular bathing was common in Japan at the time of the Meiji Restoration, few English houses had bathrooms in the mid- to late-nineteenth century. Not even the rich built bathrooms, preferring instead to rely on servants to prepare a bath in front of the bedroom fireplace so that it could be taken in complete privacy. But this entailed bringing in a bathtub, protecting the floor from the water, heating the water in the kitchen downstairs, and then lugging it all the way upstairs pitcher by pitcher. When the bath was over, the water had to be carried back downstairs and the equipment put away. Needless to say, taking a bath was not quick or easy, and people often resorted instead to partial baths—foot baths, sponge baths, and washing just the face and hands. In small houses of the working classes, baths in the nineteenth century were an occasional event, usually in a tub in front of the kitchen fire on Saturday nights in preparation for the Sabbath. This was a great advance over the seventeenth and eighteenth centuries when people rarely if ever bathed; often it was said a person had just three baths: at birth, just before one's wedding, and after death.64 In the West, bathrooms and fixed bathtubs didn't appear until the late-nineteenth century. The idea of locating the toilet, the sink, and the bathtub in the same room was an American innovation.65 Although the famous architect Andrew Jackson Downing published floorplans with a "bathroom" as early as midcentury, many of the plans for his smaller cottages contained neither a bathroom nor a water closet.66 Even in the 1890s, city apartments were being designed in the U.S. that had no bathroom, and thus the movable tub by the fire must have provided any bath the residents had.67 Conditions were far worse for the poor who were crammed many persons into a couple of rooms without the privacy to bathe—particularly given the mores that said nudity was shameful.

64. Jane C. Nylander, Our Own Snug Fireside (New York: Alfred A. Knopf, 199?), pp. 143-4-8. 65. Witold Rybczynski, A Short History of Home (New York: Viking Penguin, 1986), p. 1 6 4 . 66. Andrew Jackson Downing, Victorian Cottage Residences (1842; reproduction of 1873 ed., New York: Dover Publications, 1981). 67. David P. Handlin, The American Home (Boston: Little, Brown and Company, 1979).

A HEALTHFUL LIFESTYLE

101

It is difficult, if not impossible, to draw general conclusions about standards of personal cleanliness in either the West or Japan in premodern times, and equally difficult to make international comparisons. But on the basis of the following evidence it is possible to suggest that, although there were vast differences in standards of personal hygiene within Japan and within Western countries, on balance the standards were at least as high in Japan as the West and very possibly higher. Bathing did not become an important part of modern European culture nor was there a special room set aside for it until the nineteenth century. Even then, the provisions were primarily for the well-to-do and entailed a great deal of labor. By this time all well-to-do Japanese, and many others, had bathrooms. In the Western literature, there is no mention of provisions for bathing for lower-income families or people crammed together in boarding houses or tenements. In contrast, bathing was important in Japan from well before the Tokugawa period, and public baths that were cheap or relatively cheap were available in major cities by early- to mid-Tokugawa. Visitors to Japan in the sixteenth century were impressed by the bathrooms for the rich and powerful, and the emphasis visitors such as the Jesuit priest Joao Rodrigues put on cleanliness in describing the sanitary facilities suggests higher standards than he was accustomed to: "This custom of taking a bath is universal throughout all Asia. . . . but the Japanese seem to excel everybody else in this matter, not only in the frequency with which they bathe during the day, but even more so in the cleanliness and dignity which they observe in that place."68 At the other end of the spectrum is the dirt that Isabella Bird was not prepared to find when she visited mountain villages north of Nikko only a decade after the Meiji Restoration. But even there, the people told her "that they take a bath once a week." Bird was not impressed with Japanese baths in which "the water is used without any change by all the inmates of a house, and in the public baths by a large number of customers." And she thought friction did not make up for the lack of soap.69 However, how many people in the West in such poor circumstances were taking anything resembling a bath once a week? Isabella

68. Joao Rodrigues, S.J. as quoted in Michael Cooper, S. J., ed., They Came to Japan (Berkeley: University of California Press, 1965), p. 221. 69. Isabella L. Bird, Unbeaten Tracks in Japan (New York: G. P. Putnam's Sons, 1881), p. 171.

io2

A HEALTHFUL LIFESTYLE

Bird reported on people homesteading in Colorado with no house after nine years and on filth so terrible in a Chicago inn that she couldn't eat—experiences worse than any she encountered in Japan.70 The Westerners who visited Japan in the early-Meiji period by and large came from middle-class and well-to-do homes. They were people who took weekly baths, lived in the large but cluttered homes of Victorian times, ate well, and had for the most part strong opinions as to what was proper, civilized, and healthful. These were the people who fussed over fleas in the bedding in Japanese inns, disliked the Japanese diet, and found Japanese rooms barren and uncomfortable. But what is missing from most accounts is a comparison with similar people and circumstances in their own countries. Westerners were repulsed by Japanese food because it was not what they were used to eating, not because it was unsanitary. Yet it was in the U.S., not in Japan, that Isabella Bird was so disgusted by the food and sanitary conditions that she neither wanted to eat nor put her belongings down in a room. What is key is the diversity of conditions of housings food, and sanitation compared to present industrial societies; travelers who were repelled or appalled by conditions in countries other than their own were often unaware of the wretched conditions of many in their own societies. Though very different from the West, Japan clearly was not necessarily a worse society in which to live in terms of material culture or living conditions. When looking at the standard of living and the level of physical wellbeing, there is always the problem of whether the cup is half full or half empty. Many people lived in tightly closed, damp, dark, and rather unhealthful housing, had inadequate diets, and were poorly clothed during the Tokugawa period. Thus the argument can be made that people were very badly off. But in looking at the level of physical well-being from the historical perspective, I find more persuasive the argument that the level of physical well-being improved over the course of the period and that it was more or less comparable with that in the West even in 7o. "I got down and found a single room of the rudest kind, with the wall at one end partially broken down, holes for windows, and no furniture but two chairs and two unplaned wooden shelves, with some sacks of straw upon them for beds. There was an adjacent cabin room with a stove, benches, and table, where they cooked and ate, but this was all. . . . They all slept under the trees, and before dark carried the sacks of straw out for their bedding. . . . I make my bed on the floor, and draw a bucket of ice-cold water from the river." Isabella L. Bird, A Lady's Life in the Rocky Mountains (New York: G. P. Putnam's Sons, 1900), pp. 4 5 - 4 6 . Her description of an inn in Chicago was equally dismal. Isabella Lucy Bird, The Englishwoman in America (1856; reprint, Madison: University of Wisconsin Press, 1966), pp. 149-50.

A HEALTHFUL LIFESTYLE

103

the nineteenth century. Except for famine years, most people had adequate diets, and neither agriculture nor the commercial economy suffered because people had a lack of energy. Comfort levels increased in clothing, cleanliness, bedding, and housing. Analysis of how this affected their physical well-being is found in the following chapters on sanitation, longevity, and other demographic indicators of well-being.

CHAPTER

FIVE

Urban Sanitation and Physical Well-Being

On November 2 4 , 1 8 7 7 , R. W. Atkinson read a paper before the members of the Asiatic Society of Japan in Tokyo on "The Water Supply of Tokio." With his students at the University of Tokyo, he had tested for the presence of organic matter in samples from various parts of the Tokyo water supply, including from the Kanda system, the Tama system, and surface waters. As he explained to the members of the society, the presence of organic matter was an indication of "contamination with sewage matter,'" and thus the lower the level of organic matter, the better the water was in quality. This survey revealed that the water in Tokyo's water supply system was unusually close to pure at the source, but the farther the samples were taken from the source, and especially the surface samples, the more likely they were to 1. Atkinson tested Tokyo's water at various points in the system for the existence of solids, chlorine, ammonia, nitrogen, and so on. At this time, it was not possible to test for bacteria content, or "germs" as they were called by Atkinson, but scientists could chemically test water for organic content that would indicate the approximate contact with raw sewage and thus the likelihood that it would cause disease. R. W. Atkinson, "The Water Supply of Tokio," Transactions of the Asiatic Society ofJapan 7, pt. 1 (1877-1878): 90, 96. Several years later, O. Korschelt assessed the quality of wells in Tokyo and concluded that artesian wells were a suitable means of supplying water to areas of Tokyo not reached by piped-in water. He was even more salutary in his assessment of Tokyo's water supply than was Atkinson. O. Korschelt, "The Water Supply of Tokio," Transactions of the Asiatic Society of Japan 12, pt. 3 (1884).

