Classic Japanese Anthropology

Special Issue Contemplating Ruth Benedict’s The Chrysanthemum and the Sword

Takeyoshi Kawashima, Hiroshi Minami, Kizaemon Ariga Tetsuro Watsuji, and Kunio Yanagita1

The Issue of Hierarchy in Japanese Social Structure

Kizaemon Ariga2

Translated by James E. Roberson3 Kanazawa Seiryō University

1. Ruth Benedict’s The Chrysanthemum and the Sword was for me a very interesting work. While there are points in the book which due to inadequate investigation of resources are mistaken or exaggerated, that important points regarding the “mentality”4 of the Japanese are touched upon is witness to Benedict’s personal excellence and to the high level of American cultural anthropology. When I read this book, my anticipatory interest was in how a cultural anthropologist who had been engaged in research on primitive peoples had conducted research about a highly cultured people having ancient historical records. I recalled that when Robert Redfield

1 This special issue was originally published in Minzokugaku Kenkyū (Japanese Journal of Ethnology), Volume 14, Number 4, 1950. 2 Editor’s note: The kanji (Chinese characters) used for the author’s surname are 有賀, pronounced a-ru-ga. In the English-speaking world, however, he is better known as a-ri-ga, and he in fact called himself that when he wrote an influential paper in English (“On the Japanese Family,” Marriage and Family Living 16(4): 362-368, 1954). This confusion once prompted certain American professors to ask the author, through his Japanese friend, which pronunciation was correct. Here is what the author in essence said: in his native place of Suwa, Nagano Prefecture, he is known as a-ru-ga, but in other regions his surname is ordinarily pronounced a-ri-ga. From this the author assumed that the former was a localism, which made him think that as a nationally known professor he should switch to the latter, but he eventually decided to “let the characters [有賀] stand alone to be pronounced as the reader saw fit” (Robert J. Smith, Japanese Society: Tradition, Self, and the Social Order, Cambridge: Cambridge University Press, 1983. Page 81). For these reasons, we have here decided to follow the common English usage. 3 Translator’s note: my thanks to Masaya Shijō and Takami Kuwayama for making corrections and amendments to this translation. 4 Translator’s note: Ariga consistently and repeatedly used the English term “mentality” without translating it into Japanese. Japanese Review of Cultural Anthropology, vol. 16, 2015 22 Kizaemon Ariga visited Japan last year, he told the small gathering of Japanese ethnographers who greeted him that historical research was a weak point in American cultural anthropology. While what he said may have had various meanings, I am certain that they included an objection to calls to limit the objects of cultural anthropology or ethnography to primitive peoples. In The Chrysanthemum and the Sword, Benedict does not make any particular positive assertions regarding this point. However, as her research goal of portraying “cultural patterns” is not unrelated to history, I believe that we can wait hopefully for her views on this issue. Benedict deals with history primarily in Chapter 4 regarding the Meiji Restoration period, but wherever else necessary as well she endeavors to explain relevant historical interpretations. The historical materials that she employs include both those recorded by Westerners and those recorded in the research of Japanese scholars. For example, these include things such as Kan’ichi Asakawa’s The Documents of Iriki (入来院文書). However, as she explains, Benedict’s methodology in such cases is based on the observation of the everyday life of contemporary Japanese people – as such is the decisive methodology of cultural anthropologists who study primitive peoples – and even when utilizing older historical resources, these are used as reference materials in order to interpret the mentality of the Japanese as manifest in contemporary life. In other words, older materials as well are interpreted as among the foundations of contemporary Japanese mentality. This is strikingly different from the methods of most historians, and one may imagine that many historians will oppose such an interpretive methodology. This is because they believe that it is possible to extract history only from older historical resources and that the contemporary contains nothing prescribing the past. But is such a view correct? What must be noted regarding cultural anthropological methodology is that, in regards to research on all primitive peoples, while anthropology attempts to grasp the characteristic mentality of such people, and while anthropology includes diverse training so as to as much as possible understand the meanings or values of those people’s life phenomena from such a position, scientific research is not constituted solely by the interpretation of such primitive people’s lives from a position that is purely their own. For a cultural anthropologist to attempt to know the mentality of a primitive people, he must, in order to establish meaning(s) as scientific resources, avoid his personal researcher’s biases as much as is possible and make his interpretations from the scientific position with which he himself is currently affiliated. As regards historical interpretations of ancient periods as well, however much an historian might try to emphasize his correctness, such interpretation cannot be done only with things ancient. The historian’s contemporary rational sense must be part of his interpretations of the mentality of ancient peoples. The historian does so for all cases, even when he is not conscious of it. This is true of the interpretation of whatever period. That the interpretation of a people’s ancient history is difficult using only that history is indicated by the facts that, since historical resources regarding ancient things are insufficient, the employment of other materials has been extremely common, including resources from the ancient periods of other peoples and inferences drawn from the lives of contemporary Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 23 visited Japan last year, he told the small gathering of Japanese ethnographers who greeted primitive peoples. If one is to understand ancient history, just as with the explanation of the him that historical research was a weak point in American cultural anthropology. While lives of primitive peoples, historical interpretation must utilize our contemporary what he said may have had various meanings, I am certain that they included an objection perceptions of rationality and history. The scientific establishment of views of history is to calls to limit the objects of cultural anthropology or ethnography to primitive peoples. In important, and in order to do this, it is fundamentally important to know the contemporary The Chrysanthemum and the Sword, Benedict does not make any particular positive assertions life of any given people. regarding this point. However, as her research goal of portraying “cultural patterns” is not Is it in fact correct to think that what are commonly referred to as contemporary studies unrelated to history, I believe that we can wait hopefully for her views on this issue. are without an historical perspective? Simply because the primitive peoples who are the Benedict deals with history primarily in Chapter 4 regarding the Meiji Restoration period, research objects of cultural anthropologists “have no written history,” it is impossible to but wherever else necessary as well she endeavors to explain relevant historical imagine that there exists absolutely no sense of history of or among them. Even if one interpretations. The historical materials that she employs include both those recorded by suggests that cultural anthropology is above all the study of the contemporary, since the Westerners and those recorded in the research of Japanese scholars. For example, these contemporary itself holds historical meaning, one must have an historical perspective for the include things such as Kan’ichi Asakawa’s The Documents of Iriki (入来院文書). However, as interpretation of the contemporary as well. The issues of “culture change” and “acculturation” she explains, Benedict’s methodology in such cases is based on the observation of the that American cultural anthropology is currently intently dealing with all have historical everyday life of contemporary Japanese people – as such is the decisive methodology of meaning. These are phenomena which all past peoples tirelessly repeated and that will no cultural anthropologists who study primitive peoples – and even when utilizing older doubt be tirelessly repeated by all peoples of the future as well. historical resources, these are used as reference materials in order to interpret the mentality That the problems of the present are connected to both the past and the future is of the Japanese as manifest in contemporary life. In other words, older materials as well are something of great historical meaning. In other words, since the contemporary is, if interpreted as among the foundations of contemporary Japanese mentality. rigorously viewed, but an instance, it is a mediator between the past and the future. This This is strikingly different from the methods of most historians, and one may imagine that shows that the contemporary is in relations of mutual definition with both past and future many historians will oppose such an interpretive methodology. This is because they believe and that the contemporary moment thus holds historical meaning. It is thus impossible for that it is possible to extract history only from older historical resources and that the cultural anthropologists not to have an historical view even when taking as research object contemporary contains nothing prescribing the past. But is such a view correct? What must primitive peoples without written histories, and even when the cultural anthropologist be noted regarding cultural anthropological methodology is that, in regards to research on himself is not aware of this. It is just that, because the chronological developments in the all primitive peoples, while anthropology attempts to grasp the characteristic mentality of cultures of primitive peoples are torpid and obscure, they do not appear clearly during or such people, and while anthropology includes diverse training so as to as much as possible from research. understand the meanings or values of those people’s life phenomena from such a position, However, when the cultural anthropologist’s research object moves to that of high-culture scientific research is not constituted solely by the interpretation of such primitive people’s peoples with written histories, he must suddenly reveal his view of history. It is natural that lives from a position that is purely their own. For a cultural anthropologist to attempt to Benedict conducted her historical interpretations from her own position, and she is not know the mentality of a primitive people, he must, in order to establish meaning(s) as mistaken to have used cultural anthropological methods to study Japanese history. The scientific resources, avoid his personal researcher’s biases as much as is possible and make problem is in whether or not such were accurate or exact. There are mistakes and his interpretations from the scientific position with which he himself is currently affiliated. exaggerations in Benedict’s study of Japan, but it is important to recognize that her methods As regards historical interpretations of ancient periods as well, however much an were relatively effective. Of course, in that, one must beware of the following things, historian might try to emphasize his correctness, such interpretation cannot be done only themselves accompanied by special reasons or grounds. with things ancient. The historian’s contemporary rational sense must be part of his Here, I would like to touch briefly on methodological problems in historical research. This interpretations of the mentality of ancient peoples. The historian does so for all cases, even regards the research methodology of attempts such as Benedict’s to retrace historical origins when he is not conscious of it. This is true of the interpretation of whatever period. That the from research on the contemporary, which has already become an issue among Western interpretation of a people’s ancient history is difficult using only that history is indicated by scholars. Originally, historians generally extracted history from archaic texts, but I believe the facts that, since historical resources regarding ancient things are insufficient, the that it is of even more value to parallel this with the method of retracing origins. As already employment of other materials has been extremely common, including resources from the noted, the interpretation of history is done on the basis of our contemporary sense of ancient periods of other peoples and inferences drawn from the lives of contemporary rationality. Because of this, the sense of rationality needed for such explanation, that is the 24 Kizaemon Ariga forming of a view of history, is of the greatest importance for the historian, and the founding of a scientific view of history is needed to establish history as a science. For this, one must have sufficient and accurate resources that fully enable scientific investigation. The contemporary possesses fundamental importance, more than all other periods, in the sense of being able to systematically make available those resources needed for scientific investigation according to the researcher’s desires. Even in cases when there is no such self-awareness, the contemporary holds decisive significance for the historian. For such reasons, it is possible to establish a still more scientific view of history based on the retracing of the origins of things from the contemporary, about which deep and wide understandings are possible, back to early modern (kinsei 近世), medieval (chūsei 中世) and ancient (kodai 古代) periods. The reason such methods are especially effective in Japan is that, at least since the Asuka Period (586-), 5 there have been no great changes in the racial composition of Japan. Throughout this period, Japan only accepted the immigration of small numbers of other ethnic group members, and thus one may say that the ethnic Japanese people of today had already formed at the beginnings of Japanese history.6 If one considers that there was not an extremely large difference [in Japan’s ethnic composition]7 even in the Japanese Paleolithic, then it may be well to go even further back in time. And, regarding the special characteristics manifest by contemporary Japanese, it is not at all unreasonable to take the life of the Japanese of this period as the basis for contemplation, given that state sovereignty has resided in the Imperial family from the Asuka Period, and given that it is impossible to doubt the continuity of the tendencies of the special ethnic characteristics of the Japanese from this period, since other than today’s Allied Occupation there has not been a long-term occupation of Japan by another people. In cases where there have been violent changes in racial composition or, as a consequence, there have been changes in national sovereignty, it is impossible to imagine the continuity of homogeneous ethnic characteristics, and it is thus difficult to use the contemporary to make inferences about the ancient past. For this reason, one must think of methods that are appropriate to the nature of those changes. While the research methodology of retracing origins does not lose its validity even here, in the case of Japan [with its continuity of racial composition and national sovereignty] it is especially effective. Here lie the reasons that Benedict has been able to analyze the tendencies of the characteristics of the Japanese people more acutely than have many Japanese historians. In view of the above, The Chrysanthemum and the Sword shows us that there is no reason to limit the objects of cultural anthropology or ethnology to primitive peoples having no written

