Search the Web Search ejcjs

How to contribute to ejcjs

'Religion' and 'the Secular' in Japan

Problems in history, social anthropology, and the study of religion[1]

by

T. Fitzgerald

About the Author

In a recent discussion of the idea of tradition, R. J. Smith said:

…the rituals performed at weddings and funerals today contain many of the core symbols – and are designed to embody some of the most deeply held convictions – of the members of any society.

(Smith, 1995: 29)

A similar thing might have been said about rituals performed throughout the school system, which, like weddings and funerals, can be usefully analysed in terms of rites of passage and the dialectics of structure and communitas (Turner, 1969: 81-2). Yet in his discussion of the wedding performance described so well by Edwards (1989) Smith refers to the Shinto part of the ceremony as “a ‘religious’ ceremony” (Smith’s own inverted commas round ‘religious’) and the reception that follows as a “highly secular ritual” (1995: 28). How much weight should be placed on this characterisation? In the context of Smith’s article, I can only suppose very little. As a reader it occurred to me only as a strange blip in an otherwise nicely crafted meditation on the problem of tradition. Its interest lies in that it seems to add nothing at all to our understanding. This reader can only guess that its inclusion was an unconscious reflex, a small and relatively inconsequential ritual genuflection to a categorical imperative that might stray into the text of even the most methodologically sophisticated writer.

I assume that the Shinto part of the ceremony is associated with religion because, well, because we all know that Shinto is a religion. There are gods in it. But why describe the party afterwards as highly secular? Edwards has clearly shown that it is at the party that many of the core symbols and deeply held convictions, including the subordination of individual autonomy and the hierarchical interdependence of all the actors, are given powerful expression.

Edwards appeals to Victor Turner’s (1967, 1969) version of Van Gennep’s rites of passage in his discussion of the wedding as the liminal phase in the transformation of the principals from one kind of social actor into another (1989: 102-3). The rituals transform the principal actors, they also transform the wider society’s cognition of them, and finally they instruct both the principal actors and the participants in what is expected of marriage.

It is interesting that Turner himself describes these statuses, and their ritual transformation, as having a sacred component, and he denies that the distinction between structure and communitas is the same as that between secular and sacred (1969: 82). True, Turner himself uses the term secular rather loosely at times, as when he says that threshold people have no “secular clothing”, or when he says that in the liminal phase “Secular distinctions of rank and status disappear or are homogenized…” (Turner, 1969: 81); and again when he refers to “a ‘moment in and out of time’, and in and out of secular social structure” (Turner, 1969: 82). It seems then that Turner himself does use the terms loosely from time to time, and without any apparent theoretical gain. Indeed one might say there is a contradiction in his usages, for almost immediately afterwards he says:

The distinction between structure and communitas is not simply the familiar one between ‘secular’ and ‘sacred’, or that, for example, between politics and religion. Certain fixed offices in tribal societies have many sacred attributes; indeed, every social position has some sacred characteristics. But this ‘sacred’ component is acquired by the incumbents of positions during the rites de passage, through which they changed positions. Something of the sacredness of that transient humility and modelessness goes over…

(Turner, 1969:82)

It is surely the latter point that is his main theoretical insight, that the social space created by liminality and communitas is also the source of the sacredness that becomes invested in the hierarchical or structurally fixed statuses and roles. Turner is not concerned here with the religion-secular dichotomy. He is concerned presumably with core symbols and deeply held convictions. He is concerned with one significant aspect of the way the social order and the cosmology in which it is located are reproduced through ritual.

Edwards’ application of Turner’s ideas of liminality and communitas seems apt. At the Japanese weddings that I have attended, including my own, the social order was symbolically represented at the party that followed the Shinto ritual by the seating arrangements, in the order of precedence in speech making, and in the quite complex structure of gift exchange, as Edwards so effectively described. The mood in both the Shinto part and in the first part of the reception afterwards was solemn, deferential and serious, and it would have been meaningless to claim that while the solemnity at the Shinto part was a religious solemnity, the solemnity during the speech-making was merely a secular solemnity. Indeed, the tone of speech-making reminded me strongly of the sermons my grandfather used to preach in church – solemn, boring, but something to be endured. The endurance, a taste of ascetic self-denial (shugyō), or a rite of purification, that made the drink in the pub after the service followed by the Sunday lunch feel like a second holy communion.

At the Kyushu weddings I attended, as sake drinking progressed, the order of deference was demolished and replaced by communitas. In the ritual of sake drinking one competes, often hilariously, with one’s partners to pick up the atsukan, or flask of warmed sake, and hold it insistently over the other person’s cup, waiting for it to be lifted in order to receive the divine liquid, which should then be drunk. The other must then take the sake container and reciprocate. At the wedding parties I attended, the huge reception room was criss-crossed by men, and sometimes women, in search of a sake exchange ritual. The use of polite language became progressively more difficult to maintain as sake flowed, boundaries were crossed and status differentials obliterated. Even gender distinctions came under attack as women too became the target of sake exchanges, and structure temporarily collapsed into a benign drunken communitas.

I can hardly think of a better example of communitas in the midst of a rite of passage. But nor can I see any gain to our understanding in calling this ritual secular, in contrast to the shrine ritual which Smith calls religious[2]. What happens in the shrine and what happens in the earlier and later parts of the reception room both give expression to core symbols and deeply held convictions.

It may not be a coincidence that anthropologists who discuss religion are less likely to become submerged by conceptual problems that the category typically induces than are those who come from religious studies[3]. This is because anthropologists tend to have a more sophisticated grasp of the ethnocentric bias of categories, and a greater range of alternatives through which to lessen the perhaps inevitable distortion that occurs when the concepts of one culture are used to describe and explain another. Hendry (1987) is perfectly well aware of the special problems that ‘religion’ as a category causes, and points it out in chapters devoted to the subject in the Japanese and the wider context (1987: 117; 1999: 115-128). Jan van Bremen, in his introduction to a book in which religion appears in the title (van Bremen and Martinez, 1995) makes much the same point, that the religion-secular dichotomy simply doesn’t work in the analysis of Japanese institutions. It is as though many of us genuflect before the categorical imperative of religion even though we know through experience that it introduces unnecessary confusions, not clarity.

The answer is to problematise religion insofar as it has entered the Japanese vocabulary, either as ‘religion’ or as ‘shūkyō’, but not to assume without question that it is a neutral and valid analytical category, and more importantly not to talk as though ‘it’ is somehow ‘there’.

‘Religion’ and its tacit distinction from ‘the secular’ is a category that is being constantly used by academics and the media, as well as in ordinary conversation, to refer to a large number of quite disparate things in a wide variety of situations. Among those things there are some uses and referents of ‘religion’ that occur more frequently than others, for example belief in gods, or soteriological doctrines, or ritual in general, or some special kind of religious ritual, or cosmology, or the basic values and norms of a society, its ideology and so on. All of these ideas about the meaning and definition of religion can be found in the huge number of texts produced on the elusive subject; and frequently many different usages can be found in the same text.

Some of the usages may seem more metaphorical than others, as when for example money is described as a religion and economists as the high priests of capital, or when Marxism-Leninism is described as a religion and the Party as a church of the faithful. But few people who use the idea of ‘religion’ in their academic writing would think that it is metaphorical to call Islam, Hinduism, Japanese Buddhism and Shinto ‘religions’. When we do use the word in this way, there may lie a further tacit assumption, which is that each of these so-called ‘religions’ are species of the genus ‘Religion’, particular examples or even manifestations of a general category.

We can notice that this general category is an English language word derived from Latin ‘religio’ and with close equivalents in other European languages. There is a historical context and geographical location. It seems that in the late 16th and early 17th centuries the word ‘religion’ began to be used in new reified ways that reflected the ideological revolution of early modernity (Bossy, 1982). But many scholars in religion, and in other disciplines such as history, sociology and social anthropology, still apply it freely to any historical period in virtually any society. We can easily find books and articles about the religions of the native American peoples, about African religions, about the religion of the Buddha in ancient India, about Japanese religion in the Heian or Kamakura periods, about the religious currents of contemporary Japanese society, and so on.

What can be the justification for this?

‘Religion’ is sometimes assumed to be something we simply find in the world. Much thinking about ‘religion’ that goes on in comparative religion or religious studies has assumed that religions are things that one can study in any society, that we all know what we mean by religion when we see it and when we use the word. When someone points out that there are many anomalies where one simply doesn’t know where religion begins or ends, the reactions are various. One reaction is to say that this is true of all our categories and that the fact that there are fuzzy edges doesn’t mean to say that religions don’t exist. Some very influential writers have claimed that religion itself has an essence in the personal experience of the individual with God, but that manifestations of religion are bound to be clearer in some parts and more ambiguous in others[4]. Others, who deny that ‘religion’ has an essence, appeal to the idea that we all know what we mean by ‘religion’ because it forms part of our language game, simply by using the word we know its meanings, which form a family of over-lapping traits[5]. Others say that ‘religion’ is merely a taxonomic category that usefully organises data at a general level[6].

The point that I am concerned to make here is that, to many writers, religion has appeared as a given, universal aspect of the world and of human experience. This has taken the form of religion as a ‘natural’ aspect of the world or part of the essence of human nature, an assumption critiqued by Winston Davis (1992: 229). This habit of thinking has entered into the otherwise more sceptical and theoretically sophisticated texts of anthropologists, historians and other anti-essentialists. My argument is that religion as an idea is de facto defined in our thinking by its distinction from that which is not religion, the secular, and that this distinction itself is highly ideological and the product of specific European historical trends. Furthermore, the function of this distinction in our own democratic and capitalist ideologies is of profoundly greater importance than, for example, analytical concepts such as kinship, marriage, tribe, peasant, cosmology, sports, dance or even ritual. Though all our European-language ideas may be problematic to some extent when applied cross-culturally, some are more problematic than others.

I will leave an account of the historical problem of the religion-secular dichotomy to the end of this paper. There I will sketch out one possible scenario of the formation of the modern concept of religion in the wider context of modern western ideology. First I want to introduce examples of the problem in English language texts on Japan, a society into which the religion-secular dichotomy was imported around Meiji and where it has been inserted at certain levels of discourse – most notably the juridical, the constitutional, and the academic – without taking root in the fundamental categories of Japanese thought and behaviour[7].

I imagine a spectrum of texts on Japan incorporating to a greater or lesser extent applications of the category religion and its implicit or explicit distinction from the secular. At one end of the spectrum lie texts that are heavily indebted to the religion category and invest it with major significance[8]. At the other end of the spectrum lie the kind of texts already referred to, where the introduction of religion is either a reflex or else motivated by a sense of obligation but where there is no great theoretical commitment, and indeed where there may be a sophisticated awareness of the problem.