104

U R B A N SANITATION AND PHYSICAL WELL-BEING

105

be contaminated. That a premodern system using wooden pipes was contaminated in places should surprise no one; what is astonishing is that Atkinson concluded that Tokyo in the decade just after the Meiji Restoration had a water supply purer than did London. After all, he was comparing the water supply system of a city yet to begin using modern technology with the system in the largest city in Europe's first industrialized nation, one using metal pipes and presumably the latest technology! A strong case can be made that sanitation in Japan through the midnineteenth century was as good as or better than in the West. Today we associate good sanitation with the high living standards of the industrialized nations, but evidence from the nineteenth century indicates that this was not always the case. In England, Queen Victoria's consort, Prince Albert, died in 1861 of typhoid fever thought to have been contracted as a result of faulty sewage drains. Cholera, typhoid, and other bacterial diseases spread by filthy conditions were endemic and epidemic in American, English, French, and certainly other Western towns and cities because of polluted water supplies and inadequate sewage and garbage disposal. Although the quality of drinking water became a matter of medical concern in England in the mid-eighteenth century, it wasn't until the mid-nineteenth century, when the connection between the spread of cholera in London and the water supply was made, that serious efforts were undertaken to control and improve water quality, and it wasn't until 1921 that London's water supply was considered safe.2 In contrast, during the Tokugawa period, Japan's water supply and waste disposal methods were generally efficient and relatively hygienic in both rural and urban areas. Problems with sanitation in urban areas would be expected to most severely affect the population, but Japan's cities grew in size so that they far surpassed any Western city, and disease levels were low, surprisingly so to early visitors from the West. Japan's urbanization in the Tokugawa period may well have no parallel in world history prior to the twentieth century. Edo grew from a cluster of fishing villages surrounding a castle in 1590 to a population of approximately one million by the early-eighteenth century. In 1700, London, the largest European city, had a population of just over half 2. Anne Hardy, "Water and the Search for Public Health in London in the Eighteenth and Nineteenth Centuries," Medical History 28 (1984): 250-82.

106

U R B A N SANITATION A N D PHYSICAL WELL-BEING

that, 575,ooo, and Edo rivaled or exceeded in size the largest of the Chinese cities, Peking. Urban growth in Japan was not limited to Edo, or even the other great metropolises of Osaka and Kyoto, but included dozens of castle towns that sprang up from the sixteenth century on and port towns and cities and other earlier urban centers. By the lateeighteenth century, Japan had about 3 percent of the world's population, but it is estimated to have had more than 8 percent of the people in the world who lived in cities of more than ten thousand. By this standard, approximately 10 percent of the total population of Japan was urban in 1800, and some scholars estimate it was even higher.3 How did Japanese cities retain such large populations? Neither political turmoil nor economic problems, crop failure nor disease significantly dented the proportion of the population that lived in cities during the latter half of the Tokugawa period. In large premodern cities— certainly in both London and what we know of Tokugawa cities— birthrates were lower than deathrates, and the population could only be maintained by a constant influx of immigrants. Yet, in Japan, when the urban population in one area fell over time, it was compensated by growth in nearby towns and cities.4 In Takayama of Gifu, both birthand deathrates were low, and deathrates were often below birthrates and only climbed significantly higher during the Tempo mortality crisis.5 Though Takayama depended on in-migration for its population growth, its deathrates were comparable to those of the surrounding countryside.6 Any decline in the population of various cities during the 3. Oishi Shinzaburo, "Kyodai toshi Edo no jinko wa dore kurai ka," Rekishi to jimbutsu no (1980: 76-81); E. A. Wrigley, "A Simple Model of London's Importance in Changing English Society and Economy 1650-1750," in Towns in Societies, ed. Philip Abrams and E. A. Wrigley (Cambridge: Cambridge University Press, 1978), p. 215; G. William Skinner, ed., The City in Late Imperial China (Stanford: Stanford University Press, 1977), p. 29. Gilbert Rozman, Urban Networks in Ch'ing China and Tokugawa Japan (Princeton: Princeton University Press, 1973), p. 6, estimates Japan's population to have been 10 percent urban, whereas Sekiyama Naotaro, Kinsei Nihon no jinko kozo (Tokyo: Yoshikawa Kobunkan, 1958), p. 239, considers 12 percent of the population to have been urban. 4. See evidence for this in Thomas C. Smith, "Pre-Modern Economic Growth: Japan and the West," Past and Present, no. 60 (August 1973): 127-60. 5. Sasaki Yoichiro, "Hida-no-kuni Takayama no jinko kenkyu," in Keizai-shi ni okeru jinko, report of the thirty-seventh meeting of the Shakai Keizai-shi Gakkai (Tokyo: Keio Tsushin Kabushiki Kaisha, 1969), P- 106. 6. See rates for various areas in Japan in Susan B. Hanley and Kozo Yamamura, Economic and Demographic Change in Preindustrial Japan, 1600-1868 (Princeton: Princeton University Press, 1977), especially chapters 8 and 11.

U R B A N SANITATION AND PHYSICAL WELL-BEING

107

Tokugawa period was a result of changes in the economy rather than high urban deathrates. What I argue made the critical difference in Japan was the level of sanitation. Even if all other requirements to supply and service a city are met, one would not expect such large cities or such a high proportion of the population to live in urban areas, or for these to persist over several centuries, if a city cannot remain a healthful place to live. Yet Edo, Osaka, and Kyoto, which lacked modern water supply and sewage systems and which relied primarily on human labor or boats to carry away garbage, refuse, and night soil, maintained huge populations throughout the Tokugawa period. In the West, there exists a wealth of materials on the problems of water supply and sanitation, particularly for the nineteenth century when technology was being applied to solve problems of sanitation. Less information is available for Japan, possibly at least partly because the Japanese did not have to contend with such grave urban problems as Westerners faced. But from Japanese laws, regulations, and political concerns, a good deal of information can be extracted about Edo's water supply and waste disposal systems and about Osaka's system for disposing of night soil. When Tokugawa Ieyasu selected Edo for his capital, he recognized from the outset the problem of obtaining an adequate water supply and ordered a former retainer, Okubo Togoro Tadayuki, to construct a water supply system. In 1590, Okubo went to Edo to assess the situation and make plans. The system he began was so large in scale and so successful that it has been compared to that of the Romans.7 Although Edo was strategically situated, much of it was on low, marshy ground near the sea. Potable water was originally obtained from ponds and underground springs, but wells had to be deep because of the depth of the aquifer. The first system constructed, the Kanda system, drew its water from the Inokashira spring east of the city. Water was carried to the city limits mainly in exposed aqueducts, and then in underground aqueducts or wooden pipes within the city. The Kanda system was over forty-one miles in length and had 3,663 subsidiary ducts that drew water from the main source and distributed it to various parts of the city. At selected points water pressure was increased to lift water from one level to another, but because the Kanda system was designed for delivering water to the lower-lying sections of the city, natural wa7. Higuchi Kiyoyuki, Edo, vol. 4 (Tokyo: Haga Shoten, 1968), p. 253.

108

U R B A N SANITATION AND PHYSICAL WELL-BEING

terflow was usually sufficient. A problem with this system was that the water from the Inokashira spring was limited in quantity, and if too much was pumped out, the quality suffered—the water turned muddy. By the mid-seventeenth century, this system was inadequate for the growing city's needs, and a new system was begun in 1652.8 The second system took its water from the Tama River. Well over fifty miles in total length, it was larger than the Kanda system and carried a far greater volume of water. The Tama River system brought water up to the Yotsuya gate of Edo castle, a distance of nearly twentyseven miles, supplying first of all the shogun and then the nearby areas, Kojimachi, Yotsuya, Kyobashi, and Akasaka. When first constructed, there was sufficient water to serve the city and to irrigate rice paddies in Musashino, a farming area west of Edo. However, as the city expanded, the use of the Tama River system's water for irrigation had to be curtailed. The Tama River system proved to be inadequate, not so much because of the insufficiency of water but because of the difficulties of raising the water to higher ground as the city expanded outward. Another major stimulus for continued construction of water systems was the problem of frequent and devastating fires in Edo, whose buildings were constructed primarily of wood. In the great Meireki fire of 1657, approximately two-thirds of the city was destroyed; deaths were estimated in the tens of thousands. Subsequently, a policy was implemented to decrease the density of the population at the center, and many samurai mansions and temples were relocated on the outskirts of the city, leaving open spaces within the city to serve as fire breaks. Four new water systems were added—the Honjo, Aoyama, Mita, and Sengawa—all of which relied on the Tama River for their water supply and were subsidiary to the main system. These newer systems seem not to have been entirely satisfactory for, by the mid-nineteenth century, Edo relied primarily on the original Kanda and Tama River systems. Since the Kanda system now received some of its water supply from the Tama River, Edo's water supply came mainly from that one river. During the century it took to build Edo's water systems, engineering techniques became increasingly sophisticated. In the early part of the 8. Descriptions of Edo's water supply system are to be found in Horikoshi Masao, Ido to suidd no hanashi (Tokyo: Ronsosha, 1981); Sabata Toyoyuki, Suido no bunka: Seiyd to Nihon (Tokyo: Shincho Sensho, 1983); Ito Koichi, "Edo no suido seido," in Edo chonin no kenkyu, vol. 5, ed. Nishiyama Matsunosuke (Tokyo: Yoshikawa Kobunkan, 1978), pp. 283-308; Higuchi, Edo, vol. 4 , pp. 231-47.

