The sociology of religion: Problems and desiderata

18
THESOCIOLOGYOFRELIGION : PROBLEMSANDDESIDERATA RolandRobertson ProfessorofSociology,UniversityofYork InthisessayIlookcriticallyatthesub-disciplineknownasthesociology ofreligionandsuggestsomeguidelinesfor finding awayoutofsome currentdifficulties . Duringthepastfewyears,thedebateaboutsecularizationhasin- creasinglycometoconstituteasiteuponwhichthesedifficultieshavebeen mostevident.Responsestothesecularizationthesisandthemethodo- logicalandtheoreticalproblemsattendantuponthesubstantivecoreof thedebatehaveinthemselvesbecomeinterestingfrombothsociological andtheologicalstandpoints .Whatbegan,atleastinsociology,asa circumscribedscientificproblem-traceablehistoricallytothesescon- cerningwhetherornottherehadbeenareligiousrevivalinpost-Second- World-War-America-hasbecomeanamorphousproblemareainwhich manycrucialissuesthatlieattheheartofthesociologicalenterpriseasa wholeareembedded .Thisappearstobepartlytruealsoinrespectof theology .Ihavingattemptedelsewheretolaybarethemajoraxesof contention amongsociologistsandsociologically-inclinedtheologians, andtorelatethesetotheincreasingconcernwiththe critical stanceof sociologistsinrelationtothehumansocietiesinwhichtheyoperate .'The majoraxesofthedebateseemtobe :first,whethersecularizationis occurringornot ;second,whetherjudgementeitherwayisbasedupon primarilyempiricalortheoreticalgrounds ;third,whetherthepresence orabsenceofsecularizationisapprovedofornot ;and,fourth,whether thelatterjudgementispredicatedonsociologicalorreligious-meta- physicalstandpoints .Interpretationofthesecularizationdebateinthese termsconstitutes'second-level'sociologicalanalysis : ittreatssociological andquasi-sociologicaldiscussionsasdataandtherebyfocusesuponthe backgroundanddomainassumptionsandinputsof'first-level' analysts . Oneofthemajorimplicationsofthisexaminationisthat,asamatterof degree,muchofwhatappearsundertherubric`sociologyofreligion'is bestdescribed as sociotheology . Insofar as alltheologicalideasentailcon- ceptionsofthenatureofmanandsocietyandinsofar as allsociological ideasentailviewsabout`theultimate'-evenifonlytodenythesigni- ficanceofthelatter-thenclearlysociotheologyisnonewphenomenon . Whatisnewisthattherenowappearstobeamuchmoreconscious

Transcript of The sociology of religion: Problems and desiderata

THE SOCIOLOGY OF RELIGION :PROBLEMS AND DESIDERATA

Roland RobertsonProfessor of Sociology, University of York

In this essay I look critically at the sub-discipline known as the sociologyof religion and suggest some guidelines for finding a way out of somecurrent difficulties .

During the past few years, the debate about secularization has in-creasingly come to constitute a site upon which these difficulties have beenmost evident. Responses to the secularization thesis and the methodo-logical and theoretical problems attendant upon the substantive core ofthe debate have in themselves become interesting from both sociologicaland theological standpoints . What began, at least in sociology, as acircumscribed scientific problem-traceable historically to theses con-cerning whether or not there had been a religious revival in post-Second-World-War-America-has become an amorphous problem area in whichmany crucial issues that lie at the heart of the sociological enterprise as awhole are embedded. This appears to be partly true also in respect oftheology . I having attempted elsewhere to lay bare the major axes ofcontention among sociologists and sociologically-inclined theologians,and to relate these to the increasing concern with the critical stance ofsociologists in relation to the human societies in which they operate .' Themajor axes of the debate seem to be : first, whether secularization isoccurring or not ; second, whether judgement either way is based uponprimarily empirical or theoretical grounds ; third, whether the presenceor absence of secularization is approved of or not ; and, fourth, whetherthe latter judgement is predicated on sociological or religious-meta-physical standpoints. Interpretation of the secularization debate in theseterms constitutes 'second-level' sociological analysis : it treats sociologicaland quasi-sociological discussions as data and thereby focuses upon thebackground and domain assumptions and inputs of 'first-level'analysts .

One of the major implications of this examination is that, as a matter ofdegree, much of what appears under the rubric `sociology of religion' isbest described as sociotheology . In so far as all theological ideas entail con-ceptions of the nature of man and society and in so far as all sociologicalideas entail views about `the ultimate'-even if only to deny the signi-ficance of the latter-then clearly sociotheology is no new phenomenon .What is new is that there now appears to be a much more conscious

no

THE SOCIOLOGY OF RELIGION :

