CHAPTER 1
THE TOKUGAWA PERIOD AND JAPANESE FEUDALISM
JOSEPH R. STRAYER
Most students of the European Middle Ages would now, I think, admit that feudalism existed in Japan. They would also admit that it was "real" feudalism, and not just a set of institutions that looked like feudalism on the surface but actually worked in a very different way. Japanese feudalism belongs to the same genus as European feudalism; it is not like the marsupial wolf which resembles a European wolf but in reality is a kind of ferocious opossum.
No matter what definition of feudalism is used, Japan can he brought under its terms. I happen to prefer a rather narrow political definition, on the grounds that feudalism simply ceases to have any specific meaning when it is used to describe economic and social conditions. Great estates cultivated by poorly paid, almost landless laborers can exist without feudalism (for example, collective farms), just as factories and an urban proletariat can exist without capitalism. It is true, of course, that the political structure necessarily affects the economic and social structure: the dominant group in any organized community will control and divert production to suit its own purposes and will impose its standards and values on society. But the methods and forms of economic control and the social values which are inculcated vary with the political system and not vice versa.
In political terms, feudalism is marked by a fragmentation of political authority, private possession of public rights, and a ruling class composed (at least originally) of military leaders and their followers. It is not necessary to demonstrate that these conditions existed in Japan, perhaps even more noticeably than in medieval western Europe. Indeed, while extensive areas in western Europe were never feudalized, none of Japan remained untouched by feudalism.
It is also clear that the early and middle stages of Japanese feudalism were not unlike the corresponding stages in Europe. At first, public authority and public officials continued to exist alongside of increasingly powerful feudal lords. The central government retained power in some areas while losing it in others, and only gradually faded away to a shadow. The peak of fragmentation of authority and of private control of rights of government came several centuries after the process of feudalization had begun. All this seems familiar and reasonable to a western medievalist.
It is the final stage — the reversal of the process of fragmentation — which is strange and puzzling. It is strange because devices which proved dangerous or ephemeral in Europe-such as the existence of large principalities-were used effectively in Japan by the shogun to increase his authority. It is puzzling because the final stage lasted so long, because the changes which took place in the Tokugawa period were so slow that at times they were almost imperceptible. I judge that my colleagues whose views are represented in this book are also somewhat puzzled by political developments between 1600 and 1850. It is a period in which the forms of feudalism were carefully preserved while much of the substance vanished. But if the substance was vanishing, why were the forms necessary? Conversely, if some of the substance remained (as I think it did), why was it not disruptive, why could it be safely tolerated by the shogunate?
Let me elaborate a little on these points. There are two periods in western European feudalism which bear some resemblance to Tokugawa Japan. The first comes at the very beginning, in the ninth and tenth centuries, when the last Carolingians tried to hold their realm together by establishing great commands — the duchy of Burgundy, the duchy of Gothia, the duchy of Francia, and so on. The men who held these commands were at first not unlike Tokugawa daimyos: they were officials of the king; they or their heirs could be moved from one district to another; they could be forced to surrender or exchange some of the counties they administered. But very soon the holders of the great commands became hereditary rulers practically independent of the king, and soon after that the great commands broke up into smaller units. Burgundy might be compared to Tosa: each province was about equally distant from central authority; each had aspirations to control neighboring areas. But Burgundy paid less and less attention to royal orders, and in turn the subordinate counts of Burgundy paid less and less attention to the orders of the duke. Tosa, on the contrary, remained united and obedient. In short, Japanese feudalism seems to be marching backwards; it ends in the pattern with which western feudalism began.
The other period in western history which has some parallel to Tokugawa Japan cannot be dated so precisely; it begins with the emergence of large feudal principalities in France and Germany in the eleventh or twelfth centuries and lasts until the principalities were either absorbed into a larger political unit or became independent states. But the difference is that in Europe the period of feudal principalities was a period of intense political competition. No one was satisfied with the status quo; no one thought that the principalities could continue indefinitely as autonomous but subordinate entities. Kings did not trust princes to remain obedient; they tried to add princely holdings to the royal domain. Princes did not trust kings to respect local autonomy; they did their best to reject all possibility of intervention by higher authorities. On the whole, the issue was decided fairly quickly. During the thirteenth century, the king of France annexed so many of the great principalities — Normandy, Anjou, Poitou, Toulouse, Champagne — that more than half the country was directly governed by him and his officials. The remaining principalities could be infiltrated with royal agents and absorbed at leisure. In the end only Flanders escaped. In Germany, during the thirteenth century the princes became practically independent. From that time on the accidents of war and inheritance, not the will of the emperor, determined the fate of German principalities.