URBAN SANITATION A N D PHYSICAL WELL-BEING

109

seventeenth century, ditches were the most common conduit outside city limits. For its underground aqueducts, the Kanda system used square pipes of red pine, but other early systems had pipes made of other wood, stone, earth, and bamboo. By the time the Tama River system was constructed, the engineers designing it were using triangulation, which helped solve the problems created by differing elevations of land, and the measurements they made were "surprisingly accurate."9 Siphons were used to draw water up into Edo castle from the ducts of the Tama River system. The newer the system, the more pipes it had rather than open channels, and thus much more of the later systems were underground. The government not only was responsible for the building of the water systems supplying Edo, but also tightly regulated their use. No individual was permitted to directly draw water from the main system unless that individual was a samurai of high status, usually a daimyo or high-ranking retainer of the shogun. The public was supplied with water from wells built into the aqueduct system; people were required to go to the nearest well and draw water rather than tap the nearest duct themselves. This control ensured an adequate flow of water and proper maintenance of the system. In Edo, there was always a sufficient water supply twenty-four hours a day, and stoppage of water was so rare that there were no backup arrangements for emergencies in the city. Indeed, the Tama River system brought so much water into Edo that a waterfall in the Shinjuku Imperial Gardens was created from the surplus. When the Ebisu Beer Company was founded in the early-Meiji period, it used the water from this same system founded over two centuries earlier.10 Thus, the city of Tokyo inherited a water system both sufficient in quantity and pure enough in quality to serve its early needs. Atkinson's test of the water supply in the 1870s bears this out. Furthermore, Edo's system was so well designed that when it was modernized at the end of the nineteenth century, the only major change was to replace the wooden pipes with impervious metal ones. Thus, the Japanese were able to use the main features of a system constructed in the seventeenth century when converting to a water supply system based on modern technology more than two centuries later. The engineering feat involved in building such a large-scale and sound sys9. Higuchi, Edo, vol. 4 , p. 244. 10. Sabata, Suidd no bunka, p. 32.

no

U R B A N SANITATION A N D PHYSICAL WELL-BEING

tern was indeed remarkable, especially considering that dirt from the excavations for the channels and works was hauled away in baskets or straw mats. How did Edo's water supply system compare with that of Europe's largest city? Although London's "New River" was constructed in the same period as Edo's Kanda system, the success in bringing spring water to London in 1613 was scarcely the equivalent of the construction of the Kanda system in Edo in the same period. In London, most of the pipes were on the surface and watchmen had to be hired to deter tampering. As London expanded, the water supply became increasingly inadequate, but unlike in Edo, no ready solution was found. Pipes could be added but the supply was not increased, and so by the mid-eighteenth century Londoners could draw water only seven hours a day, three days per week. In contrast, Edo's pipes were not as strong, but they were buried within the city, and the water supply was adequate.11 Clearly Edo was a special case. Its water supply system was designed at the same time the city was, and it was ordered, constructed, and paid for by the most powerful political force in the country. Edo's system was vastly superior to the water supply system in any other city. Yet other cities had good water supply systems. With the building of the castle towns from the late-sixteenth century on, water supply systems were increasingly constructed to supply drinking water to urban populations, not only in central Japan but also in such far-flung places as the southern tip of Kyushu. Systems to supply potable water to cities in what are now the prefectures of Shiga, Hyogo, Oita, Hiroshima, Mie, Kagawa, and Kagoshima were built before 1650, and many more followed during the late-seventeenth and eighteenth centuries. In addition, numerous irrigation projects supplied water used for drinking as well as for irrigation. 12 Both Kyoto and Osaka relied to a large extent for their supply on Lake Biwa, the largest lake in Japan, termed "nature's huge reservoir." 13 In most cities, as in the countryside, rivers and wells constituted the main source of water. This is not to argue that the premodern Japanese water supply system was without its problems. A major hazard for any water supply system is the danger of pollution from human wastes. But again, Japanese cities did not have the horrendous problems of sewage contamination of the main water supply that London and Paris did, for several reasons. Water n. Ibid., pp. 32-34. 12. Horikoshi, Ido to suido, pp. 97-99. 13. Sabata, Suido no bunka, p. 97.

URBAN SANITATION AND PHYSICAL WELL-BEING

hi

in Japan tended to be purer at the source since the main islands are all dominated by mountain ranges, resulting in short, swift rivers. In addition, abundant rainfall, particularly during the typhoon season, helped flush out the rivers, preventing the pollution that occurs in slow, meandering rivers. But Japan's good quality water supply did not result solely from geographical luck. Rather, people were not disposing of night soil in rivers where it would contaminate the river as a source of water supply, hence the high quality of the sources of Edo's water supply. The most important difference between waste disposal in Japan and in the West was that in Japan human excreta was not regarded as an economic "bad"—something that one paid to have removed—but rather as a product with a positive economic value. The night soil of Japanese cities was long used as fertilizer to a much greater extent than occurred in Western countries. With the growth of Japan's population, the limited amount of arable land and the increasingly intensified use of land to feed the growing population, combined with the relative scarcity of animal wastes and other fertilizers, meant that human waste had a value as fertilizer that far exceeded its value in the West. In short, human waste was in economists' language an economic "good." Long before Edo was even established, Osaka's night soil was used as fertilizer for the surrounding farm villages. Most of it was collected, loaded onto boats, and distributed to ports in nearby farm areas. The huge volume brought to the wharves resulted in such an unpleasant odor that there were complaints. In the Tokugawa period, the magistrates deliberated upon these complaints but concluded that "it was unavoidable for the manure boats to come into the wharves used by the tea and other ships."14 In the early years of the Tokugawa period, boats were sent into Osaka loaded with farm produce that was exchanged for the night soil of the city. But as the price of fish and other fertilizers rose, the value of night soil rose correspondingly, and vegetables and farm produce were no longer sufficient to pay for it. By the early-eighteenth century, with the increase in new paddies in the Osaka area, the price of fertilizer had jumped to the point that even night soil had to be paid for with silver. The value of human wastes was so high that rights of ownership to its components were assigned to different parties. In Osaka, the rights to fecal matter from the occupants of a dwelling belonged to the 14. Wakita Osamu and Kobayashi Shigeru, Osaka no seisan to kotsii, vol. 4 of Hoso Bunka Sosho (Osaka: Mainichi Hoso, 1973), p. 127.

Mainichi

ii2

URBAN SANITATION A N D PHYSICAL WELL-BEING

owner of the building, and the urine belonged to the tenants. Feces were considered more valuable and hence commanded a higher price. Generally speaking, the price of fecal matter from ten households per year amounted to over one-half a ryo of gold. This was equivalent to just over one month's wage for a farm laborer in the early-nineteenth century. Because human waste was so valuable, rents were adjusted on the basis of how many tenants occupied a unit and were raised if the number of occupants dropped. 15 With the rapid growth of Osaka in the seventeenth century, the city government found by midcentury that it had to step in and form guilds to ensure that waste disposal was handled properly. By the end of the century, farmers from neighboring areas were forming associations for the purpose of obtaining monopsony rights to purchase the increasingly expensive night soil from various areas of Osaka. Eventually fights broke out over collection rights and prices. In the summer of 1724, two groups of villages from the Yamazaki and Takatsuki areas (between Kyoto and Osaka) fought over the territorial rights to collect night soil in the city. Other disputes arose between the guilds in the city and farmers' associations, and examples exist for the neighboring provinces of Kawachi and Settsu as well, indicating that this type of conflict was neither a localized nor isolated event.16 In the three major areas of Osaka, neighboring farm villages held the rights to collect night soil from households, but they either could not or did not want to collect all of the urine. The remainder was left to be collected by shoben nakagainin (urine jobbers). The number of jobbers gradually increased, as did jurisdictional problems. Eventually they created their own association, and in 1772 they paid a fee to the Osaka authorities to establish a kabunakama (guild based on ownership of shares) with the authority to enforce jurisdiction and to set prices. However, the rights to collect the urine from containers left for passersby on the street corners in Osaka were given to an outcast village named Watanabe, but even though the price of urine was lower than for fecal matter, there were constant conflicts over these collection

15. Although urine is usually higher in nitrogen and potash than solid excreta and is especially useful as an activator in converting crop residues to humus, it is more difficult to transport and store than solid excreta, which probably accounts for its lower price. The price of fecal matter is in Watanabe Minoru, Mikaiho buraku-shi no kenkyu (Tokyo: Yoshikawa Kobunkan, 1 9 6 5 ) , p. 297. The value of a ryo was calculated from Yamasaki Ryuzo, "Edo chuki no bukka doko to keizai hendo" in Rekishi no naka no bukka, ed. Harada Toshimaru and Miyamoto Matao (Tokyo: Dobunkan, 1 9 8 5 ) , p. 78. 16. Wakita and Kobayashi, Osaka, p. 128.