weaving-together of social-scientific ideas and theological or religiousideas. From the sociological side, two currently prominent examples will,I hope, suffice to make the point . In her book, Natural Symbols, MaryDouglas focuses upon the problem of the kind of symbol system andritual practice which are most appropriate for the social structure ofmodern societies. 2 In other words, she addresses herself to problems ofa cosmological nature on a prescriptive basis . In contrast, Robert Bellahargues in his book of essays, Beyond Belief, for an adherence amongstudents of religion to what he calls symbolic realism . 3 For Bellah symbolicrealism, as opposed to forms of symbolic reductionism, is characterized byits emphasis upon the ontological reality of all symbolic communication .Such a position has to be argued in conjunction with a commitment tothe view that the major task of the contemporary sociologist of religion isto grasp the religious meanings which are expressed in such communi-cation ; and that this is best done by synthesizing sociological analysiswith the work of those who tap ultimate meaningfulness in such spheresas philosophy, theology, and artistic literature. The crucial differencebetween the two perspectives, and it is a crucial difference, is that whereasBellah argues for the autonomy of symbols-for symbols as being the`ultimate' reality-Douglas quite clearly sees symbols as communicativemodes which are contingent upon a prior social-structural reality . Inprimitive societies the fit between social structure and symbolic structure-between, for example, the structure of kinship relations and the structureof beliefs about the cosmic significance of animals-is relatively unprob-lematic, in the sense that the social structure is clearly articulated andregulated . 4 In contrast, the structures of literate societies are more`messy' and therefore less easily related (subjectively and objectively,phenomenally and analytically) to symbols and thought-categories .Hence the prescriptive nature of Douglas' analyses of modern societies .The work of Douglas may be seen more generally as part of the con-temporary tendency among some social scientists to puzzle about andsuggest solutions to such issues as the most appropriate or optimal re-lationships between man and his natural environment ; the problematicsof the best fit between culture and social structure, between culture andpersonal experience as located in social situational contexts, and so on .In brief, then, Bellah and Douglas respectively contribute to two strandsof a social-scientific concern with what may best be described as cosmicmeaning. Bellah talks about the ways in which religious symbols conveyhuman anxieties, aspirations and preferences, while Douglas attempts todiscover the precise nature of the social-structural features of modernsocieties which lead to problems at the level of symbolic communicationin talking about these existential problems of meaning . It should be addedthat Bellah seems more concerned with religious-cultural content, whileDouglas is more preoccupied with the form and structure of symbolic cultureand associated ritual .

I describe these kinds of intellectual concern as sociotheological be-cause they are concerned in a relatively prescriptive manner with the

PROBLEMS AND DESIDERATA

III

social dimension of cosmologies . s They address questions which traditionallyoccupied theologians from a sociological standpoint . The addressingis done, moreover, particularly in Bellah's case, in a relatively self-con-scious manner-which is demarcated from the more traditional neutral-scientific posture of sociologists and anthropologists of religion .

The weaving together of sociological and, in the very broadest ofsenses, theological modes of enquiry should be contrasted with the argu-ments of those sociologists of religion who have come to see an in-builttension between sociology of religion and religious-theological orien-tations-which involves a tendency to denigrate the `ultimate significance'of sociology itself. Thus, we may note in this connection the shift in theposture of Peter Berger : from arguing that social science can fruitfully beregarded as a profane auxiliary to the Christian faith to contending thatsociology is inherently a de-transcending orientation to sociocultural lifeand therefore, from his point of view, constricting in what might be calledits `cosmic reach' . 6 In turn, we can place Berger's position against thosesociologists of religion who continue to argue for the basically scientificstudy of religion. This involves the view, implicit or explicit (thoughusually the former), that religion may be seen as an object of analysis inthe sense that sociology can transcend religious belief and experience .?

It is clear that anyone who purports to analyse X places himself in a`superior' relationship to X . It is also clear, however, that there is a widerange of variation in the ascription of objectness to psychological, socialand cultural phenomena . In its popularized meaning, positivism clearlyentails an extreme form of objectification in this respect-the object ofstudy is not permitted to `react' . No subjectivity is allowed for in thephenomenal world-or if subjectivity is countenanced it entails a highlypredetermined analytic closure . Within the sociology of religion the as-signation of such closure can be traced back to the origins of the discipline-mounted as it was in contrast to the Religionswissenschaft of such in-dividuals as Scheler and Otto, a school of thought which itself extends backto Schleiermacher . The sociology of religion developed from within thematrix of Religionssoziologie. In the latter, as Davis has pointed out, wecan include Durkheim's definition of religion as manifested collectively insocieties sui generis ; Simmel'sfunctional orientation to the study of religionand his idea that culturally defined religious sentiments and actions aregeneral to a large range of spheres of social behaviour ; and Weber'sattempts to show the interconnections between religion and other socialspheres . 8

Religionssoziologie was clearly more comprehensive and wide-rangingthan the sociology of religion came to be in its major seat of operation inAmerica during the I95os. Be that as it may, it is evident that during thepast few years the problems which defined the relationship betweenReligionswissenschaft and Religionssoziologie have come to the fore again .Problems of a basically philosophical-anthropological nature, which inturn have definite implications for the nature of the sociological enter-prise itself, are much more frequently discussed nowadays-to the point

112

THE SOCIOLOGY OF RELIGION :

that it is the case with some sociologists of religion that they have become`religious about religion' . 9