5 Translator’s note: various dates are given for the Asuka (飛鳥) Period, which lasted from the mid-6th to the early 8th centuries. 6 Translator’s note: the terms“racial” and “ethnic” are used respectively for the adjective forms of “jinshu” (人 種) and“minzoku” (民族). Depending on the context, the latter means “people,” “ethnicity,” “ethnic group,” “nation,” and so forth. 7 Translator’s note: brackets are used for the words inserted by the translator for clarity. Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 25 forming of a view of history, is of the greatest importance for the historian, and the founding histories. And, this tells us as well that neither cultural anthropology nor ethnology is of a scientific view of history is needed to establish history as a science. unrelated to history, even when dealing with primitive peoples without written histories. For this, one must have sufficient and accurate resources that fully enable scientific Regardless of a people have a written history or not, the facts of being formed as a people investigation. The contemporary possesses fundamental importance, more than all other and retaining a culture have historical meaning, and as such we must pursue such meaning. periods, in the sense of being able to systematically make available those resources needed The reason that cultural anthropology in England and America has shifted from cultural- for scientific investigation according to the researcher’s desires. Even in cases when there is historical tendencies to concentrate on contemporary-studies tendencies is because of the no such self-awareness, the contemporary holds decisive significance for the historian. For disappointments with, after Lewis Henry Morgan, theories of world historical stages of such reasons, it is possible to establish a still more scientific view of history based on the development. German ethnology as well became similarly disappointed, but gave birth to retracing of the origins of things from the contemporary, about which deep and wide kulturkreis (culture circle/complex) and kulturschichten (culture horizons/strata) theories in understandings are possible, back to early modern (kinsei 近世), medieval (chūsei 中世) and order to make more detailed cultural-historical investigations. ancient (kodai 古代) periods. However, are these two tendencies really completely different? One is attempting to more The reason such methods are especially effective in Japan is that, at least since the Asuka deeply submerge within the contemporary to explain ethnicity (minzoku) and culture, while Period (586-), 5 there have been no great changes in the racial composition of Japan. the other is attempting to more deeply pursue the past in order to explain ethnicity and Throughout this period, Japan only accepted the immigration of small numbers of other culture. However, even if one says that the study of the contemporary is submerged fully ethnic group members, and thus one may say that the ethnic Japanese people of today had within the contemporary, it is impossible to deny historical meaning. And, even if one claims already formed at the beginnings of Japanese history.6 If one considers that there was not an that it is the elucidation of primitive cultural history that pursues the past, it is impossible extremely large difference [in Japan’s ethnic composition]7 even in the Japanese Paleolithic, to deny the contemporary. This is because, as regards peoples and their cultures, things then it may be well to go even further back in time. And, regarding the special similar to those that occurred in the past are occurring in the contemporary, and things characteristics manifest by contemporary Japanese, it is not at all unreasonable to take the similar to those that are occurring in the contemporary also occurred in the past. It is of life of the Japanese of this period as the basis for contemplation, given that state sovereignty course inappropriate for us constitute these two academic styles as opposites and to choose has resided in the Imperial family from the Asuka Period, and given that it is impossible to but one of the two. doubt the continuity of the tendencies of the special ethnic characteristics of the Japanese Rather, we must further advance and link together and examine those peoples without from this period, since other than today’s Allied Occupation there has not been a long-term and those peoples with written histories. This is because the concept of an ethnic group or occupation of Japan by another people. people should not monopolize one or the other category. A people/nation (minzoku) is not a In cases where there have been violent changes in racial composition or, as a consequence, fixed and fundamental society. Together with the origins of culture, their origins are unclear. there have been changes in national sovereignty, it is impossible to imagine the continuity of However, the origins of peoples and cultures may be understood from the currently existing homogeneous ethnic characteristics, and it is thus difficult to use the contemporary to make figures of those peoples of whom we are now aware. Given that it cannot be said that such inferences about the ancient past. For this reason, one must think of methods that are inquiry has yet been fully made, and even given changes in conceptual definitions appropriate to the nature of those changes. While the research methodology of retracing accompanying advances in science, things that may be seen in contemporary form may be origins does not lose its validity even here, in the case of Japan [with its continuity of racial assumed to have already similarly been extant in the past. composition and national sovereignty] it is especially effective. Here lie the reasons that If the present is taken as a key with which to interpret the past, ethnicity and culture bear Benedict has been able to analyze the tendencies of the characteristics of the Japanese great potential to be rank-ordered from low-culture to high-culture peoples. There must have people more acutely than have many Japanese historians. been innumerable conditions defining these. It is thus impossible to imagine, for example, In view of the above, The Chrysanthemum and the Sword shows us that there is no reason to that contemporary low-culture peoples alone are able to illuminate primitive peoples or limit the objects of cultural anthropology or ethnology to primitive peoples having no written cultures of the past, and even if one creates various patterns of the latter from the former it will remain difficult to easily endorse such claim. Peoples or ethnic groups (minzoku) are not

5 Translator’s note: various dates are given for the Asuka (飛鳥) Period, which lasted from the mid-6th to the fixed but are without stop continuously dividing, mixing, compounding, blending, early 8th centuries. assimilating, expelling, controlling, submitting, branching, weakening or becoming extinct, 6 人 Translator’s note: the terms“racial” and “ethnic” are used respectively for the adjective forms of “jinshu” ( 国民 種) and“minzoku” (民族). Depending on the context, the latter means “people,” “ethnicity,” “ethnic group,” or forming or dismantling nations (kokumin ). These diverse tendencies are based on “nation,” and so forth. various conditions. It is not the case that there are no resources amongst high-culture 7 Translator’s note: brackets are used for the words inserted by the translator for clarity. 26 Kizaemon Ariga peoples as well allowing one to envision primeval cultures. This is because not all ethnic groups have manifested the marked cultural stagnation of contemporary primitive peoples. Ethnic cultural history involves not just the investigation of primitive/primal peoples of the past but also the ethnography of the contemporary. Even if primal ethnic groups and later peoples are not exactly the same, as ethnic peoples they both must have (had) the essence of ethnic peoples. Just as the cultural-historical approaches and contemporary- studies approaches of ethnology are connected, so too through their various negotiated relationships are high-culture peoples and low-culture peoples deeply inter-related. If ethnology or cultural anthropology is to take ethnic groups and cultures as research objects, its field of vision must be as such. The Chrysanthemum and the Sword has thus in this sense presented us with both a research problem, and in a certain sense, its solution.