URBAN SANITATION AND PHYSICAL WELL-BEING

113

rights. Periodically, other people tried to get these rights away from Watanabe, but the village managed to maintain its monopoly through the Tokugawa period, despite sabotage of its containers, challenges by others to its collection rights, and offers to buy the rights. 17 In Osaka, by the mid-eighteenth century, night soil was so clearly an economic good that ownership and monopoly and monopsony rights were assigned, the formation of officially recognized associations and guilds was permitted, and the price could be determined by these organizations. The price rose so high that the lowest-income farmers had difficulty in obtaining sufficient fertilizer, and incidents of stealing began to appear in the records, despite the fact that going to prison if discovered was a real risk. In Edo, judging from official records, disputes over night soil were not as serious a problem as they were in Osaka, but government did have to step in to handle problems of orderly waste disposal. Edo's waste, other than night soil, was of four types: household waste, probably mostly kitchen garbage; trash discarded along the roads and in the waste water drains; junk floating in waterways—moats, rivers, and the harbor; and waste from fires. In addition, Edo had the problem of disposing of its waste water since its ample water supply meant that enormous quantities of water had to be discarded somewhere.18 Regulations regarding waste began to appear in Edo during the midseventeenth century. The focus at this time was not waste collection per se, but the growing problem of keeping the streets, open areas, and drainage ditches free from rubbish. This was viewed as a problem within the purview of city government. At the same time, problems relating to the disposal of night soil came to the attention of the authorities. In 1648, city regulations required small huts and toilets along the banks of rivers to be torn down. The repeated issuance of this and other regulations over the next half-century indicates that Edo residents must have been slow to comply with the new, more sanitary arrangements for waste disposal.19 These regulations also indicate that demand for Edo's night soil was 17. Watanabe, Mikaihd, pp. 292-99. See also Osaka-shi, ed., Osaka shishi, vol. 1 (19111915; reprint, Osaka: Seibundo, 1965), pp. 866-68. 18. Hayashi Reiko, "Kinsei ni okeru jinkai shori," Kyutsu keizai ronshu 8 (1974): 7 2 86.

19. For information on waste disposal in Edo, see Ito Koichi, "Edo ni okeru gomi, gesui, shinyo no shori," in Nihon no hoken tosbi, vol. 2, ed. Toyoda Takeshi, Harada Tomohiko, and Yamori Kazuhiko (Tokyo: Bun'ichi Sogo Shuppan, 1983), pp. 431-55; Kawazoe Noboru, Urajjawa kara mita toshi (Tokyo: Nihon Hoso Shuppan Kyokai, 1982), pp. 152-90.

URBAN SANITATION A N D PHYSICAL WELL-BEING

probably not high during the early-Tokugawa period and much potential fertilizer must have been wasted. However, as truck gardening developed in the area around Edo and the city came to rely less on a supply of daily goods from afar, the demand for night soil rose. The market for the night soil of Edo was the farm villages surrounding the city within a radius of about ten miles. By the first half of the eighteenth century, the sudden withdrawal of a particular supply of night soil could be devastating to a farmer, because alternate supplies were not easily found. A village head in the neighboring district of Tama lost his supply of fertilizer in 1725 when the main residence of the daimyo of the Owari Tokugawa burned. Before the mansion could be rebuilt, the farmer suffered major crop losses. Contracts were given to farm villages for the night soil from specific areas in Edo. Usually each daimyo contracted out these rights, with the price determined on the basis of the market demand. For example, the Hitotsubashi daimyo sold the rights to night soil from the residence in 1742 to one Hanbei from Tanashi village in the district of Tama for the price of fifteen hundred large daikon (large white radishes), two thousand middle-sized daikon or two ryd in cash, whichever Hanbei preferred, to be paid at the end of the year. Each year this daimyo house decided the price by bidding out the rights. Each daimyo made different arrangements, sometimes with the payment made twice a year and the price as high as six ryo or more.20 As the price of night soil rose, entrepreneurs sought rights to place containers to collect urine on busy street corners in Edo, but these petitions were denied. Since Edo was the seat of government, officials were concerned with appearance more than they seem to have been in Osaka, but also with the impeding of the narrow streets and the problem of smell. One particularly innovative petition in 1789 requested permission to set out soy sauce and sake barrels, which would be less of an eyesore than urinals. The petitioner argued that allowing the collection of urine from passersby would add to the quantity of fertilizer, thereby reducing the price of other fertilizers, enabling farmers to use more fertilizer and produce larger harvests, and ultimately lower the price level in general.21 In the Kanto as well as the Kansai, as Anne Walthall has documented, night soil was a precious commodity over which farmers fought for rights and banded together for effectiveness. Here too there were 20. Ito, "Edo ni okeru gomi," p. 449. 21. Ibid., pp. 4 4 5 - 4 6 .

U R B A N SANITATION A N D PHYSICAL WELL-BEING

115

conflicts over the collection and distribution of night soil and over its pricing. And here too, "by the late nineteenth century, the demand for access to nightsoil meant that peasants sometimes tried to outbid each other in promising fees and presents to landlords in exchange for the right to empty privies."22 There are numerous stories of travelers trying to make money from their own waste; they would carry a container with them and try to exchange their urine for food or money. Local residents could make money from passersby, as happened outside Kyoto when farmers realized that they could charge for providing suitable facilities for well-to-do women on their way to Arashiyama, a popular destination for outings.23 By the late 1850s, facilities for travelers to Edo were provided so as to obtain all the fertilizer farmers could from the people on the road, as Francis Hall remarks: "At frequent intervals, intervals so frequent as to be astonishing, casks are sunk along the side of the great highway, which are used as urinals. They are emptied of their contents so often that little or no effluvium from their contents is experienced by passers by."24 At the same time that the first regulations on night soil appeared, the disposal of other forms of waste was also specified. By 1655, the people of Edo were ordered to dispose of their garbage and rubbish on the island of Eitai in Edo Bay rather than just dump it in the rivers. During the decade from 1655 to 1665, disposal policies were gradually put into effect: collection points for refuse were established in each ward, transport was contracted to specific jobbers, and the wards were ordered to bear the costs. From the ward collection points, the rubbish was loaded onto boats and transported to Eitai Island. Although the original purpose of these measures was to keep the river channels open for commerce and traffic, they resulted in the establishment of an official dump for the city located outside city limits. These policies remained in effect for the next two centuries.25 The designation of Eitai Island as a dump eventually resulted in the creation of new land from the swampy ground in eastern Edo. Several other landfills also resulted in the creation of fields, so that disposing of Edo's wastes became a very profitable business. By the 1820s at least 22. Anne Walthall, "Village Networks: Sodai and the Sale of Edo Nightsoil," Monumenta Nipponica 43, no. 3 (autumn 1988): 296. 23. Rinoie Masafumi, Kawaya shiwa (Tokyo: Rokko Shuppansha, 1949), pp. 316-28. 24. Entry for Tuesday, November 29, 1859, published in Japan through American Eyes: The Journal of Francis Hall, Kanagawa and Yokohama, 1859-1866, edited and annotated by F. G. Notehelfer (Princeton: Princeton University Press, 1992), p. 82. 25. Ito, "Edo ni okeru gomi," pp. 431-36.

II6

URBAN SANITATION A N D PHYSICAL WELL-BEING

eighty contractors were involved in collecting the rubbish from Edo, encouraged in part by the creation of official organizations, the kabunakama. Now the townspeople had only to deposit their refuse at a collection site within each ward and pay for its collection and transportation for final disposal by a contractor. Ward fees were assessed by length of frontage and location of the property, so that, in effect, people paid a property tax. Whether renters were assessed depended on the ward. 26 Finally, there was the matter of drainage and disposal of waste water in Edo, which had a number of rivers and canal-like moats flowing through the city. From the number of regulations issued, the major problem seems to have been keeping the moats and river channels free from rubbish rather than providing for drainage itself. Detailed instructions were issued: people were ordered to construct drainage channels along the fronts of their houses, under the eaves. These ditches collected runoff from the streets and roofs, as well as people's waste water. The channels, about a foot wide and a couple feet deep, partially covered by stones to prevent people from falling in, are still seen in many cities, including the outskirts of modern Tokyo. Archaeological excavations in Tokyo reveal the fine network of drainage ditches even within the compounds of what would be considered crowded working-class housing at best and slums at worst.27 Edo and Osaka were not alone among Japanese cities in their emphasis on the proper disposal of waste material and the necessity to keep streets and waterways clean and open. Even the main streets in most castle towns were relatively narrow, about twenty-four feet wide, but they were "extremely well maintained and immaculately clean." The regulations regarding the maintenance of public roads were detailed and infractions reported. In Tottori, for example, streets had to be cleared and then sprayed with water (which probably reduced the incidence of respiratory disease). In Hirado, orders were issued to the effect that all bridges, gutters, and waterways should be repaired, maintained, and cleared; to make certain that this was done, officers of the town were to inspect them constantly. " N o corner shall be left uncleaned." Judged by regulations alone, cleanliness and proper sanitation were of high priority among Tokugawa urban administrators.28 26. See also Hayashi, "Kinsei ni okeru jinkai shori." 27. Koizumi Hiroshi, Edo 0 horu (Tokyo: Kashiwa Shobo, 1983), pp. 6 4 - 7 4 . 28. Murai Masuo, "Hoken-sei no seiritsu to toshi no sugata," in Seikatsu-shi, vol. 2, ed. Morimatsu Yoshiaki et al., which is vol. 16 of Taikei Nihon-shi sosho (Tokyo: Yamakawa Shuppansha, 1965), P- 128.