What we are witnessing in the modern period is a fundamental re-appraisal of the relationship between science and `the world' . The resur-rection of themes which were involved in the relationship betweenReligionswissenschaft and Religionssoziologie are probably best viewed in thecontext of this larger debate. Standing between and linking these problemareas is that of the revitalized attention which is currently being paid tothe relationship between science and, religion . As Barbour has succinctlystated, the dominant feature of the science/religion relationship in recentdecades has been that of co-existence between, on the one hand, neo-orthodox or existentialist theology, and, on the other hand, positivisticconceptions of science :

In such a combination, the separation of the spheres of science andreligion is enforced from both sides . The metaphysical disclaimers ofmany scientists and philosophers today are welcomed by these men,for they help to `clear the field' for religion by undermining rival natural-istic faiths which once claimed the support of science . Neo-orthodoxwriters even welcome positivistic attacks on natural theology . More-over, if science leads only to technical knowledge of regularities inphenomena, and if in addition philosophy is confined to the analysis oflanguage, then religious faith is outside the scope of possible scientific orphilosophical attack. The independence of the two fields is guaranteedfrom both sides if each is restricted to its own domain .'0

Undoubtedly, this has been a resilient 'alliance'-but my argument isthat it has become much less resilient, and, like Barbour, I am concernedto contribute to and accentuate this latter tendency . It should also be saidthat the social-science/religion relationship is much more problematic andhas been less clearly established than the relationship between naturalscience and religion."

The study of religion has historically been at the heart of the sociologicalendeavour .'2 There is admittedly some irony in the fact that for the pastfour decades or so, few, if any, studies of religion have attained the statusof being widely regarded as potential classics within the domain ofsociology, while the sociology of religion has in the same period beenregarded as something of a soft version of the developing science . But it isbecoming increasingly clear that centrality of the study of religion (in thebroadest sense) is being restored . In the earliest phase of the developmentof sociology-roughly the first half of the nineteenth century-religiousbeliefs were regarded as rivals to the scientific sociological thought-styleby men such as Comte and Marx. Evolutionistically-inclined anthropolo-gists and sociologists continued this trend until in the z 8gos the con-servative reaction led to a more tempered conception of religion on thepart of the sociologists of the classical period (roughly z8go-ig2o) . Intheir different ways, sociologists such as Simmel, Durkheim and Weber

PROBLEMS AND DESIDERATA

iii

`validated' religion, and the religious matrix came to be regarded as thecore or centre of sociocultural systems . It has frequently been noted thatmany of the standpoints, perspectives and conceptual apparati which wehave inherited from the sociologists of the classical period were wroughtin the process of analysing religious phenomena. Concepts such as alien-ation, anomie and charisma ; substantive themes such as the significanceof culture as a sociological variable, personal identity, symbolism andritual-all of these were part and parcel of sociologies of religion ofnineteenth-century and early twentieth-century sociology .

In contrast, the four decades or so following the end of the classicalperiod witnessed a preoccupation with themes of a more Realpolitiknature : power, coercion, exchange, class, and so on. This same periodalso saw the development of `tough minded' empiricism and associatedmethodological and technical concerns . To be sure, a prominent strand ofsociological thought since the late i 93os has focused on the themes sug-gested by Durkheim et al ., namely, the action theory propagated byTalcott Parsons and his pupils . Much of the latter was, however, tied toa highly self-conscious commitment to sociological professionalism, thelegitimating focus of which was the value-neutrality argument of MaxWeber. The major effect of this commitment was an avoidance of themore thorny aspects of the relationship between scientific-sociologicalclaims to `know the world', and the-in principle-competing claims ofsystems of religious belief, ideological belief systems and so on .

During the i g6os sociologists became increasingly involved in debatesabout the raison d'etre and sociocultural significance of their discipline .Among the more important factors which contributed to this develop-ment were : pressure upon sociologists in most societies to justify theirexistence in terms of policy-formation ; the conspicuousness of sociologydepartments in controversies in institutions of higher education ; and,not least, the intellectual and moral challenges thrown up, particularly inthe second half of the decade, by seemingly intractable societal problems.Thus in the mid- and late-ig6os, there rapidly developed concerns withproblems of an applied or policy-oriented nature . In turn this has led toa renewed concern with the idea of sociology as a critical discipline .

It is very clear that during the past few years, sociologists-at leastin the Western world-have become increasingly drawn into debatesabout the quality of the societies in which they live and work . This drifttowards a fairly comprehensive critical sociology should not be construed asbeing necessarily of a `radical' nature. Sociological criticism encompassesall styles of intellectual work which deploy sociological insights in theassessment of the attributes of any sociocultural system or group of suchsystems. Thus, critical sociology may be either `conservative' or `radical' .What is common to all forms of critical sociology is the use of sociology asan apparatus for analysing and evaluating the operation of humancollectivities. I deal here with some aspects of the development of acritical sociological style-chosen mainly in terms of their relevance tothe sociological study of religion .