Near the committed end of the spectrum of texts I place those of Reader (1991, 1993, 1995, 1998), as when, for example, Reader asserts that:

…in reality Japanese people in general exhibit extremely high levels of religious activity and behaviour, and Japanese society and culture are intricately interwoven with religious themes.

(Reader, 1991: 5)

Such texts that exhibit high levels of commitment to religion as a category would normally be most heavily freighted with the ideological assumptions that derive from comparative religion and its founding fathers such as Max Muller, Wach, Otto, and Eliade. Conversely, at the point on the spectrum furthest away, I place texts of the kind I have already discussed such as those of Smith, van Bremen and Hendry, and also in the context of anthropology more generally Victor Turner. Between these two different ends of the spectrum I place writers such as Lawson and Macauley (1990), Winston Davis (1992), H. Neill McFarland (1967), Byron Earhart (1982), and Richard B. Pilgrim (1985), whose texts are heavily freighted by the religion category and its distinction from the secular. Such writers tend to be located in a sociological tradition, in Davis’s case the Weberian paradigm with which he engages. Lawson and Macauley are anthropological theorists, not specialists of Japan, who claim to be able to make a useful distinction between religious and non-religious rituals.

What Davis, McFarland, and Reader share with many other authors in their different ways is a commitment to the study of Japanese religion, Japanese religion and society, Japanese religiosity, or the religious world of the Japanese. On the other hand there is an important difference between Reader and Davis. Reader may be unaware of the ideological legacy he has received from the fathers of comparative religion, and one sign of this possibility is that he seems to suppose without question that religion is a universal phenomenon that has existed in all times and all places. There is implicitly conveyed in his writing a notion that religion is a natural reality, and that to deny that the Japanese are religious would be absurd and contrary to common sense. Davis, as I will show, also never fundamentally questions the religion category, only its method of application; but he is well aware of the ideology of religious studies, makes explicit reference to it, and distances himself from it. Davis approaches the subject as a historian and sociologist, indebted to (but not uncritical of) the Weberian paradigm.

So the kind of commitment has a different nuance between Reader and Davis. But they are both significantly committed to religion as an analytical category, and their books (like that of McFarland) labour the pursuit of the religious to such a degree that it often seems as though they are asserting the existence of a substantive, reified entity. What I hope to show is that this commitment to what is essentially a category mistake involves them and us in unnecessary conceptual entanglements. Their commitment to religion as a category of analysis does not seem to emerge from any clear theoretical gains. Indeed, quite the opposite. It frequently leads to unintelligibility. The problem then is to understand their persistence with it. I suggest that this can only be understood as the result of a western ideological imperative that we all to some extent share and that needs to be uncovered effectively before it can be neutralised and eliminated from our analytical vocabulary[9].

The problem lies in the way data is selected and presented when the religion-secular divide is operating, and the way the tacit theoretical presuppositions introduced by the category religion itself distort rather than clarify our understanding. This ideological field, or religion discourse, is also reproduced, as I have indicated, though in somewhat less invasive ways, in anthropology texts, and is reflected in taught courses on the anthropology of religion, anthropological sections and panels of religious studies conferences, and chapters on religion in introductions to anthropology.

While all our categories can be problematic, religion and its distinction from the secular carries more ideological weight in the modern western configuration of values than most. Modern concepts of politics, economics and law, for example, presuppose the notion of a secular sphere, and the idea of the secular is interdependent with the idea of religion. They are mutually self-defining. It is for this reason that I am arguing that we should take special note how this important piece of western ideology misleadingly appears as a piece of neutral analytical equipment, or even more problematically as though ‘it’ is somehow objectively there in the society or culture. Indeed, in the texts that I shall look at, it is simultaneously both the analytical equipment and the object of analysis.

Despite the problems with the English word ritual, effectively analysed by Bell (1992), it would be difficult to proceed with an analysis of social institutions, especially Japanese ones, without it. It seems important that one of the best recent books on ritual and ceremony in Japan explicitly problematises the religion-secular dichotomy and offers instead a powerful spectrum of ethnographies on rituals (van Bremen and Martinez, 1995: 2-3). The kinds of theories of ritual that are spelt out is important. Rituals, where they exist in the Protestant-influenced worldview, tend to be imagined as symbolic representations of meanings already formulated in doctrine. Protestant rituals, and the idea of ritual held by Protestant-influenced theorists, do not actually perform any mystical transformations, as in the Catholic mass; they merely act out symbolically the primary intellectual beliefs[10]. The ritual performances are explained by the beliefs.

This Protestant view of ritual as a secondary symbolic statement of already held, doctrinally formulated beliefs about salvation has been influential, especially in religious studies. However, anthropology has also provided a different theoretical tendency. Ever since Robertson-Smith's publication of The Religion of the Semites it has become a commonplace observation that rituals are not usually if ever ‘explained’ by the prior holding of beliefs, but are performed anyhow, the doctrinally formulated beliefs (if they exist in any coherent sense) being used retrospectively to explain and/or justify the actions. The idea that ritual practice is logically and perhaps chronologically prior to belief systems and can function quite happily without intellectualist representations at all is at least a serious proposition that might help us to understand what is happening in a society such as Japan[11].

Lawson and Macauley, like some other anthropologists and religionists, postulate a distinction between religious and non-religious rituals on the basis of the criterion of “culturally postulated superhuman agents” (1990). However, such a distinction merely puts the problem one step back but does not solve it[12]. By favouring rituals that involve ‘superhuman agents’ with some special attribute that is known as ‘religiosity’, we hold ourselves ransom to what, in the Japanese context at least, is a weakly conceptualised category. Indeed, it may be that the Japanese capacity to take the shrine and temple performance of rituals seriously in the absence of clearly articulated ideas or doctrines about aborted foetuses, printing blocks, dead ancestors, bodhisattvas, angry ghosts and children’s toy dolls can only in principle be understood in the context of the whole spectrum of ritual performances that are conducted in every institution and at every level of society.

By giving special priority to some supposedly special class of actions and experiences called ‘religious’ that can be decisively distinguished from ‘non-religious’ ones, we will be in danger of losing sight of what may be more important analytical connections, ones that might be established on the basis of metaphor and analogy across the spectrum of ritual performances.

True, ritual is also an English-language word and therefore carries semantic associations developed from within the Christian west. Jack Goody has claimed that it is a useless concept and Grimes admits that it is a vague idea in English, and that often there is no corresponding word in other languages (Grimes 2000: 261). However, whereas religion as a concept is fundamental (I shall argue) in the construction of modern western democratic and capitalist ideology, the idea of ritual seems less charged with ideological resonance and easier to separate out as a useful, indeed indispensable, analytical category. As Jan van Bremen has put it:

The study of ritual is not a search for the essential qualities of a peculiar and qualitatively different event; it is a way of examining how trivial elements of the social world can be elevated and transformed into symbols, categories, mechanisms, which, in certain contexts, allow the generation of a special or extraordinary event.

(van Bremen, 1995:3)

Rituals, or actions with ritual characteristics, are performed everyday and with scant regard for the niceties of the so-called religion-secular distinction. In this sense the concept seems far less loaded with ideological baggage than religion, and, provided it is not essentialised as some quite distinctive kind of thing, but is used adjectivally or adverbially to indicate attributes of performances that, being not merely idiosyncratic and not only utilitarian or technical, have some kind of publicly recognised symbolic significance, seems to have been remarkably successful in the anthropological context as a cross-cultural analytical category[13].

All sorts of institutions, even in western societies, straddle this problematic division between religious and non-religious rituals, a distinction that I want to argue is superfluous and possibly unintelligible. Ritual, ceremony and the celebration of sacred symbols and statuses are performed across a wide spectrum of social life and undercut the ideologically determined distinction between religion and the secular[14].

Ritual is pervasive at every level of society, not least in Japan. The attempt to distinguish between religious and non-religious ritual does not solve this problem but merely reproduces it. Yet the reflex power of our categories, and our easy confusion of categories with reality itself, seems to compel scholars to assume that their seeming naturalness and intuitiveness must be a sign of their veridicality.

This wider ideological field is well indicated in its commercial form by the contents description on the back of Reader’s well-known Religion in Contemporary Japan (1991), which typifies the marketing of this field:

What role does religion play in contemporary Japanese society and in the lives of Japanese people today? This book examines the major areas in which the Japanese participate in religious events, the role of religion in the social system and the underlying views within the Japanese religious world. Through a series of case studies of religion in action – at crowded temples and festivals, in austere Zen meditation halls, at home and work, at dramatic fire rituals – it illustrates the immense variety, energy and colour inherent in Japanese religion while discussing the continuing relevance and responses of religion in a rapidly modernising and changing society.

These unexamined clichés are formulaic repetitions, which through the power of sheer assertion, leaves no room for doubt that religion is some kind of object and even agent in the world[15]. We already know what religion is – temples and festivals, Zen meditation halls, fire rituals, the home Buddha shelf [butsudan - 仏壇]. Religion is inherently various, energetic and colourful, and responds to modernisation. It is also an agent that plays a role in society, as though religion were something separate from society but interacting with society at specific points.

Ian Reader’s contribution to Japanese Studies is in several ways an outstanding one, and his books and articles have made an important contribution to our knowledge. I do not question the validity of his research, nor its inherent interest to an international audience[16]. My concern is with the inherent theoretical problems with the concept of religion, problems that Reader did not invent and which we are all in my view entangled in to one degree or another. Reader says that “..intensely religious dimensions still operate in Japanese society” and that “there does exist a deep and continuing stream of religious motifs interwoven with (rather than separate from) other aspects of Japanese life and society” (Reader, 1991: 15).

My contention is that this author’s multiple references to religion, or the religious world of the Japanese, and his repeated distinction between overtly religious acts and those rituals that are merely secular, can only be understood as an attempt to assert a special, distinct and irreducible realm of experience, feeling and action. But though Reader himself has pointed out that the idea of religion was imported during the Meiji era (1991: 13-14), he never considers the arguments that this is a western myth, one that liberal ecumenical missionaries and others have been exporting to the rest of the world since the days of Max Muller. Furthermore, it is unjustified by the actual data that is cited by the author, which often forces him to virtually deny what he is asserting within the same paragraph.

We can only explain Reader’s determination to construct a “religious world” of the Japanese as the result of some prior commitment of the author’s. However, Reader did not invent this commitment himself as a lone individual. This commitment springs from the collective consciousness of western societies.

