Ikki: Social Conflict and Political Protest in Early Modern Japan

By James W. White

The reign of the Tokugawa shoguns was a time of statebuilding and cultural transformation, but it was also a period of ikki: peasant rebellion. James W. White reconstructs the pattern of social conflict in early modern Japan, both among common people and between the populace and the government. Ikki is the first book to cover popular protest in all regions of Japan and to encompass nearly three centuries of history, from the beginnings of the Tokugawa shogunate in the 1590s to the Meiji restoration.

White applies contemporary sociological theory to evidence previously unavailable in English. He draws on the long historical record of peasant uprisings, using narrative interpretation and sophisticated quantitative analysis. By linking the texture of conflict to the political and economic regime the shoguns created, he casts doubt on competing interpretations of a contained, orderly society.

• Language: English
• Publisher: Cornell University Press
• Release date: Mar 24, 2016
• ISBN: 9781501704581

Book preview

INTRODUCTION

In the early nineteenth century the village of Higashi Zenyōji lay in the Maebashi district of Kawagoe domain, overseen from a distance by a county magistrate (kōri bugyō) and a smattering of samurai administrators located in the domain’s castle town.¹ Though it was sited in a fertile region near the Tone River, Higashi Zenyōji had fallen upon difficult times. The whole Kantō region, around present-day Tokyo, was undergoing rapid economic change. People were moving out of farming into commercial and small-scale industrial activities and out of the villages into the bustling and more prosperous towns. Those who remained in the village saw their fortunes declining, and the exactions of village officials and incursions of landlords and moneylenders drove even more people from the villages.

Kawagoe domain was sensitive to these problems, and unsurprisingly so: its revenues depended upon the annual rice tax, and the decline of the farm economy struck directly at the administration’s bottom line.² Thus in 1820 the domain established the Agricultural Promotion Office to subsidize farming, hire agricultural workers, attract farm families from outside the district, and supervise the cultivation of disused land. Many of its activities were actually carried out by local landlords and officials, one of whom was Hachiemon of Higashi Zenyōji.

Hachiemon had led a rather checkered life. In 1784, when he was only eighteen, the death of his father and the collapse of family businesses forced him to wed and set up his own household. He lived through the crop failures and famine of the 1780s and the peasant protests that accompanied them, including one riot in a village near his own, and in 1791 he left for a stint of work in Ise, in south-central Japan. Returning to Higashi Zenyōji, he served as a village official for over a decade. By 1808, however, he was bankrupt; so he left rather precipitately for the capital city, Edo, where he earned enough to pay off his creditors. By 1819 he had risen to the position of village headman. His record was hardly unblemished: he had done a bit of time in jail, had spent 170 days in manacles for another offense, and was reprimanded by the authorities at least once. Nevertheless, 1821 found him recruiting settlers on behalf of Kawagoe domain.

It was not a good year. Drought, wind, and frost damaged the crops—especially the mulberries on which sericulture depended—and the domain, hit with a state imposition for coastal defense, decreed a tax increase: village officials were to survey and assess the crops, subject to a check by domain officials. Hachiemon sent a request to the domain for a temporary exemption because of crop conditions, but it was denied, and when he and his colleagues submitted their crop surveys the domain raised the assessments, amounting in the case of Higashi Zenyōji to a two-thirds increase in the tax bill. At the behest of his villagers Hachiemon appealed again—not against the policy or the principle of the tax but simply for a one-year extension of the old system. He received no response.

At this point the farmers began to get together. Two subheadmen of groups of farmers within the village came to see Hachiemon, and seven other villages met to plan a collective plea, though such a meeting could be officially labeled a conspiracy (totō). The farmers sent representatives to the Agricultural Promotion Office. The head of the office agreed to pass on their petition but told them it should come only from the poor of the villages. The farmers were undeterred; indeed, the core of the movement was not the poor but those of modest means whose economic viability was threatened by the tax increase. In the eleventh month a group left the Maebashi district to lodge a protest directly with the lord of the domain in Edo.

Although representatives of Higashi Zenyōji were not part of this group, Hachiemon wanted both to aid the farmers and to avoid an overt and illegal appeal. So he and two other headmen pursued and caught up with the Edo-bound group and persuaded them to hold off while the village officials tried one more time. Domain officials then solicited a petition from the people, ostensibly cooperating, although their primary motive might well have been the fear that an open appeal in Edo would cost them all their jobs. When the petition was in hand, however, and the people dispersed to their respective villages again, the domain officials rejected and confiscated it, accused the seven villages of conspiracy, illegal flight, and illegal appeal, and launched an investigation.