U R B A N SANITATION A N D PHYSICAL WELL-BEING

117

Drainage in a town street.

Metropolitan sanitation from the mid-seventeenth through the midnineteenth centuries was almost certainly better in Japan than in the West in terms of waste disposal and quality and quantity of the water supply, which resulted in a healthier environment for urban populations as measured by the size of the population and mortality rates. But Japanese city life was also more sanitary than that in the West because of customs concerning hygiene, food, and drink and because of a lack of domestic animals. Finally, government played a major role in setting and maintaining standards of sanitation in the cities. The foregoing makes evident an important difference between Japan and the West: human excrement was an economic good in Japan, much more than in the West. Since this waste was carefully collected for use as fertilizer, the water supply from source to urban outlet was far more protected from contamination that it was in the West, and people were prevented from coming into contact with waste matter while walking on the street or in other places where waste was dumped. In contrast,

n8

URBAN SANITATION A N D PHYSICAL WELL-BEING

Westerners traditionally relied on pits in the ground, such as cesspools, for the disposal of human wastes, and the danger of polluting the water supply was ever present. Even in the 1880s, Cambridge, England, was described as "an undrained, river-polluted cesspool city." In the nineteenth century, "Leicester was typical of many towns in the way it tackled the problems of excrement removal. At mid-century it had almost 3,000 uncovered cesspits, covering 1 - 1 / 4 acres." Only by the end of the century had this town managed to convert to a system of pails, which put an end to seepage into the subsoil.29 In addition, Westerners in some cities emptied chamber pots out the window onto the street, and all had to take care in crossing streets to avoid what horses left behind. Streets in parts of London had open sewers running down the middle of them as late as the early-eighteenth century.30 Streets in the "New World" were no better. In 1857 streets in New York were described as "one mass of reeking, disgusting filth, which in some places is piled to such a height as to render them almost impassable by vehicles."31 Laws put responsibility for cleaning the streets on the residents whose homes abutted them, but these were ignored, or dirt was swept into heaps and left for days, so that the dirt was rescattered. Paying scavengers did not solve the problem. One solution was to let hogs feed on refuse instead of having garbage collected. Visitors from Western Europe were disgusted, not only by the streets in New York but other cities as well, such as Louisville and Cincinnati.32 Although Pittsburgh passed an ordinance in 1807 imposing a fine on people who "cast any dead carcase, garbage, noxious liquor, or other offensive matter on any street, square, lane, or alley," no real concern was found for filth in the streets until the threat of cholera in the 1830s.33 In contrast, Western visitors to Edo, from the beginning of the Tokugawa period to the end, found its streets preferable to those in Europe. A visitor to Edo in 1609 found the main streets wide and long 29. Anthony S. Wohl, Endangered Lives: Public Health in Victorian Britain (Cambridge, Mass.: Harvard University Press 1983), pp. 73-74, 9530. M. Dorothy George, London Life in the Eighteenth Century (New York: Harper a n d R o w , 1 9 6 + ) , p . 67.

31. David W. Mitchell quoted in Lawrence H. Larsen, "Nineteenth-Century Street Sanitation: A Study of Filth and Frustration," Wisconsin Magazine of History 52, no. 3 (spring 1969): 239.

Larsen, "Nineteenth-Century Street Sanitation," pp. 2 4 2 - 4 3 . 33. John Duffy, "Hogs, Dogs, and Dirt: Public Health in Early Pittsburgh," Pennsylvania Magazine of History and Biography 87, no. 3 (July 1 9 6 3 ) : 2 9 7 . 32.

U R B A N SANITATION AND PHYSICAL WELL-BEING

119

and superior to those he saw in Spain. Moreover, they were clean. "One hardly thinks people even use them." In the nineteenth century, Alcock found the streets and roads in and around Edo carefully maintained and as good or better than in the West. He rarely had the experience of finding he couldn't walk straight through them because of refuse piled up, as was frequently the case in many parts of Asia and in Europe. 34 Much has been made of the English invention of the water closet as a solution to the problem of human waste in cities, but in the early years this system caused more problems than it solved. First, it required both a water supply and sewer system that could safely provide and remove the large quantities of water the system used. When the water closet was first invented, Londoners flushed their wastes into the Thames, thinking that at last they had gotten rid of a nasty problem in their houses. What they did not realize for decades was that the cause of the epidemics of infectious disease sweeping the city was the flushing of sewage into the upper Thames, since much of the city's water was taken from downstream. Furthermore, faulty drains caused sewer gases to waft up into homes, and people with fixed basins in their bedrooms often had to cover them with towels at night, a rather primitive method of coping with this problem.35 Nor were other countries immune: "As late as 1849, physician John H. Griscom described the unhealthy sanitary state created on Manhattan Island by 'these thirty thousand cesspools studding it up and down, and filling the atmosphere with nauseous gases.' " 3 6 "Stone comments that even after the introduction of water supply systems, conditions in cities and towns remained unsanitary until properly engineered sewers replaced cesspools, beginning about 1850." 37 Pinkney's assessment of 34. Ito Yoshiichi, Edo no Tumenoshima (Tokyo: Yoshikawa Kobunkan, 1982), p. 45. 35. Thomas McKeown, The Modern Rise of Population (New York: Academic Press, 1976), p. 125. 36. John H. Griscom, The Uses and Abuses of Air, 3d ed. (New York: Redfield, 1854), p. 183, as quoted in May N. Stone, "The Plumbing Paradox," Wintertbur Portfolio 14 (1979): 292. 37. Stone, "The Plumbing Paradox," p. 284. For good overviews, see Joel A. Tarr et al., "Water and Wastes: A Retrospective Assessment of Wastewater Technology in the United States, 1800-1932," Technology and Culture 25 (1984): 226-63; J o e ' A. Tarr, James McCurley, and Terry F. Yosie, "The Development and Impact of Urban Wastewater Technology: Changing Concepts of Water Quality Control, 1850-1930," in Pollution and Reform in American Cities, 1870-1930, ed. Marton V. Melosi (Austin: University of Texas Press, 1980), pp. 59-82.

no

U R B A N SANITATION A N D PHYSICAL WELL-BEING

Paris in the same year is that there was "one shockingly direct connection" between sewage disposal and the water supply. "The city drew part of its water supply from that main collector sewer, the Seine, and pumped it largely at points downstream from the mouths of sewers emptying into the river. Most of the remainder of the city's water supply came from sources little more inviting." 38 The Japanese system of waste disposal was not without its problems. When human excreta is used for fertilizer, there is always the danger of contamination—transmission of pathogens to the food supply and pollution of the water supply through runoff from fields or inadequate storage or transportation of the night soil. Although some human excreta was sold as fertilizer in the West, for the most part night soil was wasted, and this was particularly true in the largest urban areas after sewerage, or water carriage, systems had been installed. By the end of the nineteenth century, the West had advocates of sewage farming, but public health officials, as well as the general public, had a strong bias against using human waste for fertilizer. They "maintained that the raw sewage exposed farm employees to possible infection and that the vegetables grown on the farms could be the carriers of 'dangerous microbes or other parasites,' even though there was no clear evidence to this." 39 In fact, Asians did not dump raw night soil onto their fields but stored it at least a month, since it was common knowledge that direct application of raw excreta was dangerous.40 The combination of heat and time necessary to kill various pathogens varies. Also, scientists disagree or are uncertain as to the degree of contamination of water by an enteric virus that is needed to infect a community. There are numerous documented cases of outbreaks of disease in both Asia and the West through the use of night soil or sewage for fertilizer. Thus it is difficult to determine today how safe treatments of human excreta or manure were two hundred years ago. "The great majority of illnesses associated with sewage, however, appear to have been caused by application of raw or inadequately treated sewage wastewater, raw sludges, and night soil to crops which were consumed 38. David H. Pinkney, Napoleon III and the Rebuilding ofParis (Princeton: Princeton University Press, 1958), p. 21. 39. Joel A. Tarr, "From City to Farm: Urban Wastes and the American Farmer," Agricultural History 4-9 (1975): 610. 40. F. H. King, Farmers of Forty Centuries (1911; reprint, Emmaus: Rodale Press, 1973), pp. 193-215; Reginald Reynolds, Cleanliness and Godliness (Garden City, N.Y.: Doubleday and Co., 1946), p. 253.