The expression “interwoven with (rather than separate from)” seems to be some kind of metaphor, but actually it is a rhetorical trick playing on the words “interwoven ... rather than separate from”. The implied analogy is with weaving but the analogy is misleading. If a red stranded thread is interwoven with blue thread to form a tapestry, do we say the red is separate or not separate from the blue? Obviously it depends on the context and the purpose. Yes, they are separate threads of different colours, but they are also combined in one overall pattern. But we can in principle identify and separate the red thread by looking for it. But what are we looking for in the case of “religious motifs”? We don’t know – this is precisely our problem. We do not have the intuitive certainty that we have in our distinction between red and blue, because what constitutes religious as distinct from non-religious motifs is the contested point at issue. The metaphor is a problem because it depends on what is included in “other aspects of Japanese life and culture”. These “other aspects of Japanese life and culture” in turn depend on what is included in religion. Reader is trying to hoodwink himself and us by pretending that what constitutes these categories is unproblematic, we all know what they mean. The language of “interwoven ... rather than separate from” seems to be giving us an insight into Japanese reality, but I would question whether it actually does so.

Reader on ‘religion’ as shūkyō

Reader more or less admits that the very idea of ‘religion’ and its distinction from ‘the secular’ has been negotiated into existence in the Japanese context. He says

A problem that occurs…is precisely what is understood when terms like ‘religion’ are used in Japan. The Japanese word generally used in surveys and elsewhere to denote ‘religion’ is shūkyō, a word made up of two ideograms, shū, meaning sect or denomination, and kyō, teaching and doctrine. It is a derived word that came into prominence in the 19th century as a result of Japanese encounters with the west and particularly with Christian missionaries, to denote a concept and view of religion commonplace in the realms of 19th century Christian theology but at that time not found in Japan, of religion as a specific, belief-framed entity. The term shūkyō thus, in origin at least, implies a separation of that which is religious from other aspects of society and culture, and contains implications of belief and commitment to one order or movement – something that has not been traditionally a common factor in Japanese religious behaviour and something that tends to exclude many of the phenomena involved in the Japanese religious process.

(Reader, 1991: 13-14)

What is remarkable about this passage is that, at the very moment that Reader is showing us, correctly, that the term religion and its derivatives such as religious is an alien concept which falsely separates “that which is religious from other aspects of society”, he is at the same time, even in the same sentence, continuing to assert such ideas as “Japanese religious behaviour” and “the Japanese religious process”. He is saying, in effect, that the distinction between the religious and the non-religious is not a common factor “in Japanese religious behaviour”. These are not mere banal grammatical observations that I am making; the use of language here is an indication of a wider problem, and one feels compelled to ask why a writer of Reader’s acknowledged accomplishments should get trapped in it. The answer is that such thinking is not merely Reader’s, but is institutionalised within Religious Studies and to some extent throughout the humanities and social sciences. This is how an ideology mystifies us by appearing as self-evident and in-the-nature-of-things.

Reader acknowledges the pervasive importance of the household, the nation state and the collective racial identity throughout Japanese cultural practice (1991: 27-8). Yet the political in this sense is marginal to Reader’s purposes, for ‘religion’ is constructed as something that has little or nothing to do with power. Power after all is in the secular field (McCutcheon, 1997).

The problem is with the selection and interpretation of data. The data itself often tends to cluster around ritual forms of behaviour that are ostensibly linked to the ‘supernatural’ or its nearest Japanese equivalent. Thus, for example, the contents of the book are typically concerned with Buddhist temples, Shinto shrines, priests, ancestors, graveyards, butsudan, mountain asceticism, and new religions such as Agonshū.

The fact that some of these ritual actions make references to various unseen beings that in English we typically refer to as the supernatural fails to guarantee a separate and distinctive phenomenon called religion, as Reader’s own data indicates. For this reason he asserts “The Primacy of Action” (1991:15-20) by which he seems to suggest that it is ritual practice that defines the religious sphere. If Reader had gone on from there to consistently develop a theory of ritual, he would have been in good company, and I suggest many of the conceptual confusions in his work would be resolved, doing greater justice to the fascinating and important material that his work contains.

Reader distinguishes between religious rituals and non-religious rituals, just as he distinguishes between religious harmony and non-religious harmony (1991:30), and even has a special set-apart place for “religious sincerity” (1998). This means that there must be some further criterion available. But having trawled through many of Reader’s texts I cannot find it, and am convinced that it is a chimera. There is therefore no clear editorial principle for either including or excluding a vast array of ritual activity connected to etiquette, hierarchy, power, politics and/or the reproduction of the social order.

The Japanese are deeply religious, but they don’t know it

So on the one hand for Reader, “religion as a specific, belief- framed entity” (1991: 15) is not typical of Japan, and he has always in his publications rightly attempted to focus on the ritual performances, or what he calls “The Primacy of Action” (1991), a phrase that might have been taken straight from Robertson Smith. The practice part of the idea of being “practically religious” points in the same direction (1998). Talking about the great New Year shrine visiting activities of hatsumōde, and the national ritual of visiting the graves of ancestors at o-bon in July or August, he says:

Much of what is done at New Year and o-bon does not require any prior or fixed religious commitment from the participants: it does not even seem to require belief in the existence or influence of the spiritual beings (kami or ancestors) to whom the prayers and offerings are directed.

(Reader, 1991: 11-12)

Again and again in this book, Reader seems to be about to articulate a theoretical interpretation of the significance of the ritualisation of life in Japan that would explain why we (both the Japanese and the expert observers) do not need to talk about ‘religion’ at all. For most of what Reader means by religion is the performance of various kinds of rituals, and it is unclear what else over and above ritual he really wants to talk about.

On the other hand, Japanese people are overtly religious but they don’t know it:

Japanese people in general are quick to say they are not religious and to describe their society as one where religion either does not exist or has in some way died out. Winston Davis writes of a conversation he had with a Japanese banker on a flight to Japan in 1971:

‘When he learned that I was studying Japanese religion he shook his head and sadly assured me that there no longer was such a thing. I was later to hear the same thing from many other Japanese.’

In the 1980s I too have received similar assurances. It is almost as if many Japanese like to convey the impression that Japan is a wholly secular society in which religion has to all intents and purposes disappeared. As this book will demonstrate however, this is certainly not the case for, in reality, Japanese people in general exhibit extremely high levels of religious activity and behaviour, and Japanese society and culture are intricately interwoven with religious themes.

(Reader, 1991: 5)

Here we find the assumptions of the scholarly expert confirmed by the reference to Winston Davis in the 1970s, which for Reader strengthens his case that the Japanese are fundamentally mistaken about their own behaviour, and that they have been harbouring this mistake for a long time. It confirms the assumption that Reader and Davis know exactly what religion, religious activity, and religious behaviour is, but the Japanese do not. Indeed the Japanese have persisted in their error throughout the 1970s and 80s. And since, as we have already seen, the Japanese never did distinguish between religious and secular ideas and institutions until western pressure led to the construction of shūkyō, then presumably they had been in error until the Christian missionaries and their liberal ecumenical progeny came along to enlighten them[17].

Reader tells us that the Japanese are religious but they don’t know it. He has hinted at a possible reason, that they want to appear secular. But what does secular mean? For it needs to be remembered that the concept of religion did not come into Japan alone, pure as a shrine maiden. It came coupled with the concept of the secular. The insistence by the western powers that a civilized society separates church and state was a kind of imperial intrusion[18]. The confusion persists to this day. Thus, when a Japanese person seems to want to say that Japan is a secular society, it is far from clear what he or she means.

Reader gives instances of Japanese people denying that they are religious, or denying that there are some emotions or experiences that they, as Japanese people, separate out as of a special kind, attributes which would mark them out as people with a special religious commitment, rather than, for example, a commitment to those values, ideals, dispositions, and principles which are widely celebrated in institutional settings across the whole spectrum of Japanese cultural life.

Yet Reader continues to insist that they do act with some special distinctively religious feeling or commitment. In one story his two Japanese friends tell him that they are not religious and they have no interest in religion, but Reader knows what religion is, and he knows they are wrong:

Not long after I first arrived in Japan in 1981 two Japanese friends, knowing my interest in religion, decided to take me out for the day to visit some Buddhist temples and Shinto shrines. Before setting out they assured me that they were neither religious nor did they have any interest in religion: their participation in the trip was purely to show me some places that would be of interest to me in my studies. I have long since become inured to Japanese people telling me that they are not religious, even whilst performing acts of an overtly religious nature such as praying at a shrine or walking a pilgrimage route dressed in the traditional clothing of a Japanese pilgrim, so that nowadays I hardly take any notice of such protestations. At the time however I did, assuming that this meant my friends would act similarly to me at such places – interested, respectful but not worshipful.

(Reader: 1991: 1)

The main points that Reader is telling us here is that he, Reader, knows without question what kind of thing ‘religion’ is, or what kind of things the word refers to: praying at a shrine, walking a pilgrimage route, wearing traditional pilgrim's clothes, and performing the rituals in a “worshipful” – rather than a merely interested and respectful – manner are all “overtly religious acts.” But his Japanese friends can’t see it.

There is thus a contradictory tension in Reader’s book, on the one hand acknowledging that the evidence points to the irrelevance of the idea of a special, set-apart “religious world” in Japan, on the other hand projecting onto the data and onto his Japanese informants this set of attributes.

Reader and Tanabe (1998)

In the 1998 book Reader, with co-author Tanabe, argued that there is a “common religion” of Japan that is in principle shared by all Japanese and can be located and even advocated in both popular practice and the ancient texts (see Reader and Tanabe, 1998: 23-32). It is a religion that undercuts distinctions that are made between lite and popular religion, or between scriptural religion of educated priests and members of the upper classes, on the one hand, and the superstitious religion and magic of ordinary people. It is a religion that has at its heart “the practice of seeking this-worldly benefits”, a religion that “provides an open-access, total-care system for its members.” (1998: 31)

As sociologists Reader and Tanabe rightly question the theological (or Buddhalogical) value judgement that lies behind a distinction between true and false religion. But they perpetuate the more fundamental myth of religion itself. When Reader and Tanabe talk about “the common religion” of Japan and the Japanese, they carve out from the spectrum of ritualised actions a special category, implicitly distinct from the non-religious or secular ones. In comparison to this category mistake, the issue of ‘true’ versus ’false’ religion seems secondary.

Reader and Tanabe refer to a “religious context” as relying on “spirit-oriented explanation and interpretation of events” which is in contrast to “the scientific-rationalist worldview that underlies the modern education and political systems of Japan as a modern secular state.” (1998: 257)

However, the claim that genze riyaku (this worldly benefits) demarcates a religious rather than a non-religious sphere of activity should be challenged. Genze riyaku is certainly about exchange and reciprocity. The principles that govern relations between the living and the dead do not, on the basis of the data provided, seem fundamentally different from the principles governing relations between the living. The ancestors are related hierarchically to the household head much as parent to child, the older to the younger, the emperor to the subject, the sensei (teacher/master) to the deshi (student/disciple). The principles of obligation, reciprocity, and hierarchy operate similarly in all these relations. Recent ancestors are not necessarily remote from the living descendents. An ancestor may be more intimate and approachable than a remote company boss, and therefore apparently require less formality of approach, less awe and humility. This does not mean that the boss is more transcendental than the ancestor in an ontological sense. It’s all a matter of propriety, for both are fundamentally social beings, and the relations between them are fundamentally social relations. In both cases there is mutual dependency, obligation, and exchange, and both the boss and the ancestor provide this worldly benefits if the rituals are correctly performed.