In the first month of 1822 Hachiemon and fourteen others were arrested. The investigation—a somewhat euphemistic term for a process based largely on torture—lasted almost the entire year, and one alleged leader died in the course of it. In the end Hachiemon alone was sentenced to life in prison, where he died in 1830 at the age of sixty-four.

The 1830s were a time of dire hardship and famine in much of eastern Japan, and the Gunnai region of Kai province, just west of Edo, was no exception. A mountainous area that imported rice and depended for its well-being on wheat, tobacco, sericulture, and weaving, Gunnai was economically and commercially innovative by necessity and economically unstable by nature. It was also doubly hit by hard times: its own poor crops cut its income and its usual suppliers of rice raised prices and cut shipments. The inequality of supply and the merchants’ right to profit were not issues; what was unacceptable was the hoarders’ unwillingness to share.

The people at first responded legally, with requests for aid from the intendant (since Gunnai was shogunal territory) and demands that local merchants sell rice at reasonable prices. When these attempts failed, the people began to organize, largely under the guidance of local officials, one of whom was Hyōsuke of Inume village. Hyōsuke was born in 1797 into a traditionally elite family and was the head (kumigashira) of an intravillage group of families. Amid an air of simmering incipient conflict, he was one of the farmers who drew up a list of rules to govern the behavior of those participating, prepared flags for village groups, and obtained some swords. He was also shrewd enough to foresee the consequences of the growing popular movement: he divorced his wife, settled his debts, and arranged for his family succession before the people took action.

Action began in the eighth month with assaults on the shops of several rice-hoarding merchants in Gunnai; on the twenty-first there was more desultory destruction, and a paper circulated calling for the people to assemble and march on the Kuninaka Basin to demand price reductions and loans of rice. The response was immediate, and a movement of several thousand small and large farmers and their village officials moved down from Gunnai. The movement was not theirs for long, however: no sooner did they hit the village of Kumayado than their ranks were swelled—eventually to perhaps thirty thousand or more—by an influx of the local poor, vagabonds, and others. Discipline was lost; forced sale and loan of rice gave way to simple destruction of buildings, food, and pawn and land records; and the village basis of organization evaporated.

At this point Hyōsuke and his colleagues—and indeed most of the Gunnai people—withdrew, leaving the field to a rampaging wave of rioters, who in the last week of the month destroyed over three hundred shops and storehouses throughout Kuninaka and the provincial capital of Kōfu. This violence, though disorganized and unfocused, was not indiscriminate: the targets were rice merchants, moneylenders, landlords, and pawnbrokers. The intendants were helpless, their stations simply bypassed; with no armed force of their own, they could only appeal to neighboring lordly domains to send soldiers. When the lords responded, the rioting was quelled. Arrests followed, and eventually some three hundred participants were punished. Of those punished, only 10 percent were from Gunnai, and Hyōsuke was not among them, although a death penalty was imposed on him in absentia.

In some ways the most noteworthy stage of Hyōsuke’s activities began here. In the ninth month he fled Gunnai and spent the next several months in a peripatetic flight that belies the conventional notion of nineteenth-century Japan as a closely governed society. He crossed the mountains to the shore of the Japan Sea, then headed west to Tango province and back over to Bizen province on the Inland Sea, crossed to the island of Shikoku, traveled west through Shikoku, and then crossed back to Aki province on Honshū. He continued westward to Suō province at the end of Honshū and then turned eastward again, going through Ōsaka, southward into Yamato province, northward to the city of Kyoto, and finally to Ise province near Nagoya, where his diary ends in the eighth month of 1837. Family records suggest that he then made his way to the Edo area and eventually returned to Gunnai sometime around 1845 (Fukaya 1978). He appears to have died peacefully in 1867 at the age of seventy-one.

Hyōsuke’s skill in avoiding the executioner commands our respect. It is equally striking that he moved with almost complete freedom around Japan for at least a year, supporting himself largely with such skills as abacus, mathematics, and divination. Clearly there was a demand for Hyōsuke’s skills in almost every village, and a clientele able to pay for them, sufficient to keep him in food, clothing, and lodging. The poor of Kuninaka, who snatched the Gunnai protest out of the hands of Hyōsuke and his neighbors, were obviously a major sector of society, but not all of society.