This is not to say that feudalism, as a set of political institutions, ceased to have any influence in western Europe after 1300. It took many centuries to create the apparatus of centralized government, to build up efficient, well- trained, and reliable bureaucracies. Meanwhile local lords handled many of the details of local administration, either in their own right, or as deputies of a king or sovereign prince. As late as the seventeenth century Louis XIV, like a Tokugawa shogun, found it advisable to make his nobles spend part of the year at his court. As late as the seventeenth century the English gentry, who were not unlike the upper-level samurai of Japan, controlled local government through their positions as Justices of the Peace. But this late European feudalism had lost its basic structure and its monopoly of power. The old hierarchy of direct vassal, rear-vassal, rear-rear-vassal, etc., had vanished. Orders were no longer passed down a chain of command from count to baron to knight. The central government dealt directly with local authorities, great and small. The central government also had its parallel organizations, especially in matters of justice and finance, which reached down to the lowest levels of the population. Lords were left with enough prestige and power so that they could keep order among the peasants and collect their rents. They did not, as a rule, inflict serious punishments on peasants nor did they collect enough revenues from them to maintain large provincial armies or bureaucracies. In all these respects late European feudalism differed from Tokugawa feudalism.
Two questions emerge from this preliminary discussion: Why was the final stage of Japanese feudalism so prolonged ? Why was it so different from the Western experience?
Some obvious and familiar answers can be given at once. The pressures which force political change were either absent or at a low level during most of the Tokugawa period. There was no external threat and very little organized internal opposition. There was, for example, no Church to focus and justify discontent; the various religious groups were controlled by secular authority. There were almost no courts of justice, certainly no courts like the French parlements and English common law courts with their long traditions of independent judgment. Thus an inferior oppressed by a superior could not expect support from either religious leaders or legal experts. The rise of a business class would have been slow in any case; it was retarded even more by the virtual absence of the foreign trade which had strengthened and enriched European merchants. In the end the development of internal commerce may have been one of the most important solvents of the Tokugawa system, but this development was very gradual. Meanwhile merchants and bankers had no effective political organization and they could not use their limited economic power to oppose the government.
In short, the shogun and his advisers were less dependent on the privileged classes than a European king and Council. The shogun did not need officers from the feudal nobility to fight for him in foreign wars, nor bankers from the business community to finance them. He did not have to waste his energy in coercing bishops, judges, assemblies of estates, or semi-independent town governments. He could concentrate on one task: keeping enough of the daimyos obedient so that he would receive his share (a very large share) of the rice crop. Even if he held only one-sixth or one-seventh of Japan, this was still far more than any daimyo possessed. He could always count on the support, or at least the inertia, of most of the daimyos. He could overwhelm scattered rebels; and in the absence of any ideologies or institutions which could be used to unite latent opposition, scattered rebels were all that he was apt to meet.
These differences are immediately apparent. Others may have run deeper. The sheer size of the Japanese feudal class is amazing to anyone who has studied European feudalism; and size created problems which Europeans never had to face. It has been estimated that the maximum number of knight's fees which ever existed in England was about 6,000. This does not mean that there were 6,ooo knights, since many fees had been subdivided to a point where they gave no real service and others were held by tenants who were usually excused from service. A single han in Tokugawa Japan would have had a larger number of samurai than the largest possible number of English knights, and the samurai were all expected to perform some kind of service. There seems to be no doubt about these numbers; the records are clear, consistent, and comparable over extended periods of time. To deal with these hordes of minor vassals required a degree of planning and organization which was never necessary in Europe. Japanese feudalism had to be more structured, more impersonal, more bureaucratic than European feudalism. At least it had to have these qualities from the middle of the sixteenth century, when we have evidence of huge armies and large numbers of retainers under the control of a single lord.