Nor does this analogy mean that there are no differences between the boss and an ancestor. It might mean that they are petitioned in a somewhat different register. The sake that is offered to the boss or the iemoto sensei (teacher/master in a traditional arts or craft lineage) at o-chūgen (a midyear gift) or o-seibo (a year-end gift) may need to be specially gift wrapped in a way not necessary for an ancestor. Or we might say they are gift wrapped in different ways. The difference of register, and the significance of that difference, should certainly be part of what we want to research. It might seem ironic to bracket relations with the boss as ‘religion’ because there is greater deference involved, and those with the ancestor as merely an extension of ‘secular’ kinship relations beyond the grave.

Reader’s quotation from Winston Davis concerning the story of what the man said in the airplane can be found in Davis’s Japanese Religion and Society: Paradigms of Structure and Change (1992: 231-2) in a chapter called “The Secularization of Japanese Religion”. Like Reader, Davis has an enviable and deep knowledge of Japan, and it is difficult, when one is chasing a theoretical and methodological problem, to do justice to the many insights, pieces of fascinating knowledge, and important contributions to one’s own understanding of Japan that the book contains.

Davis’s theoretically thoughtful multidisciplinary approach is indicated by his struggles with Weber, and the debt he acknowledges to sociology, anthropology and history. Historical understanding is also of fundamental importance in his attempt to set contemporary issues in a longer-term perspective. This welcome orientation leads us to suppose that Davis intends to use religion as a heuristic device, an analytical concept that allows him to focus with reasonable precision on some specific aspect of Japanese society that can be, and needs to be, distinguished from other non-religious aspects.

Yet things are not as clear as they might be:

While this book is about Japanese religion, my goal is to frame discussion in such a way that when readers put the book down they will have a deeper knowledge of Japanese society and culture in general, and possibly even deeper insight into the nature of religion itself. The book deals with the relationship between Japanese religion, culture, and values on the one hand, and society, social change and economic development on the other. I have written some chapters as a sociologist of religion, others as a historian. Most of the volume, however, falls under the mixed rubric of historical sociology....

(Davis, 1992: 1-2)

There is here a profusion of relationships that the author intends to establish, between Japanese religion, culture, values, society, social change, economic development. Religion is posited as having its own nature here. Yet later he explicitly repudiates the claims by such Founding Fathers of contemporary religious studies as Van der Leeuw, Rudolf Otto and Mircea Eliade that religion is part of the essence of humanity, and explicitly rejects their claim that religion “has existed de facto in all cultures up to the present…” (1992: 229). This ambiguity can be found throughout the book.

The title itself should ring warning bells for the reader. The idea of coupling the terms “religion” and “society” should immediately lead us to ask what kind of relationship this is conceived to be, for inevitably by separating them in this way there is a tendency towards reification of two in principle equal entities joined in conflictual marriage. It is interesting that in his review of Davis’s book shown on the back cover, Richard B. Pilgrim should write

It provides in one place and under one cover useful studies featuring the important interrelationship between religion and society, or – to put it more correctly – Japanese religio-social reality.

This correction, which appears to spring from a minute attention to detail and nuance, instead indicates a semantic muddle pushed into a new key, for by de-coupling “religion and society” and then putting them back together again in the form of “religio-social reality”, an illusion of theoretical gain in our thinking has been established. Pilgrim seems to want to suggest that religion and society cannot be separated into distinct reified entities, yet they can nevertheless be distinguished, perhaps as different aspects of the same entity. But this still leaves us with pretty much the same problem, which is to know how to distinguish between the religious and the non-religious aspects of the same entity.

Davis and the reification of Religion

Davis has stated clearly that his work is one of historical sociology, and he has distanced himself explicitly from the sui generis school of religious studies. Yet it is quite clear when one reads Davis’s book that, despite his theoretical intentions, much of the time he is thinking of religion as a thing, an agent that acts and reacts and comes into relationship with other reified entities.

This tendency to hypostatization of religion may not be the intention, and we know that, in his discussion of how to measure secularization for example, Davis attempts to break down the rather monolithic and Protestant Christian influenced idea of religion as a system of beliefs, separating religion into smaller, more effectively measured components. He does this by separating out issues of belief from issues of action and feeling. He says “…religious praxis (shugyō) and feelings (kimochi) form the core of Japanese religion.” (1992: 236).

Yet even here the idea that religion has a “core” tends to reify, and it is difficult for the reader to know how, in Davis’s view, a core religious feeling differs from a non-religious feeling, or a religious praxis differs from a non-religious praxis. And Japanese religion is being identified in terms of religious praxis and religious feeling, which is tautologous. To say that Japanese religion is characterised by religious feeling does not help us to understand anything very much.

If religion and society are being conceptually separated, and then coupled together as though they are two distinct things that have some kind of contingent relationship with each other, we are bound to look within the book for a clear explanation as to what kind of ‘thing’ has been separated out from society in this way. Being told on page one that religion has its own nature tends to strengthen the sense that Davis is ambivalent. He is, rightly in my view, concerned to point out that emic categories are not always necessarily the best way to understand a society, and that we need both emic and etic categories (1992: 6.) Yet the problem of ‘religion’ as an etic category is not satisfactorily discussed, and throughout the book one can only gather what meanings and referents religion has by the examples he gives, and by the hints that he drops sideways on, as it were.

Davis is critical in a certain kind of context of generalisations about religion and the process of secularization, as when he says

I shall try to demonstrate by statistical and historical examples that there is no universal measure for ‘the secularization process’ – whatever that means! – and that even within a single religious system ‘the general decline of religion’ is a meaningless jumble of words.

(Davis, 1992: 231)

Yet if “the general decline of religion” is a meaningless jumble of words, how are we to understand the meaning of “a single religious system”? How is a religious system distinguished from a non-religious one? We already have uses of the word that suggest that religion is something in itself, some reified thing that acts or is acted upon. For example, in the sentence previous to the one just quoted, where he dismisses the proposition concerning “the general decline of religion” as a meaningless jumble of words, he had said “Rather than assuming that the decline of religion, however defined, is inevitable or impossible, I shall assume that it is possible, not necessarily unilinear, and therefore reversible.” (1992: 231). The almost throw-away aside – “however defined” – suggests someone who is weary of definitional problems, and wants to get on with the more important job of showing us what religion is really like by analysing ‘it’ in real contexts.

Throughout chapter 7, where on the one hand he wants to break down religion into components for more precise analysis, we find expressions that reify religion as though it were a passive or active agent, or as if it were some distinct property of other hypostatised agents: “religion as the unhappy victim of…”; “the decline of religion…”; “a quasi-religious integration”; “religious change”; “religious history”; religion as “a symbolic means for effecting social solidarity”; “cross-cultural religiousness”; “each religious system”; “the central place of feeling and emotion in Japanese religion”; “this very religious event”; “magic, like religion itself, is highly situational and quickly learns to adjust to new conditions” (Davis, 1998: 229-251.)

In trying to isolate, from the actual contexts in which he uses the word, what Davis means by religion I find a tautology running throughout the book. Religion and its distinction (implicit or explicit) from the secular, or from the process of secularisation, is both the object of analysis and the major analytical category. For example

…I shall assume that the meaning of secularization includes complementary notions of desacralization, differentiation, and the transposition of religious beliefs and behaviour to the ‘secular’ sphere.

(Davis, 1992: 231) )

Here the “meaning of secularization” depends on the meaning of “the secular sphere”, because this is where “religious beliefs and behaviour” are being transposed. This is itself a tautology, secularization being characterised in terms of “secular sphere.” Meanwhile “religious beliefs and behaviour” is both defining “secular” and being defined by it (e.g. by negation, for what is secular is what religion is not, and vice versa).

Now I have been unable to find a direct explanation as to how a religious belief differs from a non-religious belief. Telling us that feeling and action are at the core of religion only pushes the problem one step further back, because we need to know what kind of action or feeling can be considered as religious, and what kind is secular. We can, if we look through the book systematically, find a large range of examples of religion or religious characteristics; or we can infer from his discussion of secularization since the 19th century the kinds of things (the religious things) that get reduced or undermined by secularisation.

Many of the following ideas about how to assess the degree of secularization are borrowed from the folklorist Yanagita Kunio who was interested in the changes that took place in “the festival faith of the common folk” as a result of the Meiji government’s modernization programme, and its desire for “the secularization of folk practices” (1992: 237). Here the decline of religion (and thus secularization) is equated with “the secularization of folk practices”. For example (taken again from chapter 7), the shrinking of sacred space and time; reduction of time spent on festivals; the prohibition of local customs such as transvestism and orgies; the undermining of religious chants because the Tokyo dialect was declared the standard language; abolition of traditional religious holidays; the introduction of a commercialised version of Christmas; an increase in the daily eating of rice and fish which had previously been reserved for special days; the prohibition of private sake brewing; drinking alone instead of from a common jug; eating from an individual tray; pilgrims putting up at “secular ryōkan” instead of “temple inns”; disruption of local festivals due to increasing mobility; change in style of festivals from the very local and village-centred to today’s big public festivals; making money offerings in saisenbako (offertory box) as a result of people not knowing the locally correct way to approach the kami (usually translated as gods, but a potentially misleading translation); decline in belief in magical spells.

It may seem invidious to separate these examples from the context of the discussion, but some important features of the problem I wish to identify do become apparent here. There is the vagueness and elusiveness of the topic, when neither of the key terms is being brought into clear relief. For example, we can guess that a temple inn is a ‘religious’ place to stay, in contrast to a “secular ryōkan”, because it has long been one of the fundamental assumptions of the religion industry that Buddhist temples are religious places, though we don’t know clearly why. But nor is it explained why a ryōkan is secular. We are told that religion is undermined or reduced by the reduction of religious holidays. But here again we have the tautology that religion is being defined by itself. The decline of beliefs in magical spells is a sign of secularisation and thus also of the decline of religion, yet we don’t know how to distinguish between (or whether we are intended to distinguish between) magic and religion. They have different names, but the decline of one is a sign of the decline of the other.

I believe this is the key to unravelling both the deepening complexities, and yet lack of clear conceptualisation, in Davis’s text. Throughout, religion is being used both as the topic of investigation and the major analytical concept. It is thus defining itself. And it is also defining, and being defined by, the ‘secular’. We therefore have a merry-go-round in which the meaning and referent of religion is defined by non-religion (the secular), and the secular is defined by religion. Therefore the procedure is entirely arbitrary, and the large theoretical discussions that are placed upon these shifting sands are insecure. No wonder Davis dislikes definitions, and would rather just get on with the job and use the tools at hand. But the tools at hand are themselves what need to be investigated.