Bansuke, of Yonekawa village in the Nanzan district of Shinano province, was a reprobate. His family, once aristocratic, had long since fallen into the ranks of the common people. Bansuke’s generation had also slid from the status of village elder, and the expulsion from the inner circle apparently rankled Bansuke. His elder brother achieved a measure of success, but in 1838 (when Bansuke was forty-two) the family was bankrupt and moved from the village. In 1850 they returned, landless, poor, and possibly unwelcome. Obstreperous and hostile to village rules, Bansuke had been arrested some half dozen times for gambling and assorted offenses. In 1835 he had instigated a dispute over the collection of local taxes. The dispute had been resolved through negotiations, but Bansuke disagreed with the outcome, and it was not until the local intendant leaned on him, at the behest of his fellow villagers, that he had acquiesced.

When Bansuke returned to Yonekawa in 1850 he found a situation much to his taste. The people of mountainous Nanzan had long enjoyed a favored administrative status, being located in relatively lightly ruled (the intendant’s office was distant and understaffed) and undertaxed shogunal territory. They also benefited from being able to pay their taxes in coin, exchanged for the official assessment in rice at a locally set rate. But in 1843 it had been announced that the district was going to be assigned to neighboring Iida domain. The locals were perfectly aware that Iida taxes were heavier and paid in kind, and that the corvée was heavier as well. Village officials appealed both to Iida domain and to the local intendant for continuation of the previous tax system. Failing in both attempts, they appealed directly to one of the shogun’s elders (rōjū) in Edo. This appeal failed also, but in 1845 the district had again become shogunal territory.

Relief was short-lived. In 1846 Nanzan was assigned to Shirakawa domain. Again local officials appealed, and Shirakawa agreed to allow payment in coin until 1855, although at a new exchange rate that amounted to a tax hike. By 1852 the farmers were organizing, and in 1855 a petition went to the domain from village leaders and better-off farmers. The appeal was rejected and a number of the petitioners were chained or jailed; retraction and apology were demanded; and the district was hit collectively with an additional exaction. The domain also postponed in-kind payment until 1859, but the people regarded this outcome as a failure and continued to plan. Since officials’ overtures didn’t seem to be working, attention began to move down the social ladder, and it wasn’t long before eyes fell on Bansuke, if only as a fall guy.

Thenceforth, activities ran on two tracks: higher-level elements considered various appeals, and smaller farmers and tenants kept local feelings fermenting, partly by circulating the tale of Sōgorō, a legendary protest leader in days gone by. In the twelfth month of 1859 village representatives went to the market town of Haramachi to protest the exchange rate. The intendant agreed to listen, but the poorer farmers pushed the representatives to do more. More people gathered in Haramachi, and on the twenty-fourth, some one hundred of them called on the magistrate in Ichida with a further petition. His response was an abusive rejection. On the twenty-seventh most of the petitioners returned home, but some one thousand—mostly tenants and poorer farmers, including Bansuke (if he knew he was being used, he didn’t seem to mind)—regrouped. By this time village officials, fearful of an illegal appeal that would expose them to the wrath of the domain, were trying hard to control events.

Angered upon learning of the rejection of the petition at Ichida, the crowd headed off to present its own demands to the domain, now not only calling for restoration of shogunal rule but also denouncing domain cuts in irrigation and ferry subsidies, double standards in administration, punishment of previous protesters, unresponsiveness, and the appointment of oppressive officials. This kind of action was serious stuff, as a party of Iida domain officials the petitioners encountered in Chayamachi informed them: they were in conspiracy to appeal illegally and should return home immediately. The petitioners asked the officials to transmit the petition to Shirakawa domain. When the domain officials refused, the credibility of the village officials cracked, and Bansuke and his poorer allies took over.

On the twenty-ninth Bansuke’s group met with officials of Shirakawa domain. Harsh words were exchanged, and Bansuke, true to form, tried to lay hands on the magistrate. Surprisingly (or perhaps unsurprisingly, for official prestige had come to a sorry pass by the 1860s), Bansuke got away with this assault and even came away with an oral agreement to fix the tax situation as under shogunal rule. The domain’s word was hardly trustworthy, and although the protesters withdrew, they followed up in 1860 with many petitions. But the domain did as it pledged, giving in on the tax issue, replacing the magistrate, and further aiding Nanzan with cash.