2. Although it may be said that The Chrysanthemum and the Sword grasps the mentality of the Japanese from many angles, that Benedict points to hierarchy in Japanese social structure as the foundation of this mentality is especially incisive. While in contemporary times hierarchy involves the pyramidal political-social structure with the Emperor at its apex, [according to Benedict] hierarchy involves the following concrete ordering. The new Imperial system nation-state which appeared with the Meiji Restoration simplified the pre-existing hierarchy by expelling the Shōgun and feudal domain heads, and it alleviated the contradiction that had existed between loyalty to these versus loyalty to the nation-state = Emperor. However, these new changes did not weaken hierarchy but further strengthened it by centralizing power. In all fields such as politics, religion, economics and so forth, hierarchy finely regulated the obligations of each class in the relations between the people and the state. Looking at politics first, strata were ranked in the order of the Emperor, senior statesmen able to receive an audience with the Emperor, direct counselors to the Emperor, and those high and low level officers receiving letters of appointments signed with the imperial seal. These men had no need to follow public opinion, and they obeyed orders from above. The head leaders among them were neither elected by the public nor included those elected. Popularly elected members of the House of Representatives participated in legislation but they had no authority to either select or approve ministers or high officials. Another hierarchical barrier existed in the fact that, regarding legislation, such elected representatives were further restricted by the House of Peers as composed of the nobility and by Imperial Order. In and befitting this political hierarchy, there was also a system of self-governing bodies. These were the local governments. In local self-governing municipalities, the publicly elected elders selected a responsible leader. This local leader administered the local government. However, these local self-governments were not granted authority regarding the appointment of judges, police and educators, which fell under the authority of the nation- Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 27 peoples as well allowing one to envision primeval cultures. This is because not all ethnic state. The hamlet (buraku 部落)8 was a lower-level institution of local self-governance. It groups have manifested the marked cultural stagnation of contemporary primitive peoples. existed purely as a self-governing body and was not woven as a unit into the [national] Ethnic cultural history involves not just the investigation of primitive/primal peoples of administrative organization, and as such it constituted an area into which the power of the the past but also the ethnography of the contemporary. Even if primal ethnic groups and nation-state did not extend but inside of which close cooperative associations were developed. later peoples are not exactly the same, as ethnic peoples they both must have (had) the Also, the hamlet head had no responsibility to the State to collect national taxes. essence of ethnic peoples. Just as the cultural-historical approaches and contemporary- While Benedict makes many mistakes regarding rural villages, she does point out that a studies approaches of ethnology are connected, so too through their various negotiated comparatively free system of self-governance existed along with the [centralized national] relationships are high-culture peoples and low-culture peoples deeply inter-related. If political hierarchy. While in shape this duality does not necessarily vary from Western ethnology or cultural anthropology is to take ethnic groups and cultures as research objects, political forms, they were functionally extremely different. That is, because they were its field of vision must be as such. The Chrysanthemum and the Sword has thus in this sense managed according to ancient customs of allegiance formed from the ethical system and presented us with both a research problem, and in a certain sense, its solution. etiquette of the Japanese people, the privileges of ministers and high officials in the political hierarchy were respected as appropriate to their status and, even though the opinion of the 2. Although it may be said that The Chrysanthemum and the Sword grasps the mentality of the populace was not reflected at the top levels of national government, the government was Japanese from many angles, that Benedict points to hierarchy in Japanese social structure able to call on the support of the people, who accepted this. However, the State in turn had as the foundation of this mentality is especially incisive. While in contemporary times to follow the decisions of the people in those regions where the rule of popular opinion was hierarchy involves the pyramidal political-social structure with the Emperor at its apex, considered appropriate. The lifestyle of the Japanese people divided authority appropriate to [according to Benedict] hierarchy involves the following concrete ordering. each into separate spheres and determined the corresponding area of each. The new Imperial system nation-state which appeared with the Meiji Restoration Looking next at religion, here as well Benedict grasps the duality that exists. That is, she simplified the pre-existing hierarchy by expelling the Shōgun and feudal domain heads, and distinguishes between State Shintō and other religions. There is no large gap between the it alleviated the contradiction that had existed between loyalty to these versus loyalty to the latter and religion in America, as it is a matter of entrusting the freedom of personal belief nation-state = Emperor. However, these new changes did not weaken hierarchy but further to the individual, but the former fell under state jurisdiction, especially as the religion strengthened it by centralizing power. In all fields such as politics, religion, economics and worshipping the symbols of national unity and superiority. For that reason, as “ruler from so forth, hierarchy finely regulated the obligations of each class in the relations between the the unbroken imperial line,” the Emperor and his ancestor Ōmikami were people and the state. worshipped in State Shintō. Paralleling the political hierarchy, a hierarchical system of Looking at politics first, strata were ranked in the order of the Emperor, senior statesmen shrines where ritual observances were carried out, and of the priests managing these, was able to receive an audience with the Emperor, direct counselors to the Emperor, and those systematically determined based on their statuses. And, the Emperor and the Grand Shrine high and low level officers receiving letters of appointments signed with the imperial seal. of Ise were placed at the zenith of the hierarchy of State Shintō. For this reason, State These men had no need to follow public opinion, and they obeyed orders from above. The Shintō was not considered a religion, as it had no doctrine and involved nothing more than head leaders among them were neither elected by the public nor included those elected. the conduct of ritual services with purification at their center. On State Shintō holidays, the Popularly elected members of the House of Representatives participated in legislation but Emperor executed the rituals as representative of the national population. Outside of the they had no authority to either select or approve ministers or high officials. Another functions or domains of the nation-state within Shintō, the people were entrusted with the hierarchical barrier existed in the fact that, regarding legislation, such elected freedom of religion, but regarding things which from the politicians’ perspective were representatives were further restricted by the House of Peers as composed of the nobility directly related to the State, these high officials of the new hierarchy reserved to themselves and by Imperial Order. the right to rule. In and befitting this political hierarchy, there was also a system of self-governing bodies. Thirdly, let us examine economics. Here again Benedict attempted to see duality. These were the local governments. In local self-governing municipalities, the publicly elected Following the Meiji Restoration, the development of industry pursued a different path than elders selected a responsible leader. This local leader administered the local government. However, these local self-governments were not granted authority regarding the 8 Editor’s note: After the early Meiji amalgamation of villages and towns into larger administrative units, buraku refereed to “hamlet,” the smallest unit of communities of people in rural Japan. Today this term has appointment of judges, police and educators, which fell under the authority of the nation- been replaced by “shūraku” (集落) to avoid confusion with “hisabetu buraku” (被差別部落) or outcaste community. By contrast, “mura” (村), which is ordinarily translated into English as “village,” was a political unit officially recognized by the state. 28 Kizaemon Ariga that in Western countries. The Government did the planning, with government funds did the construction, and with government bureaucrats did the organizing and managing [of industry]. Foreign engineers were invited in and Japanese were sent abroad to train as technicians. As confidence was gained in the success of the state enterprises, they were sold to private companies and developed as private businesses. However, these were sold to a select minority of private investors and became a source for the growth of zaibatsu [financial- industrial conglomerates]. For this reason, capitalist production in Japan did not trod down the normal route. The Japanese government did not embark with the production of consumer goods and light industry, but gave priority to heavy industries such as shipbuilding, munitions, steel- making, rail construction, and so forth, which rapidly reached high standards of technical capability. Not all of these were sold to private enterprise, as the huge munitions industry was kept by the government. Government gave priority to those industries with some relation to government administration. Only those huge zaibatsu which had received special benefit from the government participated in this, and since there was no direct participation by smaller commercial or industrial enterprises or by entrepreneurs who were not bureaucrats, they were thought to hold low social positions. In this lower domain of industrial freedom, all varieties of light industry were operated by petty capital and low- wage labor. This duality of industry in Japan was, for Japanese people’s lifestyle, of exactly the same importance as the dualities manifest in the fields of politics and religion. As in the hierarchy manifest in other fields, when they felt the aristocracy of the financial world necessary, it appears as though Japanese politicians constructed every form of strategic industry for them, politically selected merchant families to whom to provide special benefits and, placing these people in suitable status positions, determined their relations with other(s in the) hierarchies. Above, I have simplified Benedict’s more detailed account, though regarding the future of duality and hierarchy in Japanese society, Benedict’s explanation is ambiguous. If anything, this is because, should the hierarchical system continue in full there will be no space into which duality can intervene. How is it possible for the system of local self-governance in politics to be free from state politics? There are no domains in which to be free from politics. Even individual behavior must accept such restriction when expressed as [public] action. Religious belief not affiliated with State Shintō, while perhaps granted absolute freedom if kept alive within the individual heart, was subject to political restrictions when organizing churches and [publicly] expressing belief based action. Further, it was a social necessity to determine one’s relationship with State Shintō. “Free” enterprise as well occupied but one position in the hierarchical system. It was at all times located in the midst of the overall relationships of industrial organization, was in intimate relations with government enterprise or industries divested by government, and was controlled and restricted by these. If hierarchy exists, it is necessary to provide a clear explanation of how we are to interpret Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 29 that in Western countries. The Government did the planning, with government funds did the existence of dualities in which things of completely different character are in opposition the construction, and with government bureaucrats did the organizing and managing [of to each other. Regardless of this, I do not intend to criticize the fact that Benedict has industry]. Foreign engineers were invited in and Japanese were sent abroad to train as pointed to the existence of hierarchy in Japanese society. technicians. As confidence was gained in the success of the state enterprises, they were sold Benedict views hierarchy in the Japanese social structure as having been established to private companies and developed as private businesses. However, these were sold to a during the feudal era. In this sense it is not clear whether or not she sees the hierarchy select minority of private investors and became a source for the growth of zaibatsu [financial- established in feudal times as that with roots in ancient times. In any case, it appears that industrial conglomerates]. For this reason, capitalist production in Japan did not trod down Benedict believes that the feudal era became the largest ground for the growth of hierarchy the normal route. in Japan. However, Benedict also argues that the Meiji Government politicians, who The Japanese government did not embark with the production of consumer goods and dismantled feudal society, rejected all thought which attempted to eradicate hierarchy and, light industry, but gave priority to heavy industries such as shipbuilding, munitions, steel- not just causing the continuation of such hierarchy, simplified and further strengthened it making, rail construction, and so forth, which rapidly reached high standards of technical by abolishing the Shōgun and feudal lords. capability. Not all of these were sold to private enterprise, as the huge munitions industry Given that the hierarchical relations commonly interpreted as characteristic of the feudal was kept by the government. Government gave priority to those industries with some period did not vanish with the collapse of the feudal regime, it might appear that Benedict is relation to government administration. Only those huge zaibatsu which had received special interpreting contemporary hierarchy as a form of feudal inheritance, but this is not the case. benefit from the government participated in this, and since there was no direct participation According to her research, because hierarchy has remained alive and well in the everyday by smaller commercial or industrial enterprises or by entrepreneurs who were not lives of Japanese throughout the Meiji [明治 1868-1912], Taishō [大正 1912-1926] and Shōwa bureaucrats, they were thought to hold low social positions. In this lower domain of [昭和 1926-1989] periods, and even in the postwar of today, it is not just a feeble survival industrial freedom, all varieties of light industry were operated by petty capital and low- from the feudal past. Benedict contradicts herself with this sort of interpretation, and she is wage labor. mistaken in her explanations regarding hierarchy as being visible only in relation to This duality of industry in Japan was, for Japanese people’s lifestyle, of exactly the same political control. importance as the dualities manifest in the fields of politics and religion. As in the hierarchy If it is possible to recognize vigorous forms of hierarchy both in direct relation to politics manifest in other fields, when they felt the aristocracy of the financial world necessary, it and in everyday life and, further, if even the political reforms of the Meiji Restoration were appears as though Japanese politicians constructed every form of strategic industry for them, unable to change these, then it is impossible to view hierarchy as just a political politically selected merchant families to whom to provide special benefits and, placing these configuration. Hierarchy is expressed in political form as well as in many other things. people in suitable status positions, determined their relations with other(s in the) Benedict sees hierarchy as present both before and after the Meiji Restoration and thus hierarchies. claims that the Meiji Restoration was not what Westerners would call a revolution, but she Above, I have simplified Benedict’s more detailed account, though regarding the future of does not necessarily appear to have clearly grasped the relationships between hierarchy and duality and hierarchy in Japanese society, Benedict’s explanation is ambiguous. If anything, political configurations. If one defines hierarchy as characteristic of feudal systems, then it this is because, should the hierarchical system continue in full there will be no space into is impossible to think that such hierarchy exists even now in the postwar of today. Thus, if which duality can intervene. How is it possible for the system of local self-governance in hierarchy exists still today, then one must understand it as not being characteristic of the politics to be free from state politics? There are no domains in which to be free from politics. feudal system. Furthermore, if it is taken that such hierarchy was of necessity also present Even individual behavior must accept such restriction when expressed as [public] action. in feudal society then this would have to be a form of hierarchy capable of continuing to exist Religious belief not affiliated with State Shintō, while perhaps granted absolute freedom if across all Japanese historical periods, and then it is without doubt more appropriate to label kept alive within the individual heart, was subject to political restrictions when organizing this as a characteristic ethnic (minzokuteki 民族的) tendency of the Japanese. churches and [publicly] expressing belief based action. Further, it was a social necessity to Then if so, is there such a thing? I believe there is. I myself do not have comprehensive determine one’s relationship with State Shintō. “Free” enterprise as well occupied but one evidence for all historical periods, so my remarks here are partially inferential, but I believe position in the hierarchical system. It was at all times located in the midst of the overall that this Japanese ethnic tendency toward hierarchy can be seen in any era. That such relationships of industrial organization, was in intimate relations with government explication is possible for those closely examining the life of contemporary Japanese people enterprise or industries divested by government, and was controlled and restricted by these. is conversely due to the fact that, as already seen, it is impossible for the historian who If hierarchy exists, it is necessary to provide a clear explanation of how we are to interpret interprets ancient history to use materials other than those pertaining to ancient history or 30 Kizaemon Ariga to apply to such interpretation, without any preliminary restrictions, the primitive periods of other peoples or contemporary primitive people. The reason that contemporary history has fundamental meaning for overall historical interpretation, as already touched upon, is related to Benedict’s cultural anthropological research methodology. Thus, it is deeply meaningful to use what is visible in contemporary Japanese life to advance the interpretation of Japanese hierarchy. Therefore, it would be well for me to plainly demonstrate what the basic characteristics of Japanese hierarchy are. My recent work has all been related to elucidating this issue, and I will here abridge the process through which I have taken contemporary life as foundation and have been offering proof of these special characteristics within Japanese history. To state my conclusion in advance, hierarchy is the character of the clan (shizoku 氏族) occupying the Japanese social structure. The word “clan,” here, while of course pointing to the unique or special Japanese clan (shizoku 氏族), is not the same as the commonly labelled “clan.”9 Japanese clans (氏族) consist of intimate hierarchical (leader-follower) relationships that emerge due to necessities of life (and thus for political purposes as well), from which develops a cohesive consciousness (dōtō ishiki 同統意識) or dōzoku consciousness (dōzoku ishiki 同族意識) mediated by the cult of local deities (ujigami 氏神) and temples (ujidera 氏寺). However, when such vertical relations collapse it is also easy for the cohesive consciousness to disappear. That is to say, a hierarchical character is latent or dormant even in the relations of equality in daily life. Because a dōtō ( 同統) consciousness is held it is fundamentally well to express this in terms of dōtō relations and as manifest in various scenes of daily life, but when the conditions defining these change, distinctions in system or form are engendered. Representative words that bear witness to this dōtō consciousness include oyako (親子 parent and child), oyakata-kokata (親方子方 parent-part and child-part), oyabun-kobun (親分子分 boss and follower), and so forth. At the time of the formation of the foundational alliances among households (ie 家) in small settlements (juraku 聚楽), this was manifest in main- (honke 本家) and branch- (either makke 末家; or bunke 分家) household relations. I refer to this alliance of households as a dōzokudan (同族団 dōzoku group) and wish to distinguish it from dōtō relations as such. Therefore, the dōzokudan is a neighborhood group (rinpo shūdan 隣保集団) in settlements based on dōtō relations. In other words, it is the fundamental village household alliance tied together by main- and makke branch-households. Makke branch-households themselves had bunke branches and new-comer households entered into main- and makke branch-household relations with influential old-comer households. Among bunke branch-households, there were both consanguineous branch households and non-consanguineous servant households. These all lived in the same settlement, and to the extent that subsistence related mutual assistance relations were formed with the main-household at the center, the makke branch-