Davis and the concept of exchange

But Davis pursues some theoretical ideas that have considerable interest and might help us to be freed from the conceptual nightmare in which the western ideological distinction between religion and the secular, when promoted as cross-cultural analytical concepts, seem inevitably to entangle us. One of these is his exploration in chapter 2 of the concept of exchange, which earlier I suggested might be usefully connected to worldly benefits genze riyaku in Reader and Tanabe’s text. Davis adopts Schultz’s distinction between in order to motives and because motives for exchange. The latter kind of motives involves obligation, the return of benefits and favours including those already received (1992: 18). The sense of obligation is surely fundamental and pervasive to Japanese relationships. Davis is concerned that a western idea of exchange framed in terms merely of individual economic self-interest and divorced from pervasive norms and values connected to that sense of obligation would distort Japanese reality. Talking about the values of devotion, loyalty, gratitude, and selflessness, he says

Because of the importance of such values in Japanese religion and society, we must carefully hold in check the cultural bias built into exchange theory, a theoretical persuasion which generally reflects the western market economy and its psychology of possessive individualism.

(Davis, 1992: 17)

As Chalmers Johnson has argued (1995), a large part of exchange in Japan is embedded in social relations in a way that eludes the assumptions of some western economists. The problem with Davis’s analysis is that, while on the one hand rightly suspicious of a concept of exchange that reflects the ideological assumptions of western individualism and economic theory, he frames the whole of his own discussion in the context of an equally problematic western category religion and its distinction from the secular.

Like ‘economics’ and ‘politics’, ‘religion’ has been disembedded historically into a distinct sphere of western ideology in a way that, arguably, is not characteristic of Japan. But Davis is determined that it is specifically “religious affiliations” (1992:16) to which he wants to apply exchange theory. Ideas about exchange and reciprocity, embedded as they are in the great spectrum of hierarchical relationships in homes, schools, corporations and factories, iemoto (the originating teacher-disciple lineage of arts and craft schools, for example 芸道の流派), universities, political habatsu (factions), sumo beya (sumo wrestlers’ ‘stables’), baseball teams, bureaucracies, shrines and temples, might more appropriately serve to undercut the western distinction that we seem so eager to impose on Japan between ‘religion’ and ‘the secular’. We have to decide why we suppose that the distinction between religion and the secular is so imperative for the analysis and interpretation of Japanese society, and why it should be allowed to override other categories such as exchange, reciprocity, indebtedness and hierarchy that might prove to be more illuminating. We need our analytical categories to maximally strike and illuminate the ‘connected-up’ button, not to over-determine the field clumsily in terms of our own western ideologically charged categories. By thinking of relations with the ancestors in terms of exchange and reciprocity, we can notice how much these relationships share with hierarchical and/or ie-type (in the sense that the structure and organisational principles of Japanese corporations can be seen as analogous to the traditional household, see Hirochika Nakamaki in Bremen and Martinez 1995: 146-160) relations in companies or iemoto organisations, and that the attempt to conceptually separate relations of obligation between, say, the ancestor and the descendent of the ie, and the sensei and his deshi, in terms of one being ‘religious’ and the other being ‘secular’, is to over-determine the field and maximise rather than minimise the inevitable distortion inherent in any act of translation.

Davis’s whole discussion of exchange is self-consciously framed in relation to “religious affiliations” in Japan. At this point, the guiding assumptions about what constitutes the religious in Japan are stated explicitly. Generally speaking “the Shinto parish and its guilds”, “the Buddhist family temple”, the “civil religion” of the pre-war period are the examples given of specifically religious institutions based on obligatory (because) conduct (1998: 19). In contrast, Buddhist prayer temples, Sudden Death Temples, and New Religions rest on “motivated” (in order to) conduct. Other religious affiliations include the confraternity (kō) and “various less formal, ad hoc relations with faith healers, shamans and mediums.” (1998: 16).

In an interesting way, but not one that will necessarily clarify Davis’s argument, this distinction between affiliation that springs from motivation (in order to) and obligation (because) might be considered analogous to the distinctions made in the Meiji Constitution and the Imperial Rescript on Education between ‘religion’ (Buddhist and Shinto sects as purely voluntary private organisations to which individuals choose to adhere) and ‘the secular’ (State Shinto and the observance of Confucian ritual hierarchy as obligatory on all citizens.) In this Meiji formulation, both ‘religion’ and ‘the secular’ are contested and problematic categories, for the Japanese did not find it self-evident what fitted into which category[19]. So it is interesting that the Japanese rulers themselves distinguished the imported religion-secular dichotomy in terms of what is voluntary and what is involuntary or obligatory.

What the Japanese lite themselves considered as obligatory was indeed based on collective, everyday values of hierarchical indebtedness and reciprocity, but not in terms of either-or dichotomies like the other-world/this-world, or between gods/humans, or between nature/supernature (all of which are arguably alien to Japan). Thus obligatory relations between the Emperor and his subjects belong in the same general category as obligatory relations between fathers and sons, or between ancestors and heads of the household, or between sensei and deshi, or more generally between more senior and less senior. The fundamental values seem to have been framed in terms of indebtedness, obligation and hierarchy. Instead of applying the distinction between voluntary and obligatory conduct to religious affiliations, it seems more advantageous to apply them to affiliations per se.

It might be claimed that the concept of ‘religion’ is necessary for understanding doctrinal sectarian movements such as those usually referred to as the ‘new religions’. Certainly this looks like the area which closely corresponds to the modern concept and where one can identify legally defined institutions. Yet even here, it may be that the concept of religion is doing more to obscure and confuse the data than to clarify it. This becomes evident in a well-known study Rush Hour of the Gods: A Study of New Religious Movements in Japan by H.Neill McFarland (1967). Early on McFarland points out a problem of analytical categories. Several of the groups claim to reproduce Shinto or Buddhist ideals, so are they ‘new religions’, he asks, or sects or sub-sects of Buddhism or Shinto? His solution is to call them “contemporary popular religious movements” (1967: 8). He says that ‘contemporary’ conveys that they are new and that they are rooted in distinctively Japanese traditions; ‘popular’, that though there are new and foreign elements, the movements themselves belong to a tradition of ‘folk faith’ (minkan shinkō). ‘Religious’? McFarland admits one could argue that ‘religion’ is a misnomer, but he doesn’t explain why and simply asserts that ‘religious movements’ is an acceptable expression. He says “...collectively the New Religions constitute a definable socio-religious movement – one of considerable size and potential significance” (1967: 9-10).

McFarland does not tell us how a ‘socio-religious’ movement is different from a ‘social’ movement, and this is in fact typical of scholars working with the notion of religion. But it is important to my argument: what extra meaning is being added to social when ‘religious’ is added?

McFarland places these movements in historical context, because he wants to show that they are or were originally popular expressions of dissatisfaction caused by long-standing changes in Japanese society. He variously dates the origin of these changes three and a half centuries ago (1967: 11–54) 200 years ago (1967: 11), and as beginning in the mid-19th century (1967: 54). In the process of radical restructuring beginning some time before 1854 most people were subjected to uncertainties, as passive victims not as architects of change. In this context of poverty, powerlessness, confusion of values, popular ‘religion’ produced new forms of refuge and social protest. This process became intensified after 1868. More recently, in a post-WW2 situation of economic development and prosperity, these movements have produced new rationales.

McFarland (1967) relies on the sociologist J. Milton Singer for ideas such as “Religion is part of a complex interacting system”; “religious forces” “respond” to the “social environment”, “feed back into” the social environment, dynamically set in motion changes in the social system. But actually what real analytical work is being performed by “religion” and “religious” in these statements, except to merely affirm the putative ideological distinction between ‘religion’ and ‘society’? Nowhere does McFarlane clearly explain, and in this he is no different from many other writers.

McFarland points out that, generally speaking, ‘religion’ in the west is assumed to be an individual exclusive commitment to a church defined by a doctrine concerning the after-life. He says, “profession of faith or commitment to membership is the sine qua non of meaningful religious statistics in the west” (1967: 20). But in Japan, people tend to have multiple commitments, they find it difficult to identify there own exclusive commitment, and therefore there is a problem interpreting statistics such as those of the Ministry of Education or even of McFarland’s own survey. However, this important and potentially significant cultural difference does not lead McFarland to seriously question whether one should continue to employ the same analytical concept ‘religion’.

This point becomes reinforced when McFarland points out that Buddhism is so pervasive in the Japanese cultural context and so closely identified with family life that virtually all Japanese are associated with Buddhism. On the other hand very few of these “are committed to this religion in understanding and faith.” But then what does Buddhism mean at this level of analysis? On this understanding it is difficult for the reader to be clear how Buddhism as a religion can be distinguished from ‘Japanese culture’ or ‘Japanese values’.

Given the problematic nature of the concept of religion in Japan McFarland suggests that many western analytical categories do not fit Japan (1967: 25) and that it is better “to identify the typical Japanese perspectives on religion and, in the light of these, to consider all other religious phenomena, including the institutionalised religions.” The problem is that the perspectives that he identifies as typical of Japan - the centrality of experience and the reconcilability of opposites are so broad (“They are attitudes of mind ...not credal statements or affirmations of faith.”) that it is difficult to say what is distinctive about specifically ‘religion’. He also generalises that, whereas the west values analytical abstraction and logical oppositions, Japan values holistic experience and the reconcileability of opposites. Even if these generalisations had some value in a specific analytical context (which they may), they do not help us to understand what ‘religion’ is apart from values in general.

This problem with ‘religion’ is strengthened by McFarland’s discussion of the concept of kami, which unlike the Judeo-Christian notion of transcendence does not refer to “a unique order of being or a self-contained category of phenomena.” (1967: 24) He lists 7 different meanings that kami can have. For example, kami can refer to fundamental life principles, such as fertility, growth and productivity; celestial bodies, such as the moon and the sun; natural forces such as wind and thunder; prominent topographical features such as mountains and rivers; many natural objects such as trees and rocks; certain animals such as foxes and horses; and spirits of the dead. He might also have included powerful rulers such as shoguns and emperors. But this picture merely confirms the more general impression that Japanese cosmology and symbolic ordering of the world is not well represented by some supposed distinctive meaning of religion.

McFarland continues to refer to Shinto as “this religion” while at the same time saying that its most distinctive characteristic is “the intention to be identified through ritual with the whole range of Japanese history, tradition and aspiration. Doctrinally and ethically amorphous, jealous of absolutes...” (1967: 26) It is unsurprising that some Shinto leaders objected to Shinto being classified as a ‘religion’ and claim it as “the way of life of the Japanese people.” (1967: 27) We are talking here in such generalisations that religion picks out nothing distinctive.