Bansuke was not appeased. Perhaps he was simply angry at being eclipsed (once the protest was over, leadership reverted to the village officials), or perhaps he really did feel that the aid was insufficient. In any case, 1860 found him stirring up the smallholders, and the new magistrate let the village headmen know that this agitation had better stop. Bansuke was told that his activities jeopardized the district’s gains. Thus recognized (and paid off), Bansuke cooled his followers, but in the spring he was briefly jailed anyway.

By autumn Bansuke was at it again, arguing that the tenants and smallholders had been overpaying their taxes for several years. In the tenth month a tenant gathering brought the division between tenants and landlords out in the open. The plan was to present the landowners with a set of demands, but domain officials, alerted by the landowners, arrived and arrested Bansuke and four others. The officials patrolled the villages, and the tenants caved in. The landed had won, the landless lost, and Bansuke got life in prison for his efforts. His eventual fate is unknown, but he was transferred to domiciliary confinement back in Yonekawa in 1863, at the age of sixty-eight, and freed in 1870 (Yasumaru 1975: 206).

Popular Contention in Early Modern Japan

The three stories just recounted are parts of a richly variegated text of conflict that runs throughout the early modern history of Japan.³ This period (1600–1868), during which Japan was ruled by the shoguns of the Tokugawa family, was one of state building and economic development, social change and cultural transformation, and it is hardly surprising that patterns of conflict among the common people, and between the people and those who ruled them, are closely intertwined with these phenomena. The intimate connection between text and context is exemplified by the generic term for popular contention during the era: ikki, usually translated as peasant rebellion and often defined explicitly in terms of the Tokugawa political and social system. Such contention, it can be argued, had a powerful effect on the policies, personnel, and institutions of Tokugawa-era government. Of revolution we see nothing, but the cumulative effect of the sorts of political concessions seen in the preceding stories proved decisive for the viability of the Tokugawa state. Furthermore, the prevalence of conflict belies the old stereotype of a uniquely harmonious Japanese society. Indeed, although contention waxed and waned, varied from place to place, took multiple forms and metamorphosed over time, and was throughout intimately enmeshed in contextual changes, it appears to have been almost ubiquitous in the Tokugawa era.

This book describes types and repertoires of Tokugawa-era contention, the rise and fall of great waves of conflict and protest, the actors in the drama and their targets, their organizations and motivations and tactics, and the magnitude and intensity of their efforts. It maps the distribution of conflict across counties, provinces, and domains, and across the early modern era, tracing the evolution of forms, magnitudes, and other dimensions of contention from before the Tokugawa era began until a decade after it gave way to the modern era.

Though the focus of the book is narrow, my investigation may cast light on the nature, causes, and meanings of popular contention in general. The sorts of popular contention represented by ikki are in many respects neither uniquely Japanese nor uniquely early modern. Similar phenomena have occurred and do occur in other societies and at other times. In fact, the prevalence of this sort of behavior prompts general questions: Is contentious behavior rooted in human reason or emotion, in social relations, in the actions of rulers or exploiters, or in the workings of the system of national states and economies around the globe? What does the level of conflict—low or high—say about the quality of life in a society? Is conflict a cultural interloper and solvent, or does it reflect and perhaps even cement dominant cultural patetterns? In short, what is the nature, cause, and meaning of popular contention?

Such general questions lead us back to the specific: Why this? Why them? Why here? Why now? and How? This book explores these questions by considering the structural and cultural context of Tokugawa society; the conjuncture of weather and events, man-made and natural incursions on popular livelihood; the attitudes and ideologies of rulers and ruled; and the nature of leadership in Tokugawa society. From these factors it identifies the reasons why people contended either with each other or with the authorities (their grievances, frustrations, selfless ideals and selfish interests, goals, indignations), the means of contention at their disposal (weapons, symbols, numbers, leaders, organizations, and other resources), and the opportunities for contention available to them under the status quo.

The combination of reasons, means, and opportunities explains both the quantity and quality of popular contention, although I am also concerned with which of the three is most important in which situations. Much of my attention is directed toward the forms contention took, the participants, and others of its aspects. Many previous studies have described the relationships between, for example, the behavior of state officials and frequency of political protest, and such studies provide us with expectations about the sorts of relationships likely to be found in Tokugawa Japan. I test these expectations against the Tokugawa experience.

Another sort of explanation is also called for: explanation of the consequences of conflict for Tokugawa society. Popular contention had implications for those who participated (on both sides), for the structures, policies, and personnel of government, for the subsequent course of Japanese history, and for the growth of theory in the social sciences. By looking at the implications of contention for its setting and by evaluating the interpretations offered by others, this book assesses the meaning of popular contention.