U R B A N SANITATION A N D PHYSICAL WELL-BEING

121

raw." 41 And at least two studies indicate that "the predominant method of transmission of enteric viruses appears to be a direct fecal-to-oral route." 42 Though it is difficult to assess the sanitation and hygiene standards of the populace in premodern times, information exists for elites, making possible a comparative evaluation for various nations. Despite the fact that the sanitary conditions and the customs regarding personal hygiene of the elite cannot be considered representative of a society as a whole, they are indicative of what one would expect the highest standard to be. For example, in seventeenth- and eighteenth-century France, "the palaces of the Louvre, Vincennes or Fountainebleau in places became mere latrines." A house steward wrote: "In many places in the courtyard, in the upper passages, behind the doors and almost everywhere, one can see a thousand heaps of ordure, one can smell a thousand unbearable stenches caused by the necessities of nature which everybody discharges there daily." 43 In contrast, the Jesuit Joáo Rodrigues, who was in Japan from the late-sixteenth into the early-seventeenth century, noted that the Japanese provide their guests with very clean privies set apart in an unfrequented place far from the rooms. . . . The interior of the privies is kept extremely clean and a perfume-pan and new paper cut for use are placed there. The privy is always clean without any bad smell, for when the guests depart the man in charge cleans it out if necessary and strews clean sand so that place is left as if it had never been used. A ewer of clean water and other things needed for washing the hands are found nearby, for it is an invariable custom of both nobles and commoners to wash their hands every time after using the privy for their major and minor necessities.44 Two centuries later, conditions in France were no longer so primitive, but the problems with sewage and contamination of the water supply indicated that major sanitation problems remained. In mid-nineteenth century England, even royalty was not immune from the effects of inadequate sewage disposal, as in the example of Prince Albert mentioned earlier. In contrast, in Japan in the 1870s in the privies in the "better 41 • Wylie D. Burge and Paul B. Marsh, "Infectious Hazards of Landspreading Sewage Waste," Journal of Environmental Quality 7 (1978): 7. 42. Studies by G. Berg (1966) and J. W. Mosley (1972) cited in ibid., p. 3. 43- Frantz Funck-Brentano, The Old Regime in France (London: Edward Arnold and Co., 1929), p. 156. Nicholas de la Mare was house steward to the Comte de Vermandois. 44. Joao Rodrigues, S.J., as quoted in Michael Cooper, S.J., ed., They Came to Japan (Berkeley: University of California Press, 1965), p. 221.

122

U R B A N SANITATION AND PHYSICAL WELL-BEING

class of private homes," the American visitor Edward Morse found "less annoyance and infinitely less danger . . . than are experienced in many houses of the wealthy in our great cities." His description of the privies is similar to Rodrigues's in the sixteenth century, except that attendants did not clean out the receptacle after every use, but rather it "was emptied every few days by men who have their regular routes." Morse was taken not only by the cleanliness of Japanese toilets, but also by the amount of artful carpentrywork that decorated them in homes he visited. He was not describing the toilet facilities of either the wealthy elite or the poor but those in the houses he visited on his extensive stays in Japan.45 Naturally the urban working classes, farmers, and the poor had less elaborate, less clean, and often less sanitary toilet facilities, few of which are extant today. From the examples that do exist, we know that the general principles were similar, and farmers and those of low income had even greater incentive to collect a "by-product" that could be sold or traded for goods. In farmhouses, the toilets were usually at the entrance, in the doma, or in a separate building outside. If the family was well-off and had a zashiki or guest room, there might also be a toilet nearby for the use of guests. Generally, the toilet was placed to be readily accessible to those doing farmwork, so that they didn't have to remove footgear or go all the way into the house. In order to collect urine and feces separately, there were two toilets, one a rectangular hole over which the user squatted, and a urinal, sometimes constructed so women could use it too. Wood was the most common material used, though urinals were sometimes ceramic. In the countryside, toilets could be dangerous places. They were often a pit in a small shack, and the user squatted on two boards laid over the hole and used a stick for want of paper. Not only children but adults could fall in. In the urban slums and crowded tenements, each block would have a number of toilets for use by all of the residents. This kept the stink and unsanitary aspects of home life from the small rooms or apartments of people not located next to the toilets. However, a large number of families shared only a few toilets and these were usually made of wood, which is not impervious to fluids and hard to clean; often the compounds' communal well was near or even next to the toi-

45. The quotes are from Edward S. Morse, Japanese Homes and Their Surroundings (2d ed., 1887; reprint, Rudand, Vt.: Charles E. Tutde, 1972), pp. 228, 231. See also Wohl, Endangered Lives, p. 127; Edward S. Morse, "Latrines of the East," The American Architect and Building News 19, no. 899 (March 18, 1893): 170-74.

U R B A N SANITATION AND PHYSICAL WELL-BEING

123

Nineteenth-century toilets, the best and the worst: left, an urban indoor toilet; right, a crude rural privy from a poor area, housed in a rough outbuilding.

lets. In contrast to the descriptions and examples of toilets for the upper classes as clean and well designed, at the lower end of the economic scale they were smelly and scary places.46 Nevertheless, the fact that most bodily wastes were collected and used as fertilizers rather than stored in leaky cesspools or flushed into rivers led to higher levels of sanitation in Japan than in countries in which these wastes were not a good. The Japanese didn't just rely on economic incentives to ensure the proper collection of bodily wastes; to reinforce it, they adopted a number of customs and beliefs. Since the toilet was a dark and often unpleasant place, people needed a god to protect them there; at the same time, they needed the waste for fertilizer. Thus they asked the kawaya kami (toilet god) to help them produce a bountiful harvest. This god had to be properly treated like any other, and so at the end of the year, New Year's decorations were placed in the toilet, and it was the custom in some places for family members to sit on a straw mat in front of the toilet and perform a brief ceremony to the god, each family member eating a mouthful of rice to symbolize eating something left by the god.47 If a toilet is to have a god dedicated to it, be decorated, and have 46. Examples can be found in many old buildings open to the public, open air museums such as Hida no Takayama, and some private homes. For books on toilet facilities, see Yamada Koichi, Benjo no hanashi, Monogatari mono no kenchiku-shi (Tokyo: Kagoshima Shuppankai, 1986), and Agi Kaoru et al., Nihon toire hakubutsu-shi (Tokyo: INAX, 1990). 47- Yamada, Benjo no hanashi, p. 72.

124

URBAN SANITATION A N D PHYSICAL WELL-BEING

people eat in front of it, however ceremonially, it must be kept as clean as possible. The god of the toilet was considered to be extremely beautiful, and pregnant women who wanted to have beautiful children asked the toilet god to give their babies a "high nose" and if a girl, dimples. However, if women kept the toilet dirty, it was said that their children would be born ugly and unhappy.48 What better incentive could young brides be given to keep the family's toilets clean! These practices do not mean that contamination from human wastes did not occur in Japan but that it was probably less frequent than in the West. And a variety of customs helped prevent Japanese from becoming ill even when their water supply was not free from impurities. The Japanese customarily drank their water boiled, usually in some form of tea. This was a custom remarked on by foreign visitors to Japan from the sixteenth century on. Though people in other countries drank tea and other boiled beverages, the fact that they noted this preference of the Japanese suggests more Japanese avoided drinking contaminated water than many in the West where water was often the most common or preferred drink.49 The Japanese usually ate their food cooked, with the exception of Japanese pickles, preserved by fermentation and salt, so that even if night soil was improperly applied as a fertilizer, it was less likely to make everyone sick. Within the household, each member had a set of chopsticks, rice bowl, and teacup, which no one else used, so that it did not matter much that washing was perfunctory and in cold water. Food served outside the home was frequently finger food, and chopsticks used in restaurants were usually lacquered for easy cleaning and were not put in the mouth as were forks and spoons in the West. By the mid-nineteenth century, disposable chopsticks had come into use.50 48. Ibid., p. 75. 49- Customs varied from place to place and century to century. Tea was so expensive in Deerfield, Massachusetts, in the eighteenth century that people "floated on alcohol." Michael D. and Sophie D. Coe, "Mid-Eighteenth-Century Food and Drink on the Massachusetts Frontier," in Foodways in the Northeast, ed. Peter Benes and Jane M. Benes (Boston: Boston University, 1984), pp. 41-45. On the other hand, in the Old South, "water, as was natural in the heat of a southern summer, was the most popular drink." Joe Gray Taylor, Eating, Drinking, and Visiting in the South (Baton Rouge: Louisiana State University Press, 1982), p. 43. 50. Cooper, They Came to Japan, pp. 198-99; Michael Cooper, S.J., This Island of Japon (Tokyo: Kodansha International, 1973), p. 263; Engelbert Kaempfer, The History of Japan Together with a Description of the Kingdom of Siam, 1690-92, vol. 3 (Glasgow: J. MacLehose and Sons, 1906), pp. 238-40. Although raw fish became popular during the Tokugawa period, particularly in large cities, it was most frequendy partially preserved