This lack of grip on meaning renders statements like the following about Confucianism difficult to understand:

...it has not been as a religion that Confucianism (jukyō) has functioned in Japan......the impact of Confucianism was basically political, social and ethical in character.

(McFarland, 1967: 30)

McFarland (1967) says that modern young people may not be conscious of Confucian principles, but “without knowing it, some of their most deeply ingrained habits are traceable to this source.” But we do not know what is the extra function ‘as a religion’ that Confucianism has not had. In addition to political, social and ethical, what extra is being implied that is lacking?

McFarland says that the new religions in their first Tokugawa phase were “an attempt to solve political problems in religious terms.” (1967: 55) How is he distinguishing between religion and politics, given all the qualifications he has made about the meaning (or lack of it) of ‘religion’ in the Japanese context? He says that “popular religious developments” were characterised by 1) popular ethico-religious teachings; 2) community sponsored pilgrimages (okage mairi); 3) frenzied dances intended to effect social reforms. However, again it is not clear how ‘ethico-religious’ differs from ethics. The examples that he gives are values associated with shingaku (heart learning) and hotoku (repayment of blessings) such as sincerity, genuineness, naturalness, love, diligence, self-help, the bounty of nature, avoidance of corruption, and sloth and self-pity. Hotoku fuses salvation with economic recovery. Sponsored pilgrimages are identified with group action, protest, passive resistance, “an embryonic theory of social reform.” (1967: 56). Since pilgrimage was the only reason at that time that a peasant could leave his area, going on one became “a kind of demonstration in support of their demand for personal freedom.” He interprets frenzied dances as “the feverish desire for social reform.... the popularity of the assumption that religious faith and action are directly related to the alleviation of social and economic distress.” But what is ‘religious faith’ in this context? In a sense it doesn’t matter what one actually says; if one uses a word often enough it acquires an illusion of meaningfulness.

The language of religion and religions sets up false dichotomies. Is religion an aspect of society, or is society itself the object of religion? A leading American scholar, Byron Earhart, somewhat in an attempt to give expression to this relationship between Japanese religion and Japanese society, says:

The religious world of society is the formation and use of social units (individuals and groups) for the purpose of preserving, celebrating, and transmitting religion.

(Earhart, 1984: 69)

But he could equally well have said:

The social world of religion is the formation and use of religious units for the purpose of preserving, celebrating and transmitting society.

What difference would it make? No clear meaning is being expressed by these words. We need to deconstruct these terms, which are doing nothing to promote understanding of the things that are important to the Japanese collectively themselves. The western ideologically determined slots ‘religion’ and ‘society’ do not help us understand the way Japanese institutions work, or what values of social relations are being constructed in the performance of ritual.

My case is strengthened by Earhart’s discussion of ancestor worship and the importance of the relevant social unit for the practice of all Japanese religion: the family, the village, the nation, and other groups. It seems clear from what he says that ‘religious fulfilment’ is found in the celebration of the legitimate ordering of the group, which brings us back to a society’s set of self-representations expressing its dominant values. Proper ritual relations with the ancestors symbolise and recreate the solidarity of the family (or preferably household). Presumably this has its national analogy in the performance of a whole range of reproductive and purificatory rituals for the health and safety of the nation by the Emperor, who is symbolically the father of the nation. Thus religious fulfilment is virtually identical with social fulfilment; that is, with the reproduction of the pure, harmonious Japanese social order and the symbolic elimination of conflict and pollution.

Spirit, faith, worship, peace of mind, salvation

It is impossible to establish from any of these books that there exist any specifically religious institutions, emotions, experiences or feelings. The use of the word ‘spiritual’ may seem to offer some way out of the dilemma. In their 1998 book, Reader and Tanabe identify “spiritual” elements of genze riyaku as “peace of mind, faith and salvation” (1998: 23). Religion, then, is identified as ritual actions performed to mystical beings with an attitude of “faith” (shinkō) for the purpose of obtaining “peace of mind” (anshin) and “salvation” (kyūsai). To what degree, and by what criterion, are these values and attitudes genuinely special and set-apart (religion), and how much overlap is there with ‘secular’ ritual performance?

Reader tells that one way of finding out if his two Japanese friends, who claim they are not religious, are performing their shrine rituals religiously or not is to check out if they perform them with a “worshipful” attitude. He was expecting the non-religious friends to be “interested, respectful, but not worshipful” (1991: 1). This was because they had said they were not religious. Reader is therefore implying that “worship” and its attribute “worshipful” separates the really religious performance of ritual from the merely interested and respectful.

The problem with all of these English-language terms is that their meanings are highly complex and multivalent, particularly when translated into the Japanese context. In a monotheistic Christian context, God the Creator is self-sufficient and uncreated, and the individual creature entirely dependent. The soteriological and ontological relationship indicated by the term “worship” or “worshipful” in the Christian context would therefore be significantly different from the Japanese one. Yet even in English the word ‘worshipful’ can be used as an honorific, as when we address the Lord Mayor as ‘The Worshipful Mayor’. And we talk metaphorically about a person ‘worshipping’ his lover. It seems just as likely that ‘worship’ in Japanese would be equivalent to respecting one’s guardian, whether living or dead, by humbly offering gifts, using polite language and bowing deeply and sincerely. It isn’t self-evident that a ‘worshipful attitude’ is fundamentally different if the prostration occurs before a mystical or a human guardian.

Another problem with ideas like ‘faith’ and ‘worship’ in the Japanese context is that ordinary people hardly distinguish between kami and hotoke, and that anyhow what they do is important, the rituals they perform, and that they learnt these rituals from their parents. Reader (1991: 15-20) gives examples of correct performance of ritual prescriptions taking precedence over doctrine, belief or individual salvation. Even the new religions, in which one might expect to find soteriological doctrines concerned with the moral condition and ultimate salvation of the individual to be prevalent, exhibit the primacy of ritual action very clearly (1991: 17). The category of mystical beings – kami, ancestors, buddhas, angry ghosts, the souls of dolls and printing blocks and so on – does not seem to guarantee some special status for ‘religious’ rituals because the beings are either not properly conceptualised or, if they are, then they are an extension of this Japanese social world and need to be analysed as such.

The desire itself for practical benefits, the belief that correct actions if correctly performed and with the correct condition of the actor – especially the correct attitude of sincerity and gratitude – will bring positive benefits, including peace of mind; these principles are found lying behind most ritual action. There is an implicit assumption that one’s actions will be recognised and reciprocated. Even ordinary greetings (aisatsu) rely to some extent on an act of ‘faith’ or trust, though here the appropriate word would not be shinkō, which tends to be reserved for faith in the soteriological efficacy of unseen mystical powers, but shinyō in the sense of trusting the integrity of social relationships; trusting the goodwill of the other person, that he or she understands the implications of the greeting and will reciprocate, trusting that they have been socialised in the same Japanese code of behaviour; trusting that other people will follow the rules; trusting that they know their correct place and status and that they will behave in an appropriate way. It is this deep implication of trust, that waxes and wanes relative to social distance, that underlies so much of the Japanese concept of social order and anxiety concerning outside, non-Japanese incursions.

Indeed, a faith in reciprocity and reciprocal dependency is fundamental to all societies in one way or another, and it is clearly a profoundly important principle in Japanese life (Hendry, 1987: 204), since it underlies all social interaction: gift exchange, petitions by a junior to a senior for protection and advancement; appeals to the workers from the company chairman for sacrifices; trust of ordinary people in ‘the powers that be’; the reluctance to question authority.

Of course, faith in mystical powers to bring benefits as a result of prayer is not identical to these other examples, but nor is it so different. For hierarchy and mutual obligation is a pervasive principle in Japanese life, and the hierarchical and reciprocal relationship between the supplicant and the hotoke, and the belief that ancestral rites will transform the soul of the deceased into a benevolent guardian deity, reproduces the form of human relations. Could we not say that relations with these various kinds of mystical powers is itself a way of thinking through, symbolically representing, or even constituting relations that are essentially human?[20]

Forms of reciprocity

What I am arguing is that, while ritual relations with mystical powers obviously have some distinctive features, there is no evidence that the forms of reciprocity, the values, or the feelings expressed, are of a fundamentally different kind than a second great imaginary category of ‘non-religious’ or ‘secular’ ones. A more convincing model is that of a family of over-lapping ritual performances that share ideas about reciprocity, self-sacrifice, and dependency. In such a wider perspective, freed from the compulsion to stuff rituals and experiences into either the ‘religious’ bag or the ‘secular’ one, certain conceptual confusions, idealisations and other problems of representation will be reduced.

Spirits and the spiritual

In what sense then does the word ‘spiritual’ indicate some especially ‘religious’ meaning that cannot equally refer to any widely held values that are embedded in ritual performances throughout the spectrum of social relations?

Reader and Tanabe question the dichotomy between the material and the spiritual, which they say cannot be separated (1998: 23). But as with other key terms for identifying ‘religion’ such as “worshipful” it is difficult to find a clear discussion of the meaning of “spirit” and “spiritual”. Reader sometimes identifies the “religious” with “the spiritual realm” and contrasts it with the physical (1991: 46). Presumably the “religious” then is the non-material, what cannot be seen. But we know that the dead can be seen and felt in certain situations, and thus their ‘spiritual’ characteristics seem like material ones raised to a higher degree.

However, the word ‘spirit’ and ‘spiritual’ is used in a number of different contexts. For example, the distressed spirit of an aborted foetus causes misfortune (1991: 45). It can be the “spiritual” cause of misfortune (1991: 46). Elsewhere he refers to the Japanese view “that life is a coalition of the physical and the spiritual” (1991: 45) and that spirits are attributed to animals and inanimate objects. Spirit here means “soul”, as in “the souls of the eels” that had been killed in the eel restaurant that organised memorial services on their behalf, or as in the souls of the dolls and the printing blocks that can attain Buddhahood as a result of the performance of the correct rites. Reader describes this as an “animistic view of the world” (1991: 46) in which the “spiritual realm” (sometimes described by the author also as the “psychic”) causes problems in the physical. Many of the spirits are malevolent, often as a result of ritual impropriety, and the correct performance of rituals of purification and exorcism can solve these problems. Thus it seems there is a two-way causation between the spiritual world of souls – including unhappy ancestors, eels, aborted foetuses, dolls, printing blocks – and the efficacious performance of rites.

Reader identifies these beliefs and ritual activities as “the world of religious experience in Japan” (1991: 49).