A Model of Popular Contention

This excursus takes us farther than I like from individuals such as Hachiemon, Hyōsuke, and Bansuke, but it is an essential step in making them larger—or at least more exemplary—than life. They are interesting figures in their own right, but they are also data. And data, by themselves, tell us little. To become the building blocks of understanding they must be set in context and bound together by theory.

Figure 1 sets forth the theoretical framework of this book. I do not actually test the model here; rather, I maintain that this model does in fact represent the causal process behind popular contention in early modern Japan. Indeed, I could not test the entire model at one time even if I wished, for quantitative data are lacking for some parts of it—notably interests and individual and collective decisional calculi. In these cases, we must infer what is going on. Interests are expressed in popular statements of intention and justification, but they are only stated reconstructions (sometimes with tendentious purposes) of intention and justification and not necessarily the original attitudes at all. I try to bridge these inferential gaps from both sides.

I have already noted the relevance of the institutional and structural context to popular contention; indeed, many analyses of contention remain on this level, associating overall levels of conflict in societies with the macrolevel characteristics of those societies.⁴ I am more interested in context as a determinant of interests, opportunities, and resources, but the indirect influence of context on contention is of great help in inferring the nature of this determination.

As noted in Figure 1, context can be viewed analytically as having political, economic, social, and ideological or cultural dimensions. The primary aspect of the political context is the state, and the early modern Japanese state occupied an inherently contentious position vis-à-vis the people.⁵ The people constantly attempted to broaden the scope of their freedom of action and minimize their obligations, while the state attempted to regulate them more effectively and provide less to them in return. Moreover, in the early years the scope of state activity grew dramatically, provoking repeated confrontations both with the common people and with their intermediate overlords. Nevertheless, to the extent that the state was willing to repress them consistently it was able to reduce both popular and aristocratic resistance.⁶

Figure 1. A model of popular contention

After a short time the state ceased to grow (for reasons I shall discuss), but popular conflict did not diminish, because over time the capacities and consistency of the Tokugawa state decayed, but its aspirations did not.⁷ The consequence was a condition of overreach. That is, the government chronically attempted to do more than it was capable of, and people such as Bansuke, aware of this decay of both will and power, could see their opportunities for contention expanding.

Economically, the Tokugawa context was agrarian, although the agrarian economy was undergoing significant growth, commercialization, and protoindustrialization.⁸ Scholars have documented a positive relationship between the early stages of economic growth and sociopolitical conflict.⁹ In early modern Japan, too, this was the case, although I maintain that the path of causation ran not through the socially traumatic effects of rapid change or the immiseration of development’s losers but rather through the proliferation of interests resulting from social differentiation and through the attempts of better-off commoners to abdicate their historical role as patrons and protectors of the less well-off.

The political economy, too, played a role. Its most significant aspect in early modern Japan was state extraction. Higher rates and new media of taxation always antagonized the people, but over the long run, extraction contributed to state capabilities. Consequently, I see a positive macrolevel relationship—driven on the individual level by opportunities, not grievances—between declining taxation and contention. As for the social economy, the preponderance of the data—including the prosperity of those who supported Hyōsuke in his travels—indicates that the lives of the Japanese people improved during the Tokugawa era (White 1989). Because of the vagaries of climate and (increasingly) the market, however, their vulnerability did not diminish. In consequence, the people did not become fat and contented as living standards improved but remained anxious and extremely protective of whatever gains they had made.

With regard to contention, the most important aspect of the social context in early modern Japan was neither a sweeping abstraction such as class nor a micro-unit such as the family but rather the community, the primary locus of loyalty for most of the common people.¹⁰ Community norms demanded contention under some circumstances—especially by community leaders such as Hachiemon and Hyōsuke—and demanded respect for collective interests, which, if ignored by the better-off, were seen as grounds for contention by Bansuke and his ilk. Communities possessed collective interests that crosscut class; to the extent that communities remained cohesive—and many did, even amid rapid economic growth—their capacity for contention was enhanced. And the Tokugawa state, which removed the samurai from the land and left local administration in the hands of commoners, provided the people with extraordinary opportunities to plot away from official eyes (see also Skocpol 1979).