9 Translator’s note: Ariga used here the English term “clan” in quotation marks in order to distinguish it from “the unique or special Japanese clan” which he described using the Japanese term “氏族.” Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 31 to apply to such interpretation, without any preliminary restrictions, the primitive periods of households were subservient to the main-household in all matters. As symbol of this union, other peoples or contemporary primitive people. The reason that contemporary history has the tutelary deity of the dōzokudan was worshipped in common and a common grave was fundamental meaning for overall historical interpretation, as already touched upon, is maintained at the same temple of affiliation. Among the dōzokudan of urban businessmen, related to Benedict’s cultural anthropological research methodology. Thus, it is deeply having only the latter was most common. meaningful to use what is visible in contemporary Japanese life to advance the Since dōzokudan grew with the expansion of particular enterprises, in the farming villages interpretation of Japanese hierarchy. these were centered around large land-holders, while in the cities they accompanied the Therefore, it would be well for me to plainly demonstrate what the basic characteristics of development of commercial-industrial enterprises. However, from the early modern period it Japanese hierarchy are. My recent work has all been related to elucidating this issue, and I became the common state of affairs in rural villages for landowner farming to decrease in will here abridge the process through which I have taken contemporary life as foundation scale, and since especially after Meiji landowner farming was nearly abolished, the and have been offering proof of these special characteristics within Japanese history. To establishment of non-consanguineous branch-households became rare and branch- state my conclusion in advance, hierarchy is the character of the clan (shizoku 氏族) households were almost exclusively based on consanguinity. Since the necessity for such occupying the Japanese social structure. The word “clan,” here, while of course pointing to branch-households to depend on main-households decreased, the class distinctions between the unique or special Japanese clan (shizoku 氏族), is not the same as the commonly labelled main- and makke branch-households decreased and their living relationships changed greatly “clan.”9 from those of previous periods. Japanese clans (氏族) consist of intimate hierarchical (leader-follower) relationships that For that reason, even if continuing as dōzokudan in name, most in reality became more emerge due to necessities of life (and thus for political purposes as well), from which equal in relations. Even when a particular household maintained its social status by develops a cohesive consciousness (dōtō ishiki 同統意識) or dōzoku consciousness (dōzoku ishiki maintaining the main-household tradition, most dōzokudan lost their original character of 同族意識) mediated by the cult of local deities (ujigami 氏神) and temples (ujidera 氏寺). having master and servant relationships. And, dōtō relations became inter-mixed with However, when such vertical relations collapse it is also easy for the cohesive consciousness kinship relationships, especially in rural main- and bunke branch-household relationships. to disappear. That is to say, a hierarchical character is latent or dormant even in the However, in urban areas, the opposite tendency could be seen. Commercial-industrial relations of equality in daily life. Because a dōtō ( 同統) consciousness is held it is enterprises greatly expanded since together with the development of the modern capitalist fundamentally well to express this in terms of dōtō relations and as manifest in various economy restrictions on shareholders from prior periods were lifted. This occurred on the scenes of daily life, but when the conditions defining these change, distinctions in system or basis of the dōzokudan organization, and thus unprecedentedly huge dōzokudan were form are engendered. Representative words that bear witness to this dōtō consciousness developed. For that reason, an extremely large number of non-consanguineally related makke include oyako (親子 parent and child), oyakata-kokata (親方子方 parent-part and child-part), branch-households (in western Japan called bekke 別家 separate household) were established. oyabun-kobun (親分子分 boss and follower), and so forth. In western Japan these newly independent branch establishments were called nōren-uchi (ノ At the time of the formation of the foundational alliances among households (ie 家) in オレンウチ within the shop curtain). small settlements (juraku 聚楽), this was manifest in main- (honke 本家) and branch- (either If one labels as dōzokudan those subsistence-based mutual relations among households makke 末家; or bunke 分家) household relations. I refer to this alliance of households as a formed in vertical relations, and labels as kumi (組 a group gathered to accomplish something dōzokudan (同族団 dōzoku group) and wish to distinguish it from dōtō relations as such. together) those non-subsistence-based household relations formed in mutual relations of Therefore, the dōzokudan is a neighborhood group (rinpo shūdan 隣保集団) in settlements based equality, then the foundational alliances among households in settlements may be thought on dōtō relations. In other words, it is the fundamental village household alliance tied to have taken these two forms. In the process of changes in the conditions of life for each together by main- and makke branch-households. Makke branch-households themselves had household, vertical relationships between households in kumi were engendered and it thus bunke branches and new-comer households entered into main- and makke branch-household became common to see among these equally related households the vertical relationships of relations with influential old-comer households. Among bunke branch-households, there were dōzokudan. There were also cases in which vertical relations within kumi were transformed both consanguineous branch households and non-consanguineous servant households. These into dōtō relationships, and within dōzokudan as well, with the fall of the main-household and all lived in the same settlement, and to the extent that subsistence related mutual the rise of equal relations among households, there were those that changed into kumi. assistance relations were formed with the main-household at the center, the makke branch- Such may be thought to be the inherent potential of the fundamental household alliances of settlements. This shows that it is possible to grasp the development of villages from these 9 Translator’s note: Ariga used here the English term “clan” in quotation marks in order to distinguish it from two patterns of household relationships. In one, the dōzokudan was agent of development, “the unique or special Japanese clan” which he described using the Japanese term “氏族.” 32 Kizaemon Ariga while in the other aggregate development was of farming houses of equal status. The former was a subsistence community founded on lineages of main- and branch-households, while the latter was a proto-class subsistence community (seikatsu kyōdōtai 生活共同体). But, in the villages the simple form of household affiliation based on subsistence conditions did not continue endlessly as such, as the number of households increased with development. There was both an internal increase of households and in-migration from outside. In the rise and fall of households there also arose powerful farm families and, with these at center, numerous dōzokudan coexisted. Of course, entangled with these, various kinds of kumi also existed. Dōzokudan were thus subsistence groupings (seikatsu shūdan 生活集団), and since they shared a dōtō consciousness, they naturally also were significant as political units. Corresponding to the power of these, the main-households of the dōzokudan manifest a certain social status within village life. Where the power of each main-household was equal, kumi-like mutual relations were maintained, but where the power of a particular main- household rose in prominence, the main-households of other dōzokudan within the village were placed below it. There were not a few cases in which the character of these latter mutual relationships became that of dōtō relationships. While it is not that one cannot recognize that the peasantry were severed from politics employing military arms in the contemporary or early modern periods, this was especially evident during the middle ages. I believe that the pattern whereby the main-household of a powerful dōzokudan came to exercise control over others within the mutual relations of dōzokudan was the basic structure of the clans (氏族). Thus, the expansion of the clans ran on the same rails and took the form whereby a powerful clan main-household exercised control over others within an alliance of clan houses. The significance of clan genealogies thus lie in such dōtō relations. The main- or branch-household relations of/to the local tutelary god (uji-gami 氏神) or temple (uji-dera 氏寺) were tied to this and were of the character of vertical relations of shrine and temple ranks. These had political significance and were thus not immobile but rather moved according to the relations of political control. The dōzokudan, as a basic group or collective, manifest, with the main-branch as center, class differences between main- and makke branch-households, and therefore even with the development from small to large clans the same character of hierarchy survived. This may be seen in the rise and fall of all clans from ancient times into the middle ages. Although in the early modern period this hierarchy was considerably simplified with the liquidation of mid-level feudal lords, even so the hierarchy of the warrior classes was complex. Commoners were under the immediate control of the Daimyō feudal lords but, even if the clan organization based on the mutual relations of dōzokudan—which bore the direct relations among households—cannot be seen here, they were bound in leader-subordinate relations in relation to the worship of the Daimyō household’s tutelary deity as domain deity. Commoners thus had their own dōzokudan lesser tutelary deities and the lesser tutelary deities of the larger-area villages, and furthermore also worshipped the general tutelary god Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 33 while in the other aggregate development was of farming houses of equal status. The former of the still larger area of the domain. Even though these lesser and greater deities all was a subsistence community founded on lineages of main- and branch-households, while differed, since commoners worshipped each and all with the idea that they were the the latter was a proto-class subsistence community (seikatsu kyōdōtai 生活共同体). But, in the guardian deities of each region, those smaller were included in and simultaneously villages the simple form of household affiliation based on subsistence conditions did not established with those larger and were not in contradiction to each other. continue endlessly as such, as the number of households increased with development. There In regions where the change of domain lords had not been witnessed since ancient times was both an internal increase of households and in-migration from outside. In the rise and (such as Suwa), the comprehensive tutelary deity of the territory was ceremonially fall of households there also arose powerful farm families and, with these at center, transferred (in)to lesser tutelary gods, and the same form of worship was repeated. In other numerous dōzokudan coexisted. Of course, entangled with these, various kinds of kumi also territories with much political change, there were times when this sort of ceremonial existed. transfer of the domain’s general tutelary deity to the villages occurred, but in general the Dōzokudan were thus subsistence groupings (seikatsu shūdan 生活集団), and since they pattern in which the domain lord kept the ancient shrine deities safe was common. shared a dōtō consciousness, they naturally also were significant as political units. However, the multifariousness of the shrine deities may conversely narrate if not explain Corresponding to the power of these, the main-households of the dōzokudan manifest a political change. The Daimyō houses and the commoners were separated by class certain social status within village life. Where the power of each main-household was equal, distinctions, and as they were bound together by ritual in this form there were in this sense kumi-like mutual relations were maintained, but where the power of a particular main- hidden here the characteristics of being in dōtō relations. This was also visible in the household rose in prominence, the main-households of other dōzokudan within the village relationship between the Emperor and the people after the Meiji Restoration, which had were placed below it. There were not a few cases in which the character of these latter swept from political organization the Shōgun, Daimyō and warrior clans. mutual relationships became that of dōtō relationships. While it is not that one cannot I may face the reprobation of many others for claiming thusly that such is the character of recognize that the peasantry were severed from politics employing military arms in the clans, but, without such continuity in the characteristics of clans in Japan’s social structure, contemporary or early modern periods, this was especially evident during the middle ages. the belief which has the Grand Shrine of Ise as national tutelary deity would be untenable. I believe that the pattern whereby the main-household of a powerful dōzokudan came to Here, the private tutelary deity of the main-household of a dōzokudan may be seen as the exercise control over others within the mutual relations of dōzokudan was the basic structure public tutelary deity of the whole dōzokudan, and has the same character as when the private of the clans (氏族). Thus, the expansion of the clans ran on the same rails and took the form tutelary deity of a clan main-household is thought of as the public tutelary deity of the clan. whereby a powerful clan main-household exercised control over others within an alliance of Originally, rituals were conducted by the main-household as ritual head (of the parish guild clan houses. The significance of clan genealogies thus lie in such dōtō relations. The main- or called “miyaza” 宮座), and the people affiliated with these livelihood groupings as well—even branch-household relations of/to the local tutelary god (uji-gami 氏神) or temple (uji-dera 氏寺) if they were not directly involved in the ritual as such—also performed such worship. It were tied to this and were of the character of vertical relations of shrine and temple ranks. must be recognized that the rituals of the Grand Shrine of Ise are of the same nature. These had political significance and were thus not immobile but rather moved according to Next it is necessary to examine the composition of other parts of society that coexisted the relations of political control. The dōzokudan, as a basic group or collective, manifest, with with these relationships. It is possible to observe a very great number of household-centered the main-branch as center, class differences between main- and makke branch-households, dōzokudan in various industries, and those powerful families among them were proud of their and therefore even with the development from small to large clans the same character of ancient lineages and possessed household genealogical charts. However, many of these hierarchy survived. This may be seen in the rise and fall of all clans from ancient times into genealogies recorded that they drew on lineages of the ancient nobility, thought to include the middle ages. descendants of the Imperial Household. The character of these genealogies is that of dōtō Although in the early modern period this hierarchy was considerably simplified with the relationships, and does not necessarily prove consanguineous relationships. This is because liquidation of mid-level feudal lords, even so the hierarchy of the warrior classes was even adoptees of no blood relation who served as vassals and established makke branch- complex. Commoners were under the immediate control of the Daimyō feudal lords but, even households could succeed to such lineages. Even the nameless families of commoners if the clan organization based on the mutual relations of dōzokudan—which bore the direct without lineage charts possessed genealogical charts of the same character as the main- relations among households—cannot be seen here, they were bound in leader-subordinate household of their dōzokudan or clan of affiliation. This was nothing more than the possession relations in relation to the worship of the Daimyō household’s tutelary deity as domain deity. of low-grade genealogies. Commoners thus had their own dōzokudan lesser tutelary deities and the lesser tutelary I thus wish to turn to an examination of the dōzokudan of urban businessmen. As already deities of the larger-area villages, and furthermore also worshipped the general tutelary god noted, the Japanese capitalist economy developed from family (ie 家) enterprises. The custom 34 Kizaemon Ariga of taking on apprentices, training them as clerks and setting them up as bekke (別家) branch- households (employee bunke [雇人分家] branch-households) existed from before the Edo Period. After the Meiji Restoration, restrictions on trade guilds were abolished and business families aimed for the unlimited growth of business. At the beginning of such development, large business owners, increasingly incorporating employees as commuting bekke branch- houses, wholeheartedly pursued (dōzokudan) expansion. When the business grew large, living quarters and shops were separated. Together with the expansion of their businesses, it became impossible to incorporate employees in the form of bekke branch-houses, and thus large numbers of employees with no relationship to the business owners’ households were employed. Employee-based bekke branch-houses had genealogical relations with the main- household and its consanguineous branch-households. Their cooperative relations in everyday life were extremely close, they shared temples and grave sites, and the main- branch protected the bekke branch-households in many respects. However, in workplaces without such household genealogical relationships, the familial/household based relations of private life became weakened, and vertical relations between owners and employees were formed within workplace organizations. But since even here there were close relationships of living, relationships similar to those of the dōzokudan of main- and makke branch-households were born. Even in the cities the main- and makke branch-household relations in dōzokudan were called oyakata (parent-part) and kogai (child- part) and were thought of as oya-ko (parent-child) or oyabun-kobun (boss-follower). In workplaces without dōzokudan, even without relations to tutelary deities or temples, it is well to say that, since these relations were thought of as those of oyakata-kokata, they were dōtō relations. That these relationships (of oyakata-kokata) have been observed in various contemporary workplaces is noteworthy. When one examines household enterprises that grew into large consortiums, one finds dōzokudan at their center and so-called “family corporations” (dōzoku gaisha 同族会社) directing various affiliated businesses. Looking at these large corporations, one sees that in the central dōzoku extended family, since there were main-households and consanguineal bunke branch-households—both of which possessed their own employee-based branch-households— these took the form of extended dōzokudan unions (同族団連合), that is, of clans (氏族). These main- and consanguineally-related bunke branch-households were the stockholders of the main business, and they did not themselves become directors of affiliated companies, management of which was entrusted to powerful chief executives from the main business. And, the main-office executives and main officers of the affiliated firms were incorporated into consanguineal bunke or into bekke branch-households of the main-household and thus were members affiliated with the dōzokudan or clan. The home office controlled the capital and personnel of the directly affiliated businesses, and in personnel there was a strong tendency toward dōtō relations. Regarding the capital of lesser subsidiary affiliated companies, it generally sufficed for the home office to hold approximately 10 percent of stock. Personnel [of the lower-tier Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 35 of taking on apprentices, training them as clerks and setting them up as bekke (別家) branch- companies] were controlled by the directly affiliated firms, and so while thus indirectly, households (employee bunke [雇人分家] branch-households) existed from before the Edo actual managerial power lay with the main office. These all manifest the same character of Period. After the Meiji Restoration, restrictions on trade guilds were abolished and business hierarchy, which was also apparent in the relations between company officers and families aimed for the unlimited growth of business. At the beginning of such development, employees, and between factory foremen and workers. Thus, looking at the whole, there large business owners, increasingly incorporating employees as commuting bekke branch- were dōzokudan as the core, there was the bonding of clans in such groupings, and in those houses, wholeheartedly pursued (dōzokudan) expansion. When the business grew large, living areas unable to organize as dōzokudan, organizations expanding dōtō relations or similar quarters and shops were separated. Together with the expansion of their businesses, it relations were created. And thus, the economically powerful zaibatsu all had relationships of became impossible to incorporate employees in the form of bekke branch-houses, and thus the same character as the many smaller businessmen who had to depend on them for large numbers of employees with no relationship to the business owners’ households were business. employed. Employee-based bekke branch-houses had genealogical relations with the main- These were internal relations of single zaibatsu and of enterprises related to them, and household and its consanguineous branch-households. Their cooperative relations in both large and small zaibatsu with such internal structure stood in relations of opposition, as everyday life were extremely close, they shared temples and grave sites, and the main- well as association, with each other. There were a limited number of huge zaibatsu standing branch protected the bekke branch-households in many respects. in rivalry with each other, but smaller enterprises always existed in association with larger However, in workplaces without such household genealogical relationships, the ones. This is the same thing as large clans existing in rivalry. The giant zaibatsu linked with familial/household based relations of private life became weakened, and vertical relations political parties and confirmed their social status. between owners and employees were formed within workplace organizations. But since even This of course is connected with the nature of the political revolution of the Meiji here there were close relationships of living, relationships similar to those of the dōzokudan of Restoration, which was a revolt carried out by an alliance of powerful domains in main- and makke branch-households were born. Even in the cities the main- and makke denunciation of the political responsibility of the Tokugawa Bakufu government, and in branch-household relations in dōzokudan were called oyakata (parent-part) and kogai (child- which lie the character of clan conflict. It has previously in general been said that this was a part) and were thought of as oya-ko (parent-child) or oyabun-kobun (boss-follower). In revolt by the lower-rank warriors or that it was a class-warfare type revolution based on the workplaces without dōzokudan, even without relations to tutelary deities or temples, it is well alliance of lower-rank warriors and the merchant class. However, the classes that controlled to say that, since these relations were thought of as those of oyakata-kokata, they were dōtō the leadership also all initially formed Hantō (藩黨) domain political parties (with Daimyō relations. That these relationships (of oyakata-kokata) have been observed in various clans as base) and began their actions. Most of the Daimyō themselves were not the object of contemporary workplaces is noteworthy. the revolution and were able to garner powerful positions in Meiji society. The Shōgunal When one examines household enterprises that grew into large consortiums, one finds families and the Daimyō families lost the status they held as warriors in the warrior-based dōzokudan at their center and so-called “family corporations” (dōzoku gaisha 同族会社) directing society [of pre-Meiji Japan], but since they retained part of their resources, they were able to various affiliated businesses. Looking at these large corporations, one sees that in the attain high status in the new economy and society by joining with the politicians of the Meiji central dōzoku extended family, since there were main-households and consanguineal bunke Government or with the financial capitalists related to them. Because they politically branch-households—both of which possessed their own employee-based branch-households— occupied seats in the upper House of Peers, economically became capitalists and socially these took the form of extended dōzokudan unions (同族団連合), that is, of clans (氏族). These became the aristocracy, in Meiji society their status was second only to that of the Emperor. main- and consanguineally-related bunke branch-households were the stockholders of the To explain the status of the Emperor in post-Meiji Restoration society requires that one main business, and they did not themselves become directors of affiliated companies, investigate Japanese social characteristics from the Nara Period [710-794], and such management of which was entrusted to powerful chief executives from the main business. detailed explanation is not possible here. The Emperor’s status is related to the fact that the And, the main-office executives and main officers of the affiliated firms were incorporated characteristics of clans have been retained in Japanese social structure, and in particular into consanguineal bunke or into bekke branch-households of the main-household and thus the dual-structure of politics is an expression of this. Since under the clan system the were members affiliated with the dōzokudan or clan. The home office controlled the capital separation of ritual and politics was impossible, those who ruled were of necessity those who and personnel of the directly affiliated businesses, and in personnel there was a strong worshipped the deities. There was rule by oracles. For that reason, there were complex tendency toward dōtō relations. rituals for worshipping the deities. Because the Emperor was not a deity but one who prayed Regarding the capital of lesser subsidiary affiliated companies, it generally sufficed for the to the gods, and since to worship the deities required that one be a holy person, he home office to hold approximately 10 percent of stock. Personnel [of the lower-tier underwent rigorous purification. The concept of deity ( 神) as well differs from the 36 Kizaemon Ariga