There is, however, another aspect of organised Buddhism to which the author refers, and that is to “individual spiritual awareness” (1991: 105) and to “spiritual training” (1995: 238). In the context in which he discusses this, it seems to refer to meditation as the Soto Zen sect teaches it. It is “spiritual awareness” that, according to Reader, makes Japanese Buddhism something more than rituals of social bonding and the resolution of grief (1995: 238). But though this seems to be a very different idea of “spirit” from the souls of eels and printing blocks, it could still mean something like ‘enlightenment’, or alternatively it could mean ‘mental’, as in mental training or character training. It could also mean ‘attitudinal’. This might explain why Japanese corporations send their executives and workers on Zen training programmes, to inculcate into them a truly Japanese spirit or attitude[21].

Let us compare the use of the word “spiritual” in a completely different text by Patricia G. Steinhoff in an analysis of Red Army factions in the early 1970s (Steinhoff, 1992: 195-224). This is a fascinating account of a terrible incident of group violence against its own members, and makes important points about group organisation and dynamics in Japan. The members of the red army alliance (rengō sekigun) were hiding out in the mountains. Most of them were students who dedicated themselves to the revolution with fervour and sincerity. Their commitment to the group and its aims required from them a profound self-transformation, a casting off of bourgeois mentality and life-style, a willingness to suffer many hardships and deprivations in pursuit of a communistic consciousness and society. (There is surely an analogy here to the commitment and self-transformation required by Japanese kamikaze pilots dying for the Emperor and the nation; or to the ascetic renunciation required to become a yamabushi (mountain ascetic of shugendō); or to the practice found in Ittoen[22].) In the course of her discussion of the process of kyōsanshugika (communization), invented by the group’s leader to test whether or not any member had achieved genuine subordination to the authority and the violent aims of the Red Army group, Steinhoff says:

The notion that ascetic practices and humbling physical discipline produce spiritual rewards is…thoroughly engrained in Japanese culture.

(Steinhoff, 1992: 203)

The meaning of ‘spiritual’ here indicates strength of character to submit to Japanese group values without complaint, the ability and courage to survive under adverse circumstances, to ‘gaman’ and endure with fortitude. Similar things that she is saying about a communist group might have been said about the training of sumo wrestlers and baseball players[23], the character reform of prison regimes, school training for exams – as well as Soto Zen meditation techniques and so on.

Steinhoff connects the idea of “spiritual rewards” here with the Red Army’s infliction of severe punishment and even death on its own members in the disciplinary process of kyōsanshugika. This unusual and pathological outcome may not characterise normal Japanese groups, but it was made possible in admittedly abnormal circumstances by some of the basic values of group organisation in Japan, which Steinhoff lists as:

... deference to formal authority and unwillingness to challenge it; consensus decision-making procedures that carry a high expectations of subsequent participation; indirect and ambiguous means of expressing dissent; and high levels of commitment and loyalty to the group.

(Steinhoff, 1992: 222)

The point is that the English words ‘spirit’ and ‘spiritual’ are multivalent. Reader seems to assume that the idea of the “spiritual” clinches the legitimacy of a special and set-apart “religious world of the Japanese”. Yet in his own uses it can mean different things: in some contexts, some naively constructed world of invisible souls; in another context some special mode of consciousness attained in Zen meditation; and yet again a quite different nuance of commitment to Japanese group values or corporate discipline. If “the spiritual” is to do the work that Reader seems to require as an analytical category then surely he will need to shift the focus of the analysis.

Again, there is a concept of ascetic practice (shugyō) that is believed to lead to self-purification (a state of being in the correct condition), consciousness-raising, and self-transformation. Arguably it is a sacrificial concept, where the self is sacrificed for ritual, political, economic or soteriological ends. Surely the ultimate self-sacrifice is for the nation, conceived as a transcendental dynastic domain. The values behind a sacrificial asceticism or renunciation of the self can be found operating in diverse situations in Japan. We also find that many of these rituals are concerned with the legitimation of the social order, including the nation state, the construction of status and gender identity, the continuation of the ie or the family; in short, that they are connected to social reproduction, to politics, or more generally to the legitimation of power and authority. The performance of these rituals may express profoundly important collective values, and may be accompanied by a range of feelings and emotions. There is, however, no evidence to suggest that there are special ‘religious’ kinds of values, feelings and emotions qualitatively different from ‘non-religious’ ones, or that ascetic practice is confined to Zen meditation halls or shugendō mountain contexts; or that relations with mystical powers such as kami and hotoke are only about magic and have no wider ritual meaning.

Even the idea of “salvation” is problematic in the sense that the Japanese word kyūsai, while not colloquial, does not refer exclusively to benefits obtained from mystical beings. Reader and Tanabe have, in my view rightly, extended the meaning of salvation to include the receiving of material benefits such as longer life, health, the resolution of conflicts, in this world (1998: 22). One can thus talk about both ‘thisworldly’ and ‘otherworldly’ salvation[24]. Though rather formal, it can in principle be used (like tasukeru) (to help or save) to refer to any action whereby one person or group saves another, for example from a storm at sea, from chikan (sexual molestation) on the subway, or from unemployment.

How can we explain Reader’s dedication to an alleged Japanese religiosity when we admit that it is an alien concept? Surely we must look at its function among the aliens, and ask what ideological field it is animating.

Confucianism

Reader’s short (1991: 29-30) discussion of Confucianism[25] has the potential for generating a change of track from ‘religion’ to some alternative, among which ‘ritual’ may be one contender. He points out that Confucianism’s influence

…. may be perceived running through much of Japanese society in general, instilling ideals of order and structuralising respect for one’s elders and seniors both in family and social terms and asserting the importance of harmony as a social ideal. These ideals have made their mark also in the religious sphere: harmony has been transformed into something of a religious ideal while Confucian ethical teachings and concepts of filial piety have underpinned the Japanese practice of ancestor veneration.

(Reader, 1991: 30)

Much of what Reader is referring to here as “Confucianism” might be described as a ritualised system of values that pervades all institutions in Japan. When we combine this observation with the one about the strongly indigenising, homogenising, centrifugal Japano-centric forces that operate on things which come in from outside (1991: 28), and the many references throughout his book to the serious preoccupation with racial and national identity, then the possibility of shifting the analysis from religion to some alternative such as ritual avails itself. For ritual as a concept does not require some dubious transformation from “harmony as a social ideal” into “something of a religious ideal”. The reader needs to know what this transformation consists in, and what is involved in a process whereby an ideal changes from being a social ideal to being a religious one. What qualitatively new nuance has ‘harmony’ achieved by being blessed with the adjective ‘religious’ that it did not possess before? I hold that there is nothing in Reader’s books and articles that explains this. Yet if Confucianism could be thought of as a pervasive Japanese hierarchical ritual system, it might help to free the author from his unnecessary commitment to ‘religion’.

Zen Buddhism

In his article “Cleaning floors and sweeping the mind: cleaning as a ritual process” (in van Bremen and Martinez (eds. 1995) we can find the same difficulty. Reader says

... since my own primary area of interest and research is in the world of religious behaviour and activity, the main focus of this chapter will be on ritualised cleaning processes which occur in religious environments, and the ways in which they are imbued with and express specific religious symbolisms and meanings.

(Reader, 1995: 228)

Yet he almost immediately goes on to admit, “the borderlines between religious and cultural actions are virtually inseparable in many respects.” (1995: 228) This admission may explain why the reader never does discover what these “explicit religious symbolisms and meanings” consist in.

Zen is an interesting case in point, because ‘spirit’ in Zen apparently refers to an attitude of mind, rather than communication with invisible souls. Not only does Zen problematise the concept of a field of religion defined by ‘spirits’, but it re-integrates everything into social life without specifiable remainder:

In Zen terms any activity when pursued purely and totally is the same as, indeed is no other than, meditation, and is equally a gateway to the enlightenment which is embedded in, rather than separate from, everyday life.

(Reader, 1995: 230)

What then is the principle that demarcates Zen as a specifically religious activity, in contradistinction to those activities that are non-religious or secular? Reader attempts to locate the “specific religious meanings” in Dogen’s writing in the idea that ordinary ritual acts of cleaning are a metaphor for “polish[ing] the mind” and the way to enlightenment. (1995: 232)

The problem is that enlightenment is embedded in, rather than separate from, everyday life.

What is the difference between a ritual and a “ritualised religious practice”? (Reader, 1995: 235). It may be true and useful to say that Soto Zen practitioners typically perform actions that are not only ritualized but which have an important soteriological element. That is to say, rituals of cleaning and sweeping are part of the process whereby the mundane mind is transformed into a transcendental state of insight. The problem for the reader is to know how or whether words like spirit, heart, mind and soul carry some special ‘religious’ meaning that cannot be interpreted in terms of a social ideology of a ritual kind. This point comes to a head when the author makes a distinction between overtly religious organizations and

The numerous ethical or moral welfare training organizations (shūyō dantai) which… exhibit many of the qualities and practices of religious organisations... also make use of spiritual training sessions that manifest many of the practices and concepts used in Zen temples and the new religions.

(Reader, 1995: 237-8)

Business companies use such “spiritual training sessions” to instil things like obedience, discipline and determination in their workers (Reader, 1995: 238). But it is not made clear how these values, which are anyway instilled in workers throughout the schooling and socialisation process, become specifically religious only when found in a temple.

Near the end of this essay, the supposed specific religious quality of actions and organizations (for example “the overtly religious discipline of samu in Zen temples”) is asserted in distinction from “the apparent secular occurrences” of cleaning rituals (1995: 241), such as the community cleaning rites that the author describes. But then he lists “common threads” (1995: 241) between distinct areas of Japanese consciousness to show how difficult it is to separate the overtly religious cleaning rituals from the secular ones. These common threads are listed as: generating social identity, a sense of community, concern with definition, order and purification, of “imposing order on an environment, and purifying it in accord with the prevailing social and cultural understandings of what such purification involves”, all performed with the same attitudes of endurance, service, gratitude and joy, and performed equally “in a religious institution, a school or a housing estate” (Reader, 1995: 241). However, this leaves the reader at a loss, for if these hold so much in common at what points should we claim they differ? The common threads are more powerful and interesting than some assumed difference between the “overtly religious” and the “apparently secular”.

The truth is that you can go through this paper with a toothcomb and you will not find these meanings explicated. There is no point at all where Reader can state any clear or worthwhile distinction between the religious and the cultural, or the religious and the secular (1995: 241), just as he cannot explain the difference between a ritual and a “religiously ritualized practice” (1995: 235).

The introduction of ‘Confucianism’ offered the opportunity of a shift in perspective from the notion that ‘religion’ has some special attributes which separate it from other social institutions, to the more interesting idea that harmony is just as deep and important a doctrine in the school classroom or the sumo beya as in the shrine or temple; in fact, it is crucially important in any situation where ‘knowing your place’ becomes an issue.