Yet, despite the overarching influences of government, economy, and society, popular contention is the result of individual decisions, and it is only through individual perceptions and calculations that context begins to matter. We cannot often measure popular consciousness directly, although we do have contemporary diaries, manifestoes, official analyses, and the observations of scholars then and now.¹¹ I have tried to characterize these attitudes broadly; among those of greatest relevance are popular evaluations of the propriety and efficacy of contention per se and of rights, reciprocity, and the proper nature of authority.

Along with these, I believe, the aspects of popular consciousness most relevant to contention are two. First, the common people were never mystified, or persuaded by the elites that uncomplaining subordination was right, proper, and eternally ordained. They saw through the ideology of domination and on occasion turned it against the elites themselves.¹² Second, the people were (though not narrowly rational utility maximizers) able to reason and to calculate the costs and benefits of their own behavior and also acutely moralistic in their judgments and capable of a high degree of indignation and passion, as the grimly calculating Hyōsuke and irrepressible Bansuke both suggest. Whereas rational thinking was common in the economic sphere (see T. Smith 1977), people tended to see community and polity in highly moral terms of rights and justice, contract and reciprocity. Rationality and morality, moreover, were sequenced (March and Olsen 1989, esp. chap. 3). Such ideals as morality and justice established limits of acceptability on elite activity and propriety of popular response, and it was within this framework that individuals calculated the costs and benefits of their action.

I am suggesting, it should be clear, that the different aspects of context are interrelated. In Tokugawa Japan the economic and social systems were the direct result of political decisions and were integrated with the polity; the long-term protoindustrialization and short-term fluctuations of the economic system and the consequent stratification, differentiation, and dislocation of the social system had far-reaching implications for the government. A social factor—socioeconomic differentiation—progressed faster in the cities than in the villages, and the implied proliferation of potentially conflicting interests, combined with the economic factor of market vulnerability and the resource of dense settlement and communication, seems to have outweighed the heavier urban administrative hand and the socially disintegrative effects of stratification, producing higher levels of contention in the city than in the countryside.

The ideological-cultural context was to some extent a pre-Tokugawa given, but the Tokugawa adopted certain ideological and philosophical principles designed to legitimate their rule. As with every ruling ideology, these principles were a two-edged sword (Kertzer 1988). Elites used them to rationalize their actions, but they were also a popular resource when government did not live up to them. And like the economy and society, the milieu of consciousness slipped out of the hands of the elites as the Tokugawa era drew on, until the polity was more at odds than in synchronization with all three.

On the level of interests, opportunities, and resources, the model moves from the outside world to individuals’ and groups’ perceptions of it. Indeed, much of my emphasis on the outside world derives from my view that the social institutions within which individuals lead their lives involve roles that entail certain expectations (March and Olsen 1989), and these roles and role expectations generate individual beliefs and behavior—behavior that, absent institutional considerations, can appear to fly in the face of individual self-interest, almost suicidally so in the case of Hachiemon and Hyōsuke. Individual beliefs or interests include anger, profit seeking and utility maximization, survival, institutionally imposed obligations and imperatives, and so forth.

Interests present a major problem in this book, inasmuch as they are individual characteristics, whereas the data are overwhelmingly aggregate. Many studies have successfully inferred interests from aggregate data (Gurr 1970; Hibbs 1973), but theirs were more extensive and reliable than those available for Tokugawa Japan. I have had to rely on relatively gross characteristics of individuals and groups for my inferences about their views on contention. Still, I was not without guideposts: several such characteristics have been found to be systematically related to interests, and I have approximated them here, especially when it was possible to check them against individual-level data.

The first is social class.¹³ Without reifying or homogenizing, one can see in Tokugawa society large, albeit highly internally variegated, groups who stood in more or less the same relationship to the major concentrations of wealth and power in society. They were in a common legal and economic position vis-à-vis the elites, and they were treated uniformly by the elites in ways that stimulated grievances or indignation. Consequently, common (albeit restricted) discontents and common reactions to economic or policy change can be inferred.

The second basis of inference is community. Community ways of life, autonomy, prosperity, and survival constituted common interests, from which I have inferred what sorts of government actions would be seen as threats. I have inferred interests not just from the relationship between community and outside world but also from the position of individuals and groups within the community. Degrees of economic well-being, political enfranchisement, and social integration and patterns of landholding varied greatly within communities, and so did interests. A marginal scamp like Bansuke was free to do largely as he pleased, while the obligations of leaders like Hachiemon and Hyōsuke could lead ineluctably to participation in conflict, even if the scaffold loomed beyond.