U R B A N S A N I T A T I O N A N D PHYSICAL W E L L - B E I N G

125

Moreover, the Japanese had notions about what was "dirty" and "clean," many of which can be traced back to Shinto and its concepts of pollution. Though much of the pollution in Shinto is ritualistic, what is considered polluting and what is purifying are related to contamination and cleanliness. Anything to do with blood, death, and illness—such as childbearing, menstruation, contact with a sick or dying person, and funerals—is considered unclean, and people used to be prohibited from participating in religious rituals, mixing with other people, and even returning home from a funeral without being purified. Salt, water, and fire are all considered purifying agents and used both in religious ritual and to clean and purify. This emphasis on purification is not unique to Japan; it can be found throughout Asia. But it is particularly strong in Japan and has persisted to modern times. The strong avoidance of things dirty, although at one level superstitious, most certainly has prevented many cases of bacterial or viral contamination and impeded the spread of diseases and infections in Japan. At the everyday level, it resulted in the compulsory removal of footgear when stepping onto any kind of flooring raised from the ground and the washing of hands after using the toilet.51 As described earlier, frequent bathing was a custom that became widespread in the Tokugawa period, and regular laundering of clothing by the common people began when cotton began to be widely used. The new emphasis on cleanliness must certainly have had a salutary effect on hygiene and sanitation.52 Although customs relating to hygiene within the Japanese family depended on individual conformity to have effect, public sanitation also depended on government. A major reason that clean streets and an adequate water supply of high quality could be maintained was the high degree of control that existed over the populace during the Tokugawa period.53 Government control was enhanced by two factors. First, the raison d'être of the samurai class was to govern Japan, and explicit in and fermented and not slices of fresh raw fish as is most commonly served today. In any case, fish came from the sea and would not have been contaminated by fertilizers. For Tokugawa dietary habits, see Watanabe Minora, Nihon shoku seikatsu shi (Tokyo: Yoshikawa Kòbunkan, 1964), pp. 190-269. 51. Ishige Naomichi, "Jukyo to ju-seikatsu," in Nihonjin no seikatsu, ed. Umesao Tadao (Tokyo: Kenkyusha Shuppan, 1976), p. 29; H. Byron Earhart, Japanese Religion, 2d ed. (Encino: Dickenson Publishing, 1974), p. 7; Emiko Ohnuki-Tierney, Illness and Culture in Contemporary Japan (Cambridge: Cambridge University Press, 1984), p. 24. 52. Ochiai Shigeru, Arau fuzoku-shi (Tokyo: Miraisha, 1984), pp. 71-85. 53. John W. Hall, "Rule by Status in Tokugawa Japan," The Journal of Japanese Studies 1, no. 1 (1974): 39-49.

126

URBAN SANITATION A N D PHYSICAL WELL-BEING

the neo-Confucian philosophy the samurai adopted was the concept of rulers as benevolent, as responsible to the ruled, and as moral examples. Not only did the rulers expect such conduct from themselves, but there is evidence that the ruled expected it as well and let the rulers know it. Second, by the eighteenth century there were more samurai than were needed to govern, and overstating resulted in numerous detailed regulations and sufficient officials to see that these regulations were carried out.54 The effect of this benevolent but thoroughgoing government can be found in the governance of Edo. The city was divided into machi, which were village-sized units responsible for government at the local level. This enabled authorities to have tight control over the enormous population of the city. At large intersections in the city, the premodern equivalents of the police box were set up, not only to keep an eye out for criminal activities, but also to ensure that no water pipes were leaking and that the streets were kept clear. In addition, the city authorities made use of the outcasts who lived within the city. These people served to keep the streets free from dead animals, handle corpses, perform tasks ordinary residents would not touch, and report on anything suspicious they found on their rounds. Thus, in the late-nineteenth century, Morse could comment favorably on what were the slums of the new city of Tokyo, less than a decade from when it had been called Edo. "In Tokio one may find streets, or narrow alleys, lined with a continuous row of the cheapest shelter; and here dwell the poorest people. Though squalid and dirty as such places appear to the Japanese, they are immaculate in comparison with the unutterable filth and misery of similar quarters in nearly all the great cities of Christendom." 55 Finally, it is likely that the Japanese did not have to cope with the density in cities that Westerners did. Tenements in Japan were one story high, not six as they commonly were in Europe. The families of the daily laborers in Edo were often crammed into one-room apartments 54. Irwin Scheiner, "Benevolent Lords and Honorable Peasants: Rebellion and Peasant Consciousness in Tokugawa Japan," in Japanese Thought in the Tokugawa Period, ed. Tetsuo Najita and Irwin Scheiner (Chicago: University of Chicago Press, 1978), p. 46; Kozo Yamamura, A Study of Samurai Income and Entrepreneurship (Cambridge, Mass.: Harvard University Press, 1974), pp. 47-48. 55- Morse, Japanese Homes, pp. 5-6. See Tamura Eitaro, Chonin no seikatsu (Tokyo: Yuzankaku, 1966), pp. 194-97; Harada Tomohiko, Nihon hoken toshi kenkyu (Tokyo: Tokyo Daigaku Shuppankai, J957), pp. 438 ff. Naturally there were exceptions in Europe, such as the seventeenth-century Dutch who made a fetish out of cleanliness. See Simon Schama, The Embarrassment of Riches (New York: Alfred A. Knopf, 1987), pp. 375-79.

URBAN SANITATION A N D PHYSICAL WELL-BEING

127

approximately nine feet square with a small entry for storing tools and footgear and for cooking. They shared toilets and access to water with other tenants in the block. But however densely these families were packed in, they were not on top of each other in multistory buildings nor did people live in basements as Europeans did. The ever-present danger from earthquakes and the light construction precluded tall buildings, and the authorities forbade their construction by anyone who might be inclined to be so imprudent. 56 Statistical evidence for the assertion that sanitation was substantially better in Japan than in the West during the eighteenth and nineteenth centuries comes from estimates of life expectancy, which are analyzed in chapter 6. Not only was Japan able to provide better water and disposal systems, but Japanese customs led to better hygiene and sanitation than did Western. Thus Japan was able to maintain large urban populations from the sixteenth century on, both because sanitation was better and because the Japanese had the control necessary to carry out large-scale engineering projects, implement various systems connected with water supply and waste disposal, and see that measures were enforced. In fact, Japan was surprisingly free from the devastating effects of epidemics. The plague never reached Japan, and cholera came only in the nineteenth century when it spread throughout the world. Intestinal worms and enteric infections—those that enter through the mouth and are spread through contamination of food and water—tended to be localized and to appear in endemic rather than epidemic form. This situation is what would be expected in a society that used human wastes for fertilizer. It was true for Tokugawa Japan, which would explain the relatively high deathrates for children between the ages of one and four or five.57 Children after weaning were particularly susceptible to these 56. Robert Higgs and David Booth have found that "density effects on mortality were uniformly positive and statistically significant" in seventeen American cities in 1890. "Mortality Differentials within Large American Cities in 1890," Human Ecology 7 (i979): 353See also Tamura, Chonin no seikatsu, pp. 198-99; Nishikawa Koji, Nihon toshi-sbi kenkyii (Tokyo: Nihon Hoso Shuppan Kyokai, 1972), p. 250. George, London Life in the Eighteenth Century, particularly chapter 2, provides graphic descriptions of life in London. 57. Note that deathrates in this age group were high in the nineteenth century for children in American cities and in Odessa in the Ukraine, the cause of which was considered to be "summer diarrhea." Urban deathrates in Odessa fell to 25.4 per thousand in 1891 to 1893 after the construction of aqueducts and a citywide sewerage system. These deathrates after the improved systems are only slightly lower than urban deathrates for the city of Takayama in 1773 to 1871. See Patricia Herlihy, "Death in Odessa: A Study of Popu-