Christian “God,” and is extremely humanistic, and so too the meaning of holy differed. Once the ritual was completed, the ritualist returned to being a lay person and then carried out political rule. Because in the Nara Period political governance became complicated, the combination of ritual and politics gradually became difficult to maintain. This is shown by the fact that the Department of Worship (Jingikan 神祇官) and the Department of State (Dajōkan 太政官) were divided according to the Taihō Code (Taihōryō 大宝令), and gradually national sovereignty was differentiated, wherein the Emperor appointed those responsible for political governance beneath him and he himself specialized in religious ritual. However, this did not fully free the Emperor from politics as, being the head priest of the national tutelary deities, he held authority as political stabilizer, and the political status of those who were in political positions of responsibility was maintained with this as background. This became especially clearly apparent when political leaders were endangered. This trend also became manifest among the great clans when peace prevailed. In the Tokugawa Bakufu as well, the Shōgun was internally in the position of political stabilizer while state affairs were carried out by powerholders from among the hereditary vassal (fudai 譜代) clans. This trend was also manifest among the Daimyō families, and was seen as well in the relations between main-household and headquarter executives in the large and Sumitomo zaibatsu after Meiji. Below these, this was also a frequently seen trend among even the main-households of other comparatively large dōzokudan. One must consider these in terms of the relations among the roles held on ritual occasions at tutelary shrines or for clan tutelary deities, and as such it is possible to grasp these as tendencies showing the character of clans. As noted above, the Meiji Restoration was a political revolt carried out by an alliance of powerful domains in denunciation of the political responsibility of the Tokugawa Bakufu government. However, the Emperor was supported because he was the only person in a position to act as political stabilizer. This was the reason that, in keeping with the character of Japanese social organization, the Emperor was seen as someone who could continue as though in the position of main-household (head) of the nation as a whole. Thus, to this singularly supreme position were given modern meanings and the contemporary Emperor system was established. If this is the case, then the character of clans courses through it. Benedict compared this with the sacred chiefs of Tonga and Samoa. This is interesting as comparative research. However, when she asserts that the Emperor was politically without power, since this is limiting politics to an extremely narrow meaning, she somewhat belittles the meaning of the Emperor occupying the crowning point of hierarchy. With this, it is impossible to understand the meaning of the Emperor continuing as political stabilizer after the war. Among the Japanese soon after the war there were many theses on the unnecessariness of the Emperor. It is a fact that the Emperor’s position today as political stabilizer is based largely on the support of the Occupation army, but it is not simply a matter of that. Even in those periods when the Emperor did not possess military strength, Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 37