A third basis of inference is local history and culture (Yokoyama 1977, 1986). I found the concept of cultures of contention helpful. Certain localities were characterized over the years by far more conflict than others, and although I am unsure of the mechanism of transmission (I suspect local legends, songs, chronicles, and martyrs [Walthall 1986, 1991]), it does seem clear that there is a continuity of contention in some places which is not fully explicable by the ecological or political context.

But context is central, with regard to opportunities also. They are what the outside world appears to offer or withold, and some maintain that the opportunity structure for contention can in fact be more important than either interests or resources (Calhoun 1983: 898). I have found all to be important, but four types of opportunities in particular appear singularly closely associated with popular contention in early modern Japan.

Coerciveness is the first important aspect of the opportunity structure, and Tokugawa Japan illustrates the importance of consistency and time. Inconsistently coercive regimes—sometimes lax, sometimes repressive—invite more contention than consistently permissive or draconian regimes,¹⁴ and whereas repression may simply stir people up in the short run (Gurr 1980: 52–54; Hibbs 1973), in the long run it appears to work (Tilly et al. 1975; Wortman and Brehm 1975; Rothbaum 1980). In their early and most coercive years both the Tokugawa and Meiji regimes successfully reduced levels of contention nationwide; in the interim decades the progressive ineptitude and inconsistency of the Tokugawa regime invited steadily increasing levels of contention.

Apart from coercion, the simple absence of agents of government also invites contention.¹⁵ The autonomy of the Tokugawa village gave the people much more room for maneuver than some students of the period have suggested. Lack of supervision implies high overall levels of contention. I have found that, other things being equal, communities in less tightly and coherently administered regions were more prone to conflict.

A third aspect of opportunity of which the people took advantage was elite vulnerability. Succession, fiscal crisis, or factional division often spurred the people to action, as did failures of the elites to live up to their own principles of rule.¹⁶ Finally, the context presented the people with alternatives to and objects of contention. Among the alternatives were disaster relief and the provision of petition boxes and judicial redress; where these were used adroitly the regime escaped considerable conflict. At the same time, government offices and merchants’ homes and shops were lightning rods for dissent, and their presence goes far to explain the spatial distribution of contention.

The availability of courts thus constitutes an opportunity. Having a lawyer, that is, a tool or weapon in one’s own possession, constitutes a resource. Numbers, organization and cohesion, and leadership were the resources of greatest value to the Tokugawa common people. They always outnumbered the elites (DeNardo 1985), although the fact that they were essentially unarmed dictated circumspection. The gathering of large numbers of people in cities and towns greatly facilitated the inception of contention. And particularly at the inception, the prior gathering of a crowd—at, say, a market or festival—made it far safer for a potential leader to stick his neck out with minimal fear of being singled out.¹⁷

Bonding, either psychological or social, was another crucial resource (Tilly 1978: 60ff.). The people adopted a variety of rituals of bonding—oaths, petitions, or a common cup—which no doubt bucked up their courage and demonstrated collective resolve. The units of mobilization were usually whole communities, but one must not overestimate the unity of communities—especially cities and villages where socioeconomic differentiation was well advanced. Nevertheless, the higher levels of contention in cities and in more economically advanced regions suggest that common community interests—or at least cohesion and common interests among groups within the community—survived the developmental process.

There were two common types of leaders in popular contention, institutional and entrepreneurial. Hachiemon and Hyōsuke personify the institutional; Bansuke, the entrepreneurial. The formal positions of institutional leaders entailed the obligation to lead their communities in time of crisis; in light of the personal cost of inaction, leadership even in the face of severe punishment was a completely rational path (White 1988b). Entrepreneurial leaders were a much more fluid type, emerging from society’s margins in time of turmoil and motivated more by conviction or impulse than by obligation. I believe (see also Fukaya 1978) that the very process of socioeconomic change which corroded the cohesion of communities also produced more Robin Hoods like Bansuke, who lived by their wits on the edge of the law and the community. Their presence provided the common people with sufficient leadership to make contention possible.