H4 THE SOCIOLOGY OF RELIGION :

With the benefit of hindsight it seems remarkable that such principles asvalue-neutrality and the denial of the philosophical and epistemologicalramifications of sociology survived so long. One of the major ways inwhich such resilience was maintained was undoubtedly the low degreeof attention given to cognitive variables in the development of sociologyduring the period 192o to the early ig6os . The somewhat paradoxicalcombination of an emphasis upon normative factors with variations onthe value-neutrality maxim constituted the major props of this neglect .The scientific ethos-which sociologists among all scientists haveprobably subscribed to most self-consciously-has itself heavily stressed abasic cognitive element . That is, it has stressed the utmost significance ofcanons of rational enquiry, of processes of knowing, of comprehending,and so on. On the other hand, there has been an enormous reluctance toconsider and analyse the `equivalents' of such aspects of `everyday',phenomenal life .

Martins has brilliantly etched many of the problems involved in therelationship between the internal structure of sociological knowledge andpractice and sociological perception of the structure of societies, noting inparticular the inconsistency in the position of those who argue in favour ofa consensual methodological model of and for sociological enquiry and whoyet, at the same time, are highly critical of those sociologists who hold toa consensual image of the operation of human societies and collectivitiesgenerally . 14 One of the major implications of this kind of analysis is thatthe boundaries between analysis and object-of-analysis become in animportant, but complex, sense broken down. The process of breakingdown consists in the facilitation of `flows' from the sociological to thesociocultural spheres and vice versa . If one demands consistency in therelationship between intra-sociological and image-of-extra-sociologicalstructures one opens one's own criteria for sociological investigation toinfluence by `the world' and conversely one's image of `the world' is madesusceptible to change and re-shaping in the light of one's investigationalcriteria . It should be stressed, however, that a situation of great fluidity inthis respect would make sociological, or any other kind of social-scien-tific, inquiry impossible . What is more likely to happen is that sociologistshaving been made aware of the problem of the degrees of synchronizationbetween, on the one hand, sociological theory and methodology and,on the other hand, the operation of sociocultural life-i.e. the object ofsociological analysis-will be more prepared to apply critically theirtheoretical and methodological canons to the `real world' . In other words,the various fallacies which sociologists inflict procedurally upon them-selves become equally applicable to non-sociologists . To take an elemen-tary example, one of the long-standing methodological tenets of sociologyhas been the proscription against the commission of the fallacy of mis-placed concreteness (deriving from the philosophy of Whitehead) . Indealing with a religious movement which is ultra-biblical in its doctrinalstance it would behove the sociologist to apply the strictures relating tothis methodological fallacy to the movement. 15

PROBLEMS AND DESIDERATA

115

I will deal here with a sociological and anthropological theme whichhas to do both with the problem of synchronization between the real andthe analytic and the history of the sociology of religion .

The attacks mounted against schools of so-called intellectualism in thefield of anthropology have been so successful that it is not until recentlythat a few anthropologists have been able to gain some respectability forwhat is pejoratively known as neo-Tylorism .l 8 The reference to Tylor isestablished in terms of Tylor's representativeness of a school of thoughtwhich above all placed emphasis upon primitive man's religion and magicbeing primarily cognitive in their underpinnings . In other words, whatbecame known in sociology, in the special and idiosyncratic senses in whichParsons used the term, as positivism was the major issue . 17 Anti-intellectu-alists disparaged the phenomenal-congitive emphasis on many grounds-but the most important way in which this was done was through a rathercomplex exercise in Western ethnocentricism. Since the only way inwhich anthropologists apparently considered man could operate cogni-tively was in terms which are part and parcel of Western culture, it wasthought to be completely wrong to impose this upon the native inhabi-tants of primitive societies and the cultures which those inhabitants hadproduced and sustained . It was after all in the view of the anti-intellectu-alists precisely this kind of thinking which had produced the, what wasto them, obnoxious and invalid pre-logical mentality thesis of Levy-Bruhl . 18 But it could be said, and indirectly this is what Levi-Strauss hasbeen saying for many years, that rejection of the Levy-Bruhl argument inanti-intellectualist terms is really to go even further than Levy-Bruhl isalleged to have done . 19 For not only does the argument deny that nativesthink pre-logically-it really denies that they think at all in the sense thatWesterners do. It is worth saying at this juncture that Levi-Strauss,regardless of the minutiae of his binary-opposition approach to the studyof complexes of myth, cannot be regarded as having provided a com-plete restoration of the cognitive apparatus of the members of primitivesocieties-for his Freudian bias constrains him to talk of the cognitiveprocesses of primitives as being sub- or pre-conscious ; that is, Levi-Straussallows relatively little in the way of conscious explanatory and problem-solving thought processes and patterns to the primitive . In any case, Levi-Strauss `gives', so to speak, the problems to the primitives : their real and`ultimate' problem is to overcome the opposition between culture andnature. 20 It is assumed that all societies accomplish this-althoughprimitives happen to do it according to Levi-Strauss, more elegantly-with less `symbolic waste'-than modern Westerners . Embedded in suchan approach is the view that it is only out of a Western form of culturethat a comprehension of the thought processes of primitive societies ispossible. Willy nilly the Levi-Strauss view is only sustainable in terms ofthe development of Western culture having provided us with a form of arationality which facilitates our comprehension of the thought processesof primitives .