Another factor which must have made Japanese feudalism, even in its middle period, more structured and bureaucratic than European feudalism was the difference in the arrangements for providing economic support to the vassal. Household retainers were common in early European feudalism and they never entirely disappeared, but they were not an important element of the feudal class after 1100. Most vassals by that time had fiefs, and very rapidly the vassal became the virtual owner of his fief. In these circumstances, it would have been difficult to change the value of a fief and utterly impossible to try to create classes of fiefs in which each member of the class had a more or less uniform income. In most countries records were so poor that no one could have ascribed an exact value to any one fief or group of fiefs. Even in England, which was unique in having a survey of feudal possessions (Domesday Book), estimates of income were not very precise. England was also unique in groping toward the idea that the value of one type of fief — the knight's fee — should be standardized. But the evidence suggests that some lords thought an annual income of 10 pounds enough for a knight, that others used 20 pounds a year as a standard, and that in fact there were great variations in the yearly income of knights. Above the level of the knight there was no standardization whatever. When the concept of peerage developed and it became important to know whether a vassal was a knight or a baron, the king and his minsters were obviously perplexed. A man might be summoned as a baron on one occasion and be omitted from the list of barons on the next. Personal ability and political influence were as important as wealth in determining status, and no one would have said that an income of say 110 pounds a year automatically made a man a baron. In other countries very little was known about the actual income of individual vassals and there could be extremely wide variations among the incomes of men who theoretically belonged to the same level in the feudal hierarchy. Even in the late stages of European feudalism, when more was known about the value of holdings, there was no attempt to fix maximum and minimum incomes for certain ranks. After meeting what were, by Japanese standards, the very modest demands of his superior, the lord could exploit his holdings as he saw fit and keep any surplus which he had created. He remained closely tied to his estates and the change from feudal lord to landlord still left him with great prestige and influence in local government.
The rice economy of Japan seems to have lent itself more easily to exact measurement of feudal incomes. The vassal was to have so many measures of rice, not so many villages, or ploughlands, or acres. Such a system would work only if there were careful and repeated surveys of rice production, surveys which were far more thorough than Domesday Book and a great deal more common. Such surveys could be made only by higher authority, and as evidences of this authority they immediately reduced the autonomy of the local fief-holder. More important, the Japanese system made it easy to break the ties between vassals and the land. If a vassal was to have only a fixed rice income, then it made little difference whether he took it directly from a village or the daimyo's storehouse. Thus it was possible to concentrate the majority of the feudal class in castle-towns and to keep close control over the minority which still lived in the country. There was not much chance for an independent, self-sufficient squirearchy to develop under this system. And just as the daimyo broke the ties between his vassals and the land, so the shogun weakened the ties between the daimyo and his province. If the daimyo had to spend half of his time and much of his income in Edo, he was scarcely a territorial lord.
Finally, the concept and nature of authority in Japan differed tremendously from the European pattern. In Europe, the symbol of authority was the right to hold a court. The growing power of kings and princes in the later Middle Ages could be measured by the degree to which their courts extended their jurisdiction. The feudal hierarchy was broken down through legal procedures; when the king's courts could protect minor vassals, the immediate superiors of minor vassals lost much of their authority. The first attempts to define sovereignty were bound to the idea of law: he is sovereign who can render final decisions without appeal in all law suits, or make law for the common welfare. It is true that there was an argument as to whether kings were bound by law, but it was a sign of true kingship to respect the law, whether it was binding or not. And, as a practical matter, it caused a considerable amount of inconvenience for kings when they did not respect the law, especially the law which protected the rights and possessions of the feudal class. The legal systems which had been built up in the thirteenth and fourteenth centuries were sufficiently complicated so that a man with wealth and status could be very annoying if he decided to defend his rights. A number of royal officials would have to conduct investigations, find witnesses, prepare briefs, argue about the meaning of documents, and hear appeals from preliminary decisions. It was not difficult for a determined man to keep a case going for ten or twenty years. As a last resort, European opinion condoned rebellion if the rebel could find any sort of legal justification for his act. Of course, the king could usually win the law suit or suppress the rebellion if he made a determined effort, but he had neither the time, the resources, nor the trained personnel required to carry every case to a successful conclusion. Open or tacit compromise was often the only prudent solution. In short, the states of late medieval and early modern Europe were, to use the German phrase, law-states. Authority was based on law.