This realm—of interests, opportunities, and resources—is the level of the model on which the clearest theoretical confrontation takes place, the most problematic stage of the causal process. It is here that those who espouse deprivation, normative, moral-economy, or value-, justice-, or indignation-driven explanations of contention confront those who espouse resource-mobilization, utilitarian, decision-focused, or rational-choice explanations.¹⁸ As noted, I consider both. This strategy is not novel, the confrontation notwithstanding; such participants in the debate as Ted Gurr and Charles Tilly (1993: 36) have acknowledged that both grievances and opportunities play a part; Edward Muller has systematically built both into his own model of aggressive political participation (1979), and modeler Jon Elster too (1989: 14, 20) has asserted that human actions in general are determined by desires (what one wants to do) and opportunities (what one can do or, more accurately, what one perceives that one is able to do). The acknowledgment of the multidimensionality of causation has in fact become general.¹⁹

Simply accepting all three dimensions, of course, begs the question of their interrelationships. I maintain that in early modern Japan interests led to opportunities (if you want to hit someone you will look for a chance to do so) and to resources (you will look not only for a chance to hit him but for something to do it with); opportunities influenced interests (on the one hand, consistent repression leads, over time, to lower aspirations; on the other hand, if you need money and the county magistrate leaves the vault open, why not grab some?) and resources (repression reduced the number of willing protesters; yet government establishment of a petition system transformed individuals into signatures, a new resource); finally, resources influenced both interests (if you have a hammer, suddenly a lot of things seem to need pounding) and opportunities, in that, for example, an official might become marvelously more open to a group whose unity was clear (Elster 1989: 17).

The final question to be asked about these three dimensions regards relative weight. In my view,

interests, or sudden changes in interests, were by themselves unlikely to lead to contention except in desperate, any-port-in-a-storm situations. The only common exception was when the obligations of the role one occupied (e.g., village headman) dictated action regardless of the probability of success.²⁰

Changes in opportunities alone could lead to contention, since all individuals have preferences and, everything else being equal, if the opportunity to achieve a preexisting goal presents itself, someone will take a shot at it.

Resources alone were highly unlikely to lead to contention; even if provided with a gun, one is unlikely to shoot if there is no desirable target in view.

Interests and opportunities, opportunities and resources, and all three together could lead to contention, but

interests and resources together were unlikely to lead to contention by themselves. Regardless of all else, one had to have a chance of success (except when one’s role required one to participate). Very few people, however angry or well armed, wanted to play Don Quixote. Some observers of popular contention have presumed the participants to be mindless bundles of impulse; neither leaders such as Hachiemon (and even Bansuke) nor even the urban rabble in a food riot fits this image. Desperation, perhaps; death wish, no.

To some extent individuals and groups continuously calculate interests, opportunities, and resources, and rarely does the calculation lead to contention. Some sort of catalyst or last straw, which generates a sudden change in interests, opportunities, or resources (or in the numbers of people in certain interest-opportunity-resource situations) is necessary to make individuals and groups suddenly adopt unconventional, contentious means. The sorts of stimuli of greatest importance in early modern Japan included natural disaster or dearth (particularly when combined with elite unresponsiveness); sudden changes in levels of deprivation (most Tokugawa commoners lived pretty close to the bone even in good times); inflexible imposition of old policies or adoption of new ones; unilateral elite violations of what I shall argue amounted to a social contract between elites and people; failure of elites to fulfill their obligations vis-à-vis the people; sudden threats to the subsistence or economic viability of individuals, families, or communities; and the appearance of division or indecision among elites.²¹

Given such catalysts, how did people decide whether or not to act? Dividing the normative and rational aspects of the decision is vain (Migdal 1988: 26ff.), for values, myths, symbols, and morals interact with rewards, sanctions, material constraints, and opportunities. Moreover, what was reasonable or rational depended on an individual’s or group’s position in the social hierarchy of wealth, prestige, and power; whether one was a potential leader or follower; the cultural context of the decision; and the stage of the conflict process at which the individual or group was faced with a decision. There was a big difference between stepping out to confront a phalanx of armed samurai and falling in with a rioting mob passing boisterously by.

My challenge was to avoid completely particularistic explanations of the decisional calculus of contention. At some level, every single such decision was sui generis, but my goal was generalization. I maintain that the decision to contend—either as leader or follower, initiator or joiner—had four general characteristics: it was rational, public, strategic, and institutional.²² I have already argued for rationality. Contention was also the pursuit of goods, such as tax cuts, that benefited the entire public, whether participants or not. This quality no doubt tempted some to sit back and let others take the risks, but this temptation was reduced by the fact that it was in everyone’s interest to participate because the bigger the incident the fewer the risks and the greater the odds of success, and because the price of failure could be survival.²³ Behavior was also strategic; that is, it depended on people’s expectations of others’ (allies’ and antagonists’) behavior, and in a society as densely organized as Tokugawa Japan, people