Few, if any, social scientists would wish to deny this elementary point-

116

THE SOCIOLOGY OF RELIGION :

that Western culture has over many hundreds of years provided an`analytic lift' which makes possible the study of other societies and adouble-levelled form of sociocultural reflexiveness . 21 The point is thatmany would now claim that too much has been made of it . Probablythe first group in the modern period to make serious attempts to probeand restore some sense of balance to this problem were those philosopherswho tackled the problem of rationality . Although beginning in largepart as an attack on the idea of a social science as such, the debate aboutrationality has been most fruitfully and illuminatingly conducted withreference to the problem of understanding `other cultures' .22 In turn thedebate has now coalesced with the much larger and even more crucialproblem of the psychological, social and cultural nature of analyticinquiry itself. The upshot of this conjoining of the psychology, sociologyand philosophy of science is that the gap between the nature of scientificenquiry and the nature of phenomenal enquiry has been substantiallyreduced. On the one hand, the injection into the study and adumbrationof the principles of scientific enquiry of 'non-logical' considerationsproduces a diminution of the analytic stature of the scientist in cross-cultural perspective ; on the other hand, interest in ethnoscience, thescience of the concrete, ethnomethodology, and the like, elevates theanalytic stature of the `ordinary man' . In fact, so much has the picturechanged that some can argue that the Western scientist has somethingto learn analytically from the ways in which primitive cosmological systemsare constructed, sustained, and modified .23

The sociology of religion has tended during the past few decades toignore the cognitive side of religious commitment-although a few sociolo-gists have attempted to inject cognitive factors into the sub-discipline . 24There has been a dominant tendency to consider symbols and values asmore relevant, at least in work of a theoretical as opposed to descriptivenature . 25 Where beliefs have been emphasized their structures have beenrelatively unexplored . Even more disconcerting, the kinds of religiousbelief which have been tapped have almost invariably been of the surface,institutional kind. That is, sociologists interested in beliefs have typicallylimited their analyses to the `official' beliefs manifested in organizationally-promoted settings .26 The remainder of this essay considers these and as-sociated characteristics of the sociology of religion .

In sociological analyses of religion there has almost invariably been atendency to assume that religion, or rather religious belief and/or exper-ience, is in some way phenomenallyfundamental. Fundamentality is character-ized most succinctly and definitively in the `ultimate concern' position-as exemplified in the work of sociologists such as Parsons, Bellah, Yingerand Lenski-taken over from the theology of Tillich . 27 In slightly differ-ent vein, Geertz' characterization of religion as a cultural system-in-volving an emphasis upon the templatic function and significance ofreligion-also assumes religion to be the most basic orientational matrix ofindividuals in relation to sociocultural life and its natural environment,

PROBLEMS AND DESIDERATA

117

and to operate similarly at the level of the social collectivity . 28 The stresson the cybernetic significance of religion in Bellah's more recent essays"and Parsons' cogent argument that Geertz' characterization of religionis more or less a restatement of the Weberian . `grounds-of-meaning'stance point to the fact that the fundamentality of religion is a relativelytaken-for-granted feature of the culture of much of sociology, past andpresent. 30 This is not to argue-as far as some sociologists of religion areconcerned-that religion is always `there' . In fact, among those whoadhere to the `fundamentality-of-religion' stance there is a good deal ofvariation in this respect . Parsons, and, even more so, Bellah, seem to argue,in full-blown functionalist style, that one should define religion preciselyin reference to its topmost hierarchical significance in (cybernetically-conceived) systems of human action ; in terms, that is, of what it does forthe system. Parsons, in fact, actually now includes the supernatural orultimate-reality realm as a boundary circumstance of the total system ofaction." Thus, what is culturally most fundamental (whatever that maysubstantively, in either sociological or phenomenal respects, mean) is bydefinition `religion' . In contrast there are those for whom the degree ofreligiosity in a total action system is a matter of variability (which it islogically impossible for Parsons and Bellah to admit, or at least logicallyimpermissible). Such `admission' of variability has, of course, come to ahead in the modern period in terms of the secularization debate .32

Those who advance theses concerning the occurrence now or in the pastof processes of secularization rarely, if at all, confront the problem as towhether `something' takes the place of religion . They could argue in modi-fied functionalist form that problems of ultimate meaning and experiencehaveto be faced and phenomenally `resolved' in sociocultural life and that,therefore, the `something' does what religion did previously . Those whospeak in terms of `secularized religion' would fall into this category .Durkheim came close to this perspective in his concern to develop a newform of culture in French society . They could, alternatively, argue, asMarxists almost invariably have, that the sociocultural and personal needfor religion disappears as a concomitant, determinant or consequence ofsecularization . Weber's (derivative) ideas concerning the disenchantmentof industrial society are clearly relevant to this approach . Another wayof considering the problem is in terms of the query which Schelsky posesconcerning the possibility of institutionalizing what he calls `continualquestioning' .33