Christian “God,” and is extremely humanistic, and so too the meaning of holy differed. Once his social position did not fall below those of the warriors. Even if one exclaims theses of the ritual was completed, the ritualist returned to being a lay person and then carried out imperial irrelevance, to the extent that he has power to act as political stabilizer, the political rule. Emperorship must be allowed to continue. I perceive this from the character of Japanese Because in the Nara Period political governance became complicated, the combination of society and politics. ritual and politics gradually became difficult to maintain. This is shown by the fact that the Summarizing in simple fashion extremely complicated social phenomena is difficult, but Department of Worship (Jingikan 神祇官) and the Department of State (Dajōkan 太政官) were as I do not have space to further describe this problem, I wish to sum up and point out that divided according to the Taihō Code (Taihōryō 大宝令), and gradually national sovereignty within various types of occupation groups, when for subsistence vertical relations are closely was differentiated, wherein the Emperor appointed those responsible for political bound, then oyakata-kokata relations (dōtō relations) have arisen. This has been observed in governance beneath him and he himself specialized in religious ritual. However, this did not political parties, the military, government agencies, groups of artists, miners and many fully free the Emperor from politics as, being the head priest of the national tutelary deities, other social groups. That such is also seen among racketeers and gamblers is also well he held authority as political stabilizer, and the political status of those who were in political known. When dōtō relations become strong, private familial relations may be internally positions of responsibility was maintained with this as background. engendered, and there are cases in which they become almost equivalent to dōzokudan. This became especially clearly apparent when political leaders were endangered. This Because dōtō relations and dōzokudan were intimately related, there were cases in which trend also became manifest among the great clans when peace prevailed. In the Tokugawa dōzokudan were established on dōtō relations, as well as cases in which this did not occur. Bakufu as well, the Shōgun was internally in the position of political stabilizer while state Just as when dōzokudan came to overlap hierarchically, if the occupational group was large, affairs were carried out by powerholders from among the hereditary vassal (fudai 譜代) clans. then within it dōtō relations could become vertically ranked by size. The kokata of a high This trend was also manifest among the Daimyō families, and was seen as well in the ranking oyakata could become a lower-ranking oyakata with his own kokata. However many relations between main-household and headquarter executives in the large Mitsui and strata there might be, the hierarchical ordering was clear if the dōtō relations were of the Sumitomo zaibatsu after Meiji. Below these, this was also a frequently seen trend among even same lineage, but if independent dōtō relations were involved, social rank was determined by the main-households of other comparatively large dōzokudan. One must consider these in the relative power of the respective oyakata. This was the same regarding the main- terms of the relations among the roles held on ritual occasions at tutelary shrines or for clan households of dōzokudan. For example, within the same dōzokudan, the main-household of that tutelary deities, and as such it is possible to grasp these as tendencies showing the character dōzokudan stood at the head, but when comparisons were made with the pedigrees or fortunes of clans. of the main-households of other dōzokudan then the rank-ordering of social position was As noted above, the Meiji Restoration was a political revolt carried out by an alliance of determined accordingly. powerful domains in denunciation of the political responsibility of the Tokugawa Bakufu Even where occupational groupings differed, a similar standard was used in comparing government. However, the Emperor was supported because he was the only person in a and determining household status (kakaku 家格). For example, assume that an agricultural position to act as political stabilizer. This was the reason that, in keeping with the character household and a merchant household were compared. While the roots of the household of Japanese social organization, the Emperor was seen as someone who could continue as status of each lay in the rank-ordering of lineal relations of main- and branch- households though in the position of main-household (head) of the nation as a whole. Thus, to this within each respective dōzokudan, when the two households were compared and ranked singularly supreme position were given modern meanings and the contemporary Emperor relative to the hierarchy of dōzokudan or clans, a makke branch-household of a higher-ranking system was established. If this is the case, then the character of clans courses through it. clan would be accorded the same status as the main-household of a lower-ranking clan. Benedict compared this with the sacred chiefs of Tonga and Samoa. This is interesting as Household status within these hierarchies was the grounds for intermarriage and it was comparative research. However, when she asserts that the Emperor was politically without usual for formal marriages to be made between households of the same status. However, it is power, since this is limiting politics to an extremely narrow meaning, she somewhat belittles believed that when someone from a lower-ranking household came to have ability they the meaning of the Emperor occupying the crowning point of hierarchy. With this, it is attempted to intermarry with someone from a higher-ranked household and thus raise their impossible to understand the meaning of the Emperor continuing as political stabilizer after own status; and that when a high-status household fell into ruin, their social status was the war. Among the Japanese soon after the war there were many theses on the lowered by intermarriage with a household of lesser status. Thus, household status was not unnecessariness of the Emperor. It is a fact that the Emperor’s position today as political fixed and unmovable but could shift up or down with each generation. This is one of the stabilizer is based largely on the support of the Occupation army, but it is not simply a fundamental characteristics of clan hierarchy. matter of that. Even in those periods when the Emperor did not possess military strength, 38 Kizaemon Ariga