128

URBAN SANITATION A N D PHYSICAL WELL-BEING

diseases, but if they did not die early, they tended not to succumb to them. However, the fact that the cities with single water sources, such as Edo, Osaka, and Kyoto, did not experience rampant epidemics meant that the water supply must have been generally good.58 This chapter focuses on sanitation in the cities, but the same would hold true for the countryside. In Japanese farm villages, there was "no refuse problem." 59 Villagers either burned their refuse for fuel or used it for fertilizer, depending on what it was. Because animals and fowl were not widely raised for food and draft animals were at a minimum, human waste was an important source of fertilizer, since the only other way to obtain fertilizer by the mid- to late-Tokugawa period was to purchase it. There must have been some contamination of wells and other sources of the water supply, but the need to hoard every bit of potential fertilizer would have kept contamination to a minimum. Other customs regarding hygiene and sanitation prevailed in the village as well as the city, and there was the added benefit of villagers living in less densely crowded conditions than city dwellers did. What has obscured the realization that the level of sanitation was high in premodern Japan, particularly in the cities, is the fact that the system is backward by twentieth-century standards. In 1985, only 34 percent of Japanese communities had modern sewer systems, and, ironically, the residents of a town named Tamagawa Josui (Tama River Water Supply) were still without a sewer hookup.60 Even after World War II the Japanese continued to use night soil as a fertilizer and thus were seen as backward by Westerners who made no use of human excreta. But it was the very success of the premodern methods for dealing with night soil that made the Japanese slow to modernize their toilet and sewage systems. The Japanese did not have the imminent need to spend the vast sums necessary to install flush toilets and construct water-carriage sewage systems to remove the waste water. Indeed, the very success of the premodern waste disposal system inhibited modernization in this area, for despite the shortcomings in their public sanitation facilities, the Japanese today have the longest life expectancy of any major nation in the world. lation Movements in a Nineteenth-Century City," Journal of Urban History 4 , no. 4 (August 1978): 436-37, and Sasaki, "Hida-no-kuni Takayama no jinkó kenkyü," p. 106. 58. Ann Bowman Jannetta, Epidemics and Mortality in Early Modern Japan (Princeton: Princeton University Press, 1987). 59- Itó, Edo no Tumenoshima, pp. 14-15. 60. Susan Chira, "Most Japan Houses Still Lack Comforts of Those in the U.S.," The New fork Times, October 30, 1985, pp. 1, 41.

C H A P T E R SIX

Demographic Patterns and Well-Being

Demographic statistics are a reflection of more than population trends; they can also reveal much about the standard of living and the level of physical well-being of a society. The lower the deathrate and the longer the life expectancy, the healthier the population, and health is the most important determinant of the level of physical well-being. How many children people have, how or whether they attempt to control their family size, and other aspects of demographic behavior, together with other evidence, indicate much about the standard of living within that society. These kinds of demographic information are available for both eighteenth- and nineteenth-century Japan, though much of the evidence is partial and limited to scattered parts of the country. However, the evidence that does exist, particularly when compared to that for other parts of the world, indicates that the premodern Japanese were trying to maintain or improve their standard of living and that their level of physical well-being did not lag substantially, if at all, behind that of Western nations in the nineteenth century. In a study of material culture, it is impossible to detail all of the information available on the population of the Tokugawa period. Over the past several decades, numerous studies have been made of the village, temple, regional, and national data available. The most numerous and detailed of these studies are by the pioneer in Japanese historical demography, Hayami Akira, and most of the rest of us who have analyzed the Tokugawa population are his students. My own earlier re129

D E M O G R A P H I C PATTERNS A N D WELL BEING

search was on the historical demography of Tokugawa Japan, and here I summarize my own findings and those of others as they relate to the question of how high the level of physical well-being was during this period.1 In addition to detailed local information, we know the general trend of the national population in the Tokugawa period. In the seventeenth century, it grew from somewhere between fifteen and eighteen million at the establishment of the Tokugawa Bakufu to over twenty-six million commoners by 1721, which is very rapid growth by premodern standards from either starting figure. In the eighteenth century, growth leveled off. Eleven national surveys taken between 1721 and 1846 indicate that the population held steady at the 1721 level, fluctuating around the twenty-six million mark. However, at about the time of the last of these surveys in the mid-nineteenth century, the population again began to grow.2 The standard interpretation for this trend is that the population grew rapidly during the century of peace and economic growth following unification and then leveled off because of a series of major famines and stagnation in the economy. Why it began to rise again in the nineteenth century is not usually addressed. Too much emphasis should not be placed on the national figures. The various kuni or provinces showed diverse trends within Japan: some areas, particularly in western Japan, showed significant growth from 1721 to 1846, while others, particularly in the northeast, seemed to decline. Sekiyama Naotaro aggregated the provincial figures into ten regions, and the results show a very different picture from the "stagnation" that an earlier generation of scholars found.3 (See table 1.) Seven of the ten regions showed a pattern of growth from the eighteenth into the mid-nineteenth century. The three exceptions were the Kanto region, which contained the metropolis of Edo, the Kinki region, in which the cities of Kyoto and Osaka were located, and Tohoku, the entire northeast. Premodern metropolises are known for their negative natural increase, caused by a shortfall of births over deaths, and thus to maintain their populations, migration from surrounding areas was necessary. 1. Susan B. Hanley and Kozo Yamamura, Economic and Demographic Change in Preindustrial Japan, 1600-1868 (Princeton: Princeton University Press, 1977). 2. Akira Hayami, "Population Changes," in Japan in Transition, ed. Marius B. Jansen and Gilbert Rozman (Princeton: Princeton University Press, 1986), pp. 280-317. For Bakufu survey data, see Sekiyama Naotaro, Kinsei Nihon nojinko kozo (Tokyo: Yoshikawa Kobunkan, 1958), pp. 137-39. 3. Sekiyama, Kinsei Nihon, pp. 140-41.

DEMOGRAPHIC PATTERNS AND WELL-BEING

131

The Tohoku region is the only nonurban region whose population seems not to have grown during the eighteenth and early-nineteenth centuries. It is known to have suffered more from crop failure than much of the rest of the country, but the Bakufu survey data for this region are suspect. Whereas the population in 1846 was just under 89 percent of what it had been in 1721, there was a 38 percent jump in the population of Tohoku during the twenty-six years between the last Bakufu survey and the first survey of the new Meiji government in 1872. Thus it would be better to conclude that instead of "stagnation," most of Japan underwent a slowing in its population growth rate during the eighteenth and nineteenth centuries. What caused this slowing and what were the implications in terms of the well-being of the population? Village studies from a wide variety of samples around the country suggest that a major reason for a leveling off of growth in much of the country was deliberate population control that had the effect of maintaining or raising the level of physical well-being. This can be seen by an examination of the birth- and deathrates, life expectancy estimates, and fertility patterns for various samples that exist for this period. Crude birth- and deathrates reveal that for a surprising number of samples, both followed a very similar trend. In villages, the birthrate was apt to be just slightly higher than the deathrate, whereas in the city of Takayama, for example, the reverse often held true.4 Had there been no migration, these balances would have led to slow but positive growth in rural Japan and a gradual decline in the urban population. Neither trend would have been striking enough for the residents to notice except in the long run, and certainly in the cities in-migration covered any shortfall in births in most places. What is striking is not just the overall balance of birth- and deathrates for most samples, but the fact that birth- and deathrates were both relatively low. The birthrates were not characteristic of the rates found in much of the non-Western world during the post-World War II era, when many countries experienced birthrates of well above forty per thousand.5 Nor were the deathrates high in most places, with the exception of areas and years in which there were clearly mortality crises, such as during the early 1840s following the years of bad harvest in the 4 . See Hanley and Yamamura, Economic and Demographic Change, especially chapters 8 and 11. 5. U.S. Bureau of the Census, World Population: 1977, Recent Demographic Estimates for the Countries and Regions of the World (Washington: U.S. Department of Commerce, 1978).

In i-H tp TP CO so m o m S co in 00 tN " I IN < 5 m Os, Os, in o. in. 00, "t- in in co" in oo" Os" in" so" so" es" o" 00 IN 00 CO Os 00 Os T P , IN T p , o. C so. N, 0> 00, Os rf in" co rf CO rf rf T j f co" CO M O SO t N T P Os sO in in 00 CO tN - O In C Tp O s O s T P T P cs Tp, CS tp. CO ^ tP, Os t N , O . SO, oo" t N so" < N oo" o" tp" ^ tt NN co" crf Tp co SO O sO oo CO CS In C O CO T*< in. t N , in. 00 •t. Os, T P , Os, co" so" Tp" n" n" ^ cs" rf CS n In TP 00 00 -t Os co CN O 00 CO Os O o CN T P Os os ir; 00 In, co. C co. Os, T P , in^ 00, sO 00, i-H C S I N o" co" TP" n" Os" co" o" V tN TP sO tN IN CS TP SO ^ CO so. 00, so. CO TP, CO TP, O, tP, Os, i 1 co" In" tP C N TP" rf ^ rf Os rf — CN tN TP O 00 S 00 O •>t C os tN tN T P O m cs C S ^ C S 00, O, T p , NJ