The sociology of religion suffers-like many other sociological specializ-ations-from lack of co-ordination between theoretic statement andempirical enquiry . Theory tends to be pitched in highly nominalisticfashion at the level of relatively a priori forms of reasoning about thesocietal significance of religion; whereas empirical work has been pitched,as I have already noted, at the conventionally-understood surfaces ofreligious commitment. The major deficiency of the former is that itsemphasis upon the fundamentality of religious belief and/or experiencehas little, if any, foundation in empirical enquiry. The second, empiricist

118

THE SOCIOLOGY OF RELIGION :

stance is deficient in its institutional and 'surface-oriented' foci . This is adeficiency not simply for the reason that religion, on virtually all extantdefinitions, can be, and empirically is, manifested outside of collectivitiesself-declared to be religious . A more profound aspect of the deficiency oftypical empirical enquiries of the modern period is that they fail to tapprecisely the questions raised by the `pure' theorists .

Undoubtedly much of the dissatisfaction with the current state of thesociology of religion as expressed here hinges upon the very use of theterm `religion' . I would argue that many problems which lie at the heartof the sociology of `religion' are severely limited by focusing specificallyand directly on the phenomenon of religion or religiosity . 34 This, itshould be strongly emphasized, should not be seen in the present contextas a plea for a particular form of definitional approach to the term`religion', nor is it in any respect an argument against being concernedwith conceptual mapping, analytic consistency, and so on . What I amsaying is that the themes addressed by the Parsonians and neo-Parsoniansare of fundamental sociological importance ; but that the phenomenalsignificance, the `everyday' fundamentality of these factors has hardlybeen tapped, let alone tested . The empiricists have operated in `official'institutional terms and have not attended to the kinds of problems raisedby the theorists.3s It follows that what is needed is an empirical form ofenquiry into the kinds of theoretic issues raised concerning phenomenalfundamentality.

In assessing the state of a disciplinary domain, one inevitably operatesin terms of a process of typification . Thus, there are undoubtedly a fewexceptions to the way in which I have characterized the present conditionof the sociology of religion. There are a few sociologists who have attemptedsome empirical application of grounds-of-meaning schemata, and thereare a few who have attempted to give some empirical content to the re-ligion-as-ultimate-concern approach (for example, Yinger) . 36 But I wouldstill maintain that we are well short of any kind of fruitful breakthrough .What kinds of recent development are relevant in terms of achieving thekind of breakthrough at which I have been hinting? Of all recent switchesin emphasis among those whose interests cluster in the sociology-of-religion domain probably the most likely and promising contributors arethose who have attempted to knit the study of religion to the sociology ofknowledge-a tendency which has increasingly constrained our atten-tional foci in the direction of a comprehensive sociology of culture . Otherrelevant developments include anthropological work in the fields of cog-nitive anthropology and 'ethnoscience' . Less obviously, but, as I willtry to substantiate, even more promisingly, the stance known as ethno-methodology is likely to be particularly helpful-although in a formwhich has to be severely modified if it is to assist us in the respects whichthe sociology of religion needs assistance . 37

There are many dangers that what is described by its practitioners asethnomethodology will dissipate into a form of `sectarian limbo' unlesssociologists manage to attenuate their strong tendency to regard styles of

PROBLEMS AND DESIDERATA

iig

analysis in some way as the preserve of those who adumbrate them .Simply regarded as a school of sociological thought ethnomethodologyis fairly easy to damn . 38 What I am concerned with is the distillation ofthose elements of the ethnomethodological approach which are relevant tothe problems I have pinpointed in the sociology of religion .

Prima facie ethnomethodology is relevant to the kinds of problem whichhave emerged in recent studies of religion by virtue of its commitment tocomprehending systematically the ways in which individuals and groupspresent themselves in reference to their practical everyday concerns. 39In ethnomethodological terms the key concept here is that of indexicalexpression . Garfinkel's exploration of what the phenomenal term DOA(Death on Arrival) means to those who have to cope with the problemof assigning 'deathness' is a particularly good example of this . Translatedproblem-wise to the institutional sphere of religion, one could envisagesimilar analyses on a sociological basis of 'God-talk' (sociological, that is,in contrast to the many philosophical explorations of God-talk) . Thisin itself might well be an extremely illuminating and worthwhile enter-prise. But it would not, however, quite meet the requirements which I amattempting to stipulate here .

The failure to meet the requirements stems from what I take to be amajor deficiency, or at least a lacuna, in the procedures of ethnometho-dology. This concerns limitations arising in turn from the Schutzianunderpinnings of ethnomethodology . 40 Indeed both Berger-Luckmannand ethnomethodological forms of Schutzianism suffer from the majordisadvantage of their being so highly relativistic. In different ways, bothstances are committed to viewing the sociocultural world in terms offinite provinces of meaning-such that there is a built-in tendency eitherto examine particular institutional sites or clusters of ethno-epistemiccommunality . This both inhibits the development of a generalized, or atleast generalizable, set of concepts and/or findings, and makes it exceeding-ly difficult, if not impossible, to probe ethno-ontologies and ethno-epistemolo-gies in any thoroughgoing manner. Literally interpreted, the style knownas ethnomethodology probes only phenomenal methods of expression andprocedure in discrete domains of sociocultural life . The tendency to stressmethods sways attention from non-domain-specific fundamentalities-although practitioners of ethnomethodology do appear to commit them-selves in principle to the analysis of domain-specific ethno-ontology andethno-epistemology. The Berger-Luckmann approach while addressinggeneralo the various domains of sociocultural life-mainly of the `everyday'kind-does not deal very much with the cultural substance of what Holznercalls epistemic communities. 41 Instead it concentrates upon the structureof phenomenality.