As dōtō relations were the foundation of dōzokudan, even those oyakata-kokata relations not having dōzokudan were compared with other oyakata-kokata relations and their relative abilities of livelihood and politics. That the hierarchical status of dōtō relations was thus determined was no different than in the cases of dōzokudan. In many and various occupational groups today, the oyakata-kokata nomenclature is no longer used, although in some special groups such as racketeers, gamblers, and so forth, it is still clearly employed, and in agricultural villages is still widely used. And, while such nomenclature is not obviously employed in government bureaucracies, the military or companies established in modern times, it exists still in everyday consciousness. These days, there is frequent use not of oyabun but of the English word “boss” (ボス) to convey critical connotations, but such is done within the context of the everyday awareness of oyabun as being synonymous. Even if not spoken as such, that it is possible to view these words as connecting the status retained in public workplaces with the social status of one’s household is due to the fact that underlying these is the system of clan hierarchy. That is, the occupational status of the military or the government were connected with court rank, order of merit, and so forth, and there was a strong tendency to socially express family/household ranks with these. All classes or ranks of the aristocracy were tied to family/household, and it was the custom to thus indicate the social status of all those affiliated with a particular household. The court ranks or orders of merit granted by the Emperor were, together with the various occupational groups, hierarchically ranked, and there was a strong tendency for the ranks of individuals’ households to also be so ordered. And, there was a deep-running tendency for rank-order in the social hierarchy among the lower classes not affiliated with court ranks or orders of merit to be based on the dōzokudan or the dōtō relations to which one’s family/household was bound. I would be happy if the reader would refer to my other work regarding these issues. I believe that we must even more rigorously investigate and come to an understanding of the issue of hierarchy in Japanese social structure that Benedict has pointed to.

3. Benedict suggests that the mentality of the Japanese emerged from hierarchy. This is a very interesting interpretation. She created a table of “Japanese Obligations and Their Reciprocals” [see page 116 in Benedict’s original]10 from this point of view. This issue is too complex, but Benedict has grasped important points. Because the social organization of a people is something that they have created based on their conditions of life, it is natural that their views of life or worldviews are prescribed by it. And, the fewer the opportunities these people have to observe the outside world the greater their limitations. For this reason, Benedict was correct to search for the mentality of the Japanese within Japanese hierarchy. Mentality must be understood psychologically, but it is also important to grasp it sociologically.

10 Ruth Benedict, The Chrysanthemum and the Sword: Patterns of Japanese Culture. Boston: Houghton Mifflin, 1946. Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 39