Soteriology

An important element of ancestor rites for ordinary people is soteriological, in the sense that they offer a passage out of this world into the next, which may be thought of as a kind of salvation. When people think of ‘religion’, they are sometimes, perhaps often, thinking of the other world as a place where the sadness and separations of old age and death can finally be overcome, and where the lone, and perhaps lonely, individual in this world can be reunited with loved ones. But for ordinary people the next world is a kind of extension of this world. It is not a mystical absolute that transcends time and space. But still, this is a kind of soteriology for the individual.

In Diana Lynn Bethel’s account of social relations in an old people’s home called Aotani (1992: 109-134), she describes the observances, both collective and individual, at the home’s butsudan (1992: 129-130). It acts as a window through which the elderly can communicate with their own ancestors, and many hope that a dead husband or mother will come back to collect them and take them on. However, the residents also hold collective rituals for those people who died in the Aotani home, and who had their memorial tablet (ihai) placed in the Aotani home’s butsudan, instead of being taken away by surviving relatives from outside. The old people’s home thus has its own ancestors, and Bethel shows how, within the formal institutional structures, the residents construct their relationships on analogy with a household, using fictive kinship terms that reproduce the age and gender hierarchy of the household among themselves. The role of the butsudan in the lives of individuals, and of the society of Aotani, is complex, and while the theme of individual salvation to the other world is one important element, it can only be properly appreciated within a full ethnographic context of the collective reality of the kind that Bethel provides.

Ancestor rites are conducted within complex sociological situations. They are rituals, performed collectively or personally, and they have many different implications. They are a way of negotiating or constructing and reconstructing the social relations of the group, whether the ie, or the kaisha (company) (Nakamaki 1995), or the old people’s home, or the neighbourhood (Stefansson, 1995); they can provide a mystical source of help for individuals and groups; they are also a focus for an individual’s personal soteriological aspirations and hopes. Soteriology, as a doctrine of individual liberation is present in the form of images of loved ones, or perhaps an authoritative figure, returning to take them on to the next world. But the primacy of action would suggest that such soteriological elements are only meaningful when we also understand the other co-existing elements in the social and ritual context. The distinction between the religious and the secular does not seem to help with this.

‘Religio’ and ‘Religion’ in western history[26]

Finally, I return from Japan to Europe. We need some sense of the historical trajectory of the invention of our modern categories, which presumably were exported during the eras of colonialism and post-colonialism. To construct such a trajectory is far outside my capabilities, and probably those of any single author, even a major historian. But we need some kind of a story in the context of which the problems we have in applying ‘religion’ and other ideas to Japan make sense. The story that I have made up here is based directly or indirectly on historical accounts but requires exhaustive debate and modification if it is to be acceptable.

The historian John Bossy noted in a 1982 article the change in nuance that occurred historically in the meaning of ‘religion’ especially at the end of the 16th and the beginning of the 17th century:

Religio in classical Latin is a sense of duty or reverence for sacred things; derivatively, some object which inspires this frame of mind; thence a cult, or worship in general. Essentially it is a feeling, a frame of mind….In early Christianity it meant worship, a worshipful attitude….In medieval Christianity this usage disappeared. With very few exceptions, the word was used to describe different sorts of monastic or similar rule, and the way of life pursued under them: the ‘religious’ were those who pursued such a life…in the 16th century…where the Latin form Christiana religio is found, it must be translated “Christian religion”, not “the Christian religion”: we have now had to invent the appalling ‘religiosity’ to fill the gap where ‘religion’ once stood…How then did the word come ….to get its capital ‘R’; to become, as Cantwell Smith says, a Great Something, a generic object formed by abstraction from a number of specific object-instances, members of the class “a religion”, meaning…‘a particular system of faith and worship’? …The actual motor of our change was, it seems clear, the simple existence of a plurality of embodied, and embattled, faiths…Objectification arose, then, out of the need to describe one’s own or other people’s way of belief and life, as if from outside, in circumstances where a plurality of such ways had come into existence.

(Bossy, 1982: 4-5)

Bossy further argued that, since the unity of Christendom could no longer be assumed, it became apparent to observers that a shared public sphere or space outside those increasingly reified and demarcated areas of people’s lives called religions became a necessity, otherwise there would be anarchy. For civil society to function there had to be something called Civil Society. In short, Bossy argued that religion and religions had to be invented for there to be society and societies.

We can see from this that the idea of a religion and religions was not the only category to become reified. Previous to the early 17th century ‘society’ meant company or relationship, as when we might say “I enjoyed the society of Mr and Mrs Brown on Christmas day”, or “society in this village is not congenial”. It refers to human relationships, but has not been objectified into the concept of a Society. Again this reification seems to have occurred in the late 16th early 17th century, when there are examples of ‘a society’ or of ‘societies’ and ‘Society’ with a capital ‘S’. Like religion, society was being reified and constructed as a system of institutions and practices, things that can be studied, described, compared and so on.

I am not a historian so I shall be rasher than historians might like to be, and consider what other elements of the story might form a hypothesis that can be knocked down, modified or allowed to stand in the light of historical research. Other factors were at work. Relations between Europe and non-Christian parts of the world – what shall they be called? Cultures? Societies? Tribes? Pagans? Infidels? – were becoming increasingly frequent, and new ideas were needed to make sense of these different societies. ‘Religions’, ‘societies’, ‘cultures’ all became reified entities as a result of this new multiplicity that needed to be demarcated, identified, described, reported, analysed, compared, explained, controlled, used.

And then there were also the needs of the new bourgeois classes to pursue their own capital agendas without interference from Church-legitimated authorities. This point connects the trial and execution of Charles I with the American demands of “no taxation without representation”. The principles were fundamentally the same. Political representation based on election was a new form of legitimation incompatible with the traditional bases of authority and deference that increasingly seemed arbitrary. Along with new categories, often coined from old words, there is a new kind of rationality, class formations, new ideas about capital and exchange, new forms of banking, new ideas about property, the state, new ways of legitimating authority, a new basis for legitimating knowledge in ‘nature’ (natural science), and of course the concept of secular civil society.

Probably we can find connections, some closer and some more remote, between these changes in nuance of a string of important and less important categories. What I want to suggest is that the distinction between religion and the secular is a special case, because the new ideas of ‘secular society’ independent of what increasingly came to be referred to as ‘religion’ constituted a framework within which new political, legal, economic institutions were able to appear persuasive and even natural. By separating out certain things into the basket ‘religion’ you are able to remove the arbitrary interventions of the church, and of the king who is legitimated by the priesthood, from the organisation of the state. The modern state and the idea of civil society are based on a different rationality, and science becomes increasingly unencumbered by the traditional knowledge of the Bible and the Church fathers. Science and rationality are subject to different forms of legitimation than traditional knowledge.

In this way the religion-secular idea can be seen as a contested, manufactured category that forms an important part of the emerging worldview of what, after many centuries, has become our western democratic capitalism. These words acquired profoundly different nuances in the modern configuration of ideas and values; and religion required a new set of priests for its propagation, including academics like myself who write and publish papers on the religions of the world, or liberal ecumenical Christian theologians who wish to build a world theology from these different, truncated objects that have been torn from their embeddedness in quite different ideological contexts, such as the Japanese or the Indian, the Nuer or the Ashanti.

If this is in general correct, then it should teach us various things. One is that, if we want to try to understand the language game of ‘religion’, that is, to understand its contemporary uses and the manipulations that the category engenders, we have to be aware of its historical context, because if we are not, then we will have a truncated picture of how it actually works as a category in relation to the other categories that have been evolving in tandem to form a different ideological system. This leads to another connected point, that religion as a concept does not stand alone simply as something given in the nature of things; it stands in relation to the specific western modern configuration of values and categories that also has a history.

I suggest that religion and the secular are two of our categories that have a specially close linkage, that they mutually define each other, that in order to construct a concept of secular society in which trade, law, government and science were freed from the arbitrary interferences of the Church or the King, people like Hobbes, John Locke, the Deists, the American writers of the Constitution, the French Enlightenment philosophes, needed a new idea of ‘religion’ or ‘religions’ in the plural to help them do that job of making a new idea of secular society; that the problem as to what counts as religion is also the problem as to what counts as non-religion or the secular.

The decision about what is and is not categorised as religion is highly ideological, yet the word is used freely and rather uncritically as though we can all easily find religions in any part of the world and at any period of history. Most academics use the word religion without much consciousness of the way that their usage is arranging historical and ethnographic data according to a pattern that fits into the assumptions and needs of western capitalism, or western ideas about gender, the individual, about rights, or western theories of exchange and markets. This becomes of special importance in the context of colonialism and neo-colonialism, since it facilitates a distortion in our understanding of non-western ideologies and cultures.

The export of the religion-secular distinction in the colonial situation

The issue takes on added interest if we are aware of the number of anthropologists and historians who have pointed out that many – perhaps most – societies in Asia and, I believe, Africa, do not have an indigenous word for ‘religion’, but when confronted by powerful invaders with their merchants, armies, missionaries and administrators who claimed to represent civilization, have been compelled to search for and construct a suitable word from their own traditional discourses[27]. The idea that there are some special phenomena in all societies that can be described as religious or religions was not something that the people in that society decided on by themselves, but an idea received from the west, or at least negotiated into existence by local lites with the help of trading enterprises, military officers, Christian missionaries, colonial administrators and others. What constituted ‘religion’ in any given society in Asia was not a self-evident fact to the people there, neither was it indeed to the outsiders, but had to be discovered and/or invented[28].

However, the traders and many of the colonial administrators may have been more interested in establishing the institutions of secular civil society, itself a highly ideological concept including specific concepts of exchange and markets, but appearing to many as simply the natural and rational way to organise things. In most cultures that became colonised, what we call law, economics, and politics were not separated out into distinct spheres but were embedded in a different indigenous way of representing the world. It was this disembedding that was a necessary programme for the imperial power if it was to impose western-style laws, create capital markets and forms of exchange, and to ‘educate’ the people in the new school systems.

By and large the Japanese avoided colonization by adopting western categories and institutions that were supposedly the mark of more highly developed civilizations and placing them in the omote (one’s front) or tatemae (one’s public face) mode. The American written constitution with its separation of church and state and its guarantee of freedom of religion satisfies the west that Japan is really just like us, and conforms to our western assumptions about the world. But the problems of ethnographic and historical interpretation that occur when we assume that there is a religious world in Japan that can be meaningfully distinguished from another secular world suggests that the reality is not like that.

Notes

1. This article is concerned with English language constructions of Japan. Though the authors considered here are non-Japanese, I have analysed the English-language texts of Japanese authors in other publications (Fitzgerald, 1993, 1995, 2000). A most important part of research on the uses of religion and/or shūkyō needs to be done in Japanese language texts and everyday speech to analyse its functioning at different levels of discourse, such as the juridical, academic and everyday colloquial.

e-mail the Author