The common characteristics of the new phenomenological sociologyof knowledge (Berger-Luckmann) and ethnomethodology centre upon thethrust towards a comprehension of the deeper layers of social existence .This thrust is in some respects in alignment with the motivation to getat grounds of meaning in the Weber-to-Parsons tradition . The challenge

120

THE SOCIOLOGY OF RELIGION :

lies in bringing them much more into alignment-such that the empiricalprobing of the former, or more accurately the empirical probing whichthe former logically and prescriptively entails, is constrained in the cultural-substantive directions indicated by the latter . In the dogmatic terms posedby each school this may seem a formidable enterprise . On the ethnometho-dological side, in particular, the high-flown objections to sociologicalconstructionism (which is an objection to all non-ethnomethodologicalsociology) can be seen as the main barrier to breakthrough and syn-thesis. 42 Less problematical, but still containing difficulties, is Parsons'objection to examination of the elementary forms of social life-since heregards these evidently as forming merely the substratum of socioculturalsystems. But in one way or another sociology has always been plagued bycontroversy and cleavage along that analytic axis. 43

It is clear that in the terms posed here, the ethnomethodological formof probing has to be woven together with a degree of constructionism-and the probing has to be mounted in terms of constructed categories,but not necessarily carried through to completion in terms of the initiallyconstructed categories . To take the most important example one couldimagine at this stage : even though I have here spoken much about groundsof meaning, it does not follow that there really are, existentially, groundsof meaning, in the sense of individuals operating from, or on the basis of,a set of cultural premises about the working of social life . We mightwish-and presumably an ultra-Parsonian would have to so wish-tolabel sociologically the operative premises (the fundamentalities) asgrounds of meaning : but that does not mean that individuals actually inan ontological sense have grounds of meaning . All empirical enquiry hasto start somewhere-but it does not mean (although most of modernsociology does precisly this) that it has to end `there' .

How would one go about constructing a beginning set of categories?One of the very few explicated schemata is that of Jan Loubser (whodraws on the work of Bellah and Kenneth Burke) . 44

There are criticisms of or doubts about this schema which could easilybe presented here. However, assuming the prima facie attractiveness of sucha schema, or a variant of it, what problems would arise in connectionwith its deployment?

The major problem is how one gets at the dimensions of meaningwithout actually assuming that they are subjectively felt or discerned . Itwould be easy to imagine an operationalization of the concepts in columns

Dimension of Meaning Existential AnxietyReligious Action

Providing Meaning

Being Meaninglessness FaithBelonging Loneliness WorshipDoing Guilt MissionLiving Survival Therapy

PROBLEMS AND DESIDERATA

121

i and 2--especially if we were to take Berger's argument that a problemof `meaningfulness' has arisen `not only for such institutions as the stateor of the economy but for ordinary routines of everyday life. The problemhas, of course, been intensely conscious to various theoreticians . . .but there is good reason to think that it is also prominent in the minds ofordinary people not normally given to theoretical speculations and in-terested simply in solving the crises of their own lives .' 45 In this argument,Berger echoes Weber's statement concerning `the secret anguish of modernman' . 48If we knew that these and many other similar observations about

`modern man' were true then methodologically at least there would belittle difficulty in or objection to simply going into `the field' and question-ing individuals about their sense of meaning and attendant existentialanxieties (to stay with Loubser's presentation) . We could then regardcolumn 3 of the Loubser schema as manifesting testable hypotheses-although we might wish in the spirit of this essay to erase the closureimposed by Loubser's use of the term `religious action' .

We thus arrive back at the phenomenal status of grounds of meaning .Simply in terms of Loubser's schema we need to know whether we haveto take grounds of meaning as given-that is, as universal problems ofsociocultural life which all individuals and/or sociocultural systems haveto meet. Could it be that grounds of meaning are only `resorted' to whenan existentional anxiety is felt? Or (as expressed in Homans' attempt tocope with the ritual/anxiety problem) 47 should we entertain the possi-bility that `action' in Loubser's sense actually creates anxiety and/or aground of meaning?