As dōtō relations were the foundation of dōzokudan, even those oyakata-kokata relations not If one re-organizes Benedict’s table of “Japanese Obligations and Their Reciprocals,” there having dōzokudan were compared with other oyakata-kokata relations and their relative are things that occur in vertical/hierarchical human relationships and those that occur in abilities of livelihood and politics. That the hierarchical status of dōtō relations was thus equal human relationships. Benedict has gimu ( 義務) obligations as those of vertical determined was no different than in the cases of dōzokudan. In many and various relationships and giri (義理) obligations as those of equal relationships, but this is mistaken. occupational groups today, the oyakata-kokata nomenclature is no longer used, although in Japanese people are said to have incurred and reciprocated giri obligations in both cases. some special groups such as racketeers, gamblers, and so forth, it is still clearly employed, However, in vertical relations in particular, on (恩) debts of obligation were incurred and and in agricultural villages is still widely used. And, while such nomenclature is not reciprocated and were a type of giri, and if one did nothing regarding the on incurred then obviously employed in government bureaucracies, the military or companies established in one failed in his giri duty. modern times, it exists still in everyday consciousness. These days, there is frequent use not Benedict also believes that there is on in equal relationships, but without the formation of of oyabun but of the English word “boss” (ボス) to convey critical connotations, but such is some kind of vertical/hierarchical relation there was no such thing. As Benedict says, on was done within the context of the everyday awareness of oyabun as being synonymous. Even if conceived as received from one’s superiors and as unrepayable. But this prioritized as not spoken as such, that it is possible to view these words as connecting the status retained superiors those of higher rank in a hierarchy. This is apparent in cases where service was in public workplaces with the social status of one’s household is due to the fact that rendered in repayment of on. Thus, in cases of chū (忠) duties to the Emperor and kō (孝) underlying these is the system of clan hierarchy. duties to one’s parents, the former was given priority. In cases involving dōzokudan or clans That is, the occupational status of the military or the government were connected with as well, if there was a higher-ranked superior above the subject, priority was given to that court rank, order of merit, and so forth, and there was a strong tendency to socially express superior. family/household ranks with these. All classes or ranks of the aristocracy were tied to The concepts of chū and kō are distinguished, based on their borrowing from the Chinese family/household, and it was the custom to thus indicate the social status of all those language, but it may be thought that Japanese people’s original concepts did not make this affiliated with a particular household. The court ranks or orders of merit granted by the distinction in meaning. In the case of main- and makke branch-households of a dōzokudan, the Emperor were, together with the various occupational groups, hierarchically ranked, and children of the branch-households had to prioritize their giri duties to the main-household there was a strong tendency for the ranks of individuals’ households to also be so ordered. and to postpone their giri duties to their own parents. These both contained set giri And, there was a deep-running tendency for rank-order in the social hierarchy among the relationships. That is, it was believed that there was an on-gaeshi (恩返し) duty to repay on lower classes not affiliated with court ranks or orders of merit to be based on the dōzokudan debts of obligation and that the former service to repay was limitless. This distinction was or the dōtō relations to which one’s family/household was bound. I would be happy if the none other than that of the prioritization of superiors. reader would refer to my other work regarding these issues. I believe that we must even In , kō involved the limitless service due to one’s consanguineal ancestors or parents, more rigorously investigate and come to an understanding of the issue of hierarchy in and this became the cardinal moral code for the sake of clan (sōzoku 宗族) bonding. In Japan, Japanese social structure that Benedict has pointed to. consanguineal ancestors were worshipped, but this was a private observance concerning only a given family/household. Among the events of their dōzokudan or of related households, 3. Benedict suggests that the mentality of the Japanese emerged from hierarchy. This is a most important was the worship of the dōzokudan’s tutelary deity. This originally was the very interesting interpretation. She created a table of “Japanese Obligations and Their private tutelary deity of the main-household of the dōzokudan but became simultaneously the Reciprocals” [see page 116 in Benedict’s original]10 from this point of view. This issue is too public tutelary deity of the dōzokudan as a whole. The central actor in rituals for these complex, but Benedict has grasped important points. Because the social organization of a tutelary deities was the head of the main-household. That members of makke branch- people is something that they have created based on their conditions of life, it is natural that households had first to pay service to the main-household head was because the existence of their views of life or worldviews are prescribed by it. And, the fewer the opportunities these the branch-households was originally made possible by the guardianship of the main- people have to observe the outside world the greater their limitations. For this reason, household and because as long as the main-household held power it was leader of the Benedict was correct to search for the mentality of the Japanese within Japanese hierarchy. dōzokudan. Mentality must be understood psychologically, but it is also important to grasp it The same is true regarding butsudan ( 仏壇) family altars. There, worship of the sociologically. genealogical (keifu-jō-no 系譜上の) ancestors who were the roots of the family’s main- and 血縁の branch-household relations was prioritized over the direct consanguineal (ketsuen no ) 10 Ruth Benedict, The Chrysanthemum and the Sword: Patterns of Japanese Culture. Boston: Houghton Mifflin, 1946. ancestors of the makke branch-household. The genealogical ancestors were called oya, but for 40 Kizaemon Ariga branch-households these were not limited to consanguineal ancestors. Previously there were many who thought that these were consanguineal ancestors but, while this may have been the case for main-households, ancestors for (especially non-consanguineal) makke branch- households were not limited to being consanguineous. There were many cases where, despite being the main-household of a certain dōzokudan, a family was established as the makke branch-household of another clan within the internal clan organization standing above it. Thus, if the tutelary deities they worshipped were also those worshipped by other superior dōzokudan main-households, then it is certainly not likely that such deified ancestors were a particular family’s own consanguineal ancestors. Since to be a tutelary deity did not require that one be a consanguineal ancestor, and as the concepts of consanguinity or non-consanguinity were not fixed, this had meaning as genealogical source (keifu-teki hongen 系譜的本源). Otherwise, it becomes impossible to understand the meaning of worshipping as tutelary deities ancestors completely unrelated by blood among dōzokudan or clans. Thus that kō was prioritized in China while chū was prioritized in Japan derives from the fundamental social organizational character of each society. Given this, to comprehend the fact that chū obligations toward the Emperor were prioritized, one must understand the character of Japanese social organization. However, that during the warrior period chū obligations toward the Emperor were not necessarily prioritized is due to the fact that warriors became intermediating regional lords and the Emperor was not directly related to the people’s everyday lives. It is important to know the changed historical reality that came with the fall of government. In giri obligations incurred in relations of equality, it sufficed to return the approximately equivalent amounts of such debt. As on is not created at all in such cases, there was no incurring of on from colleagues as Benedict suggests. The incurring of on debts between colleagues is a greatly exaggerated expression marking their inter-personal relationship as vertical in nature. Benedict further lists giri to one’s name [see page 116 in Benedict’s original], and explains that this indicates all those actions to restore one’s honor when such has been harmed. However, this is not giri. While their actions may have involved the taking of oaths and revenge against their enemy, the revenge of the Akō Rōshi forty-seven samurai cannot be thought of as giri to one’s name. That they—reading the mind of their leader, Lord Asano, who was unable to himself retaliate against the insult done him—took revenge against their lord’s enemy was the result of their thoughts to repay the on debts of obligation (on-gaeshi 恩 返し) received over many years from their lord and so in other words was a matter of chū duty. The thought simultaneously arises that the insult that their lord received was at the same time a disgrace or shame (haji 恥) upon themselves as his vassals, but one cannot say that the vindicating of their names (haji o susugu 恥をすすぐ) was a giri duty. Benedict does not explain using hierarchy either just why the on received from a superior was unrepayable or why superiors were prioritized. In the contemporary period, the old-age Kizaemon Ariga The Issue of Hierarchy in Japanese Social Structure 41 branch-households these were not limited to consanguineal ancestors. Previously there were pension a retired government bureaucrat receives has been called on-kyū (恩給). However, many who thought that these were consanguineal ancestors but, while this may have been this word was visible in the relations of lord and vassal during the warrior period in the the case for main-households, ancestors for (especially non-consanguineal) makke branch- system wherein a warlord would distribute part of his domain as pension. This division of households were not limited to being consanguineous. fief lands existed hierarchically, and a vassal would also divide small parts of his lands for There were many cases where, despite being the main-household of a certain dōzokudan, a his own retainers. family was established as the makke branch-household of another clan within the internal Such lord-vassal like relations existed in all classes, and at the lowest level involved those clan organization standing above it. Thus, if the tutelary deities they worshipped were also of landowner and tenant. During the medieval period tenant farmers were affiliated with those worshipped by other superior dōzokudan main-households, then it is certainly not likely landowners’ dōzokudan, and at the same time landowners were affiliated with the clan of that such deified ancestors were a particular family’s own consanguineal ancestors. Since to their headmen, with lower ranking landowners affiliated with the clans of higher ranking be a tutelary deity did not require that one be a consanguineal ancestor, and as the concepts landowners. In the bonding of such lord-vassal relationships, vassals carried out service for of consanguinity or non-consanguinity were not fixed, this had meaning as genealogical their lords, while lords parceled out parts of their lands as pensions for their retainers. In source (keifu-teki hongen 系譜的本源). Otherwise, it becomes impossible to understand the the early-modern period intermediary landowners were dispensed with and landowners meaning of worshipping as tutelary deities ancestors completely unrelated by blood among became directly attached to the Daimyō. In the relations between landowners and tenants, dōzokudan or clans. there were those binding dōzokudan, and thus the parceling of lands was also to be observed. Thus that kō was prioritized in China while chū was prioritized in Japan derives from the However, even when not involving dōzokudan based relations, because the tenant would have fundamental social organizational character of each society. Given this, to comprehend the been granted land registration (ownership rights) from the landowner, this too had the fact that chū obligations toward the Emperor were prioritized, one must understand the character of being a division of lands to retainers. Whether the tenant rendered service in character of Japanese social organization. However, that during the warrior period chū the form of labor or of produce, such would have followed the notion of on. obligations toward the Emperor were not necessarily prioritized is due to the fact that Because land registration (nauke 名請) here is a different word for receipt of on (ontakauke 恩 warriors became intermediating regional lords and the Emperor was not directly related to 高請), it also referred to the direct receipt of favor from a Daimyō. This as well could be the people’s everyday lives. It is important to know the changed historical reality that came repaid as taxes in kind or in the form of corvée labor and involved the concept of on, which with the fall of samurai government. was especially clear among the warrior classes. Initially there was division of land as In giri obligations incurred in relations of equality, it sufficed to return the approximately pension for kyūnin (給人) high-ranking retainers. This was later changed to rice stipends, but equivalent amounts of such debt. As on is not created at all in such cases, there was no this was still a form of on-kyū (恩給) pension, and was established hierarchically throughout incurring of on from colleagues as Benedict suggests. The incurring of on debts between the organization of Daimyō clans. colleagues is a greatly exaggerated expression marking their inter-personal relationship as Thus, on was tied to hierarchy and involved each and all rewards from lord to vassal, vertical in nature. whether spiritual or material. To the extent that such lord-vassal relationships continued to Benedict further lists giri to one’s name [see page 116 in Benedict’s original], and explains exist, on was conceived of as unrepayable during the retainer’s lifetime, or even during the that this indicates all those actions to restore one’s honor when such has been harmed. lifetimes of his descendants. This is not something that grew out of warrior society but may However, this is not giri. While their actions may have involved the taking of oaths and be thought to be yet older, having emerged from the character of clans. The warrior society revenge against their enemy, the revenge of the Akō Rōshi forty-seven samurai cannot be is special only in its guarding of lord-vassal relations through the use of powerful rights of thought of as giri to one’s name. That they—reading the mind of their leader, Lord Asano, dominion and jurisdiction. Even in the modern period, the “mentality” that feels the receipt who was unable to himself retaliate against the insult done him—took revenge against their of life-wages from superiors to be a form of on strongly continues to exist. What was clearly lord’s enemy was the result of their thoughts to repay the on debts of obligation (on-gaeshi 恩 referred to as on-kyū pensions were limited to the old-age pensions referred to above, but in 返し) received over many years from their lord and so in other words was a matter of chū the leader-follower relations of various occupation groups this tendency to feel on in wage duty. The thought simultaneously arises that the insult that their lord received was at the payments was strong. One may consider this to be the social foundation of the concept of on. same time a disgrace or shame (haji 恥) upon themselves as his vassals, but one cannot say However, as noted above, hierarchy arose from the intimate vertical relations born of that the vindicating of their names (haji o susugu 恥をすすぐ) was a giri duty. everyday life. Because of the fact that when such vertical relations collapsed so too did Benedict does not explain using hierarchy either just why the on received from a superior feelings and relations of on, one must recognize that it was giri that regulated such actors’ was unrepayable or why superiors were prioritized. In the contemporary period, the old-age mutual gimu obligations and duties. Even if one can observe neither dōtō relations nor 42 Kizaemon Ariga dōzokudan in these equal relationships, considering the point that hierarchy of the same character was engendered when they changed into vertical relations, one must see that the same hierarchical character lay latent in such equal relationships. In the general progress of post-Meiji Restoration Japan, large-scale hierarchies developed within the internal organizations of various enterprises, but together with this there also arose many areas of equal relations not affiliated such hierarchy. I believe that this is one cause for Benedict’s thoughts about the existence of duality. However, even if these two [vertical and equal relations] existed separately within the same stratum, such was not just a relationship that led to mutual transformations when prior defining conditions changed but, even remaining as such, were affiliated with it as a part of an even larger organization, and in such cases even relations of equality were fundamentally marked with the character of hierarchy. It follows that in these complex relationships were manifest the characteristic tendencies of the Japanese. These days, equal relations have been very often manifest in policies to break down vertical social relationships, but since hierarchy only lies hidden in such relations of equality, they contain the potential to again transform into vertical hierarchical relations should the relevant defining conditions change. Such is already apparent in the political parties and union movement of today. Many of the equal relationships that may be seen today are unhealthy and unstable because they are beleaguered by many life conditions constraining the establishment of individuality. For this reason, should the conditions of life change it will be easy for these to instantly enter into unhealthy dōtō relations. There are people who believe that democracy exists only in equality, but democracy is not possible in an equality that has no true freedom. Investigating why the hierarchy of Japanese clans developed is difficult, but it is certain that hierarchy did not just arise from the single reason that the lives of the Japanese were manifestly those of poverty. However, for democracy to thrive, it is certain that these conditions must be conquered and that individual life must be firmly established. Because the issue of hierarchy is intertwined with a great many other issues, understanding it is not something that can be completed with the simple depiction presented here.

(10 September 1949)