The drift of argument in this paper has constrained us to contend thatwe have to hold such queries as being open to empirical enquiry. In termsof Parsonian and neo-Parsonian sociology this means giving equal atten-tion theoretically and methodologically to controlling and conditioningprocesses ; individuals may be more or less control-oriented or condition-oriented. Very generally speaking, the mode of enquiry which ought to befollowed, then, is set out for different purposes in Smelser's Theory ofCollective Behaviour . That is, we should envisage a hierarchy ranging fromgrounds of meaning through to the `social technology of everyday life' . 48There is clearly some (fuzzy) parallel here between the way this stand-point has been presented and Bernstein's distinction between elaboratedand restricted coding processes . 49 That is, individuals with explicablegrounds of meaning are relatively elaborate-those who operate at thesocial-technology level are restricted . However, in so far as what I havesaid here can be related to the elaborated-restricted distinction the mostchallenging problem would be to tap the dynamics of the relationshipbetween the two coding processes . That is, for any given individual onewould wish to know whether there are occasions of switching off at one leveland moving to another or `living' entirely at one level. One would wishto know, for example, whether relatively `unquestioning' individualsencounter gaps which need ethno-explanatory filling-such that there is

122

THE SOCIOLOGY OF RELIGION :

(objectively, from the standpoint of the 'constructionist' sociologicalobserver) escalation to a higher level . What is the nature of the `eventswhich hit us'? Are they universally of the same kind and how do peoplecope with those hurtful (in a very broad sense) events? What puzzlespeople ?

The question of puzzlement is critical in the present context ; anda particularly good example of its critical status is provided by the verytitle of the Mass Observation report : Puzzled People .b0 There is in factno evidence in that report that `the people' are at all puzzled . The puzzleis sociological, not necessarily phenomenal, or existential. How, thesociologist has asked, do individuals maintain seemingly inconsistentbeliefs? By what right or analytic-interpretative precept do we judgethese beliefs to be inconsistent? Maybe they are, maybe they aren't-butas sociologists we have no way of knowing without delving much moredeeply and `sensitively' into questions of belief-salience, day-to-daysignificance, and so on .

It has been increasingly stressed by anthropologists in recent years thatfar too much attention has been paid to (anthropologically and sociologi-cally constructed) religious belief systems or cosmic maps and far toolittle attention paid to what is sometimes called practical religion-the`religion of everyday life' . 51 Both anthropology and sociology have beenguilty of posing problems for the people they study. In sociology probablythe best example of this is the soteriological tradition which we haveinherited from the work of Max Weber. As is well known, one of themajor themes-possibly the major theme-in Weber's sociology of re-ligion was that of the problem of theodicy . In examining the majorreligious traditions, Weber almost invariably tended to raise questionsabout the way in which the members of a society `solved' the theodicalproblem . 52 But what evidence do we have to suggest that people alwaysperceive that there is a theodical problem to be solved? To be sure, it isrelatively easy to pinpoint aspects of belief systems which appear to relateto what we as sociologists, theologians, or whatever, discern from ourstandpoint to be a theodical problem. In this connection it is useful toinvoke Berger's characterization of the Buddhist `solution' . Speaking of`original Buddhism' Berger says :

Gods, and demons, the whole cosmos of mythology, the multitude ofworlds of the Indian religious imagination-all these disappear, not byexplicit denial but by being declared irrelevant . What remains is man,who, on the basis of his right understanding of the laws of being . . .rationally sets out to fashion his own salvation and ultimately to attainit in nibbana (or nirvana) . There is no place here for any religiousattitudes except the coolness of rational understanding . . . . (T)heproblem of theodicy is solved in the most radical manner conceivable,namely by eliminating any and all intermediaries between man andthe rational order of the universe. . . . (T)he problem of theodicy

PROBLEMS AND DESIDERATA

123

disappears because the anomic phenomena that gave rise to it arerevealed to be but fugitive illusions. . . . 83

Berger loosely poses the question as to whether this form of thoughtconstitutes an `overreaching' of rationality-and the fact that magicalanimism has frequently co-existed with one or other of the major variantsof Buddhism shows that such a system of thought is `too rational' foreveryday life.5' Besides which, Theravada Buddhism has tended to beadhered to by laymen in terms of a `lower' form of Buddhist religiousactivity, involving merit-making-that is, the material support of templesand monks. Tambiah has elaborated the conventional wisdom in thisrespect by showing a number of reciprocal oppositions relating the detailsof religious behaviour to the social structure of village life in Thailand ."

The religion of the everyday is clearly more easily tapped in, say,Thailand, than its equivalent in the highly differentiated societies ofthe industrial sector of the world. Nevertheless, it seems clear that theintellectualist tradition in its Weberian form has tended to precludeattempts to comprehend the modes and categories of thought of the`ordinary' individual in modern societies. Through our interpretationsof Weber we have come to see the modern world as consisting in : first,variants upon the Judaeo-Christian tradition in relatively full-blown,doctrinally-articulated form ; second, those who have become imbuedwith an orientation of disenchantment to the sociocultural tradition .The fact that this dichotomized situation does little justice to the nuancesof Weber's sociology of religion is in the present context not particularlyimportant . In recent years, sociologists of religion have both widenedand deepened their analyses. The deepening has consisted in an increasingconcern with the protoreligious substratum-superstition, astrologicalinterests, and so on. The broadening has consisted in an expansion ofthe range of religious (and irreligious) movements and cults which arebeing studied .B' But such developments do not fully meet the require-ments suggested in this essay . What is now needed is an approach whichprobes `below' the surfaces of specific beliefs in `religious entities' andwhich yet at the same time relates these general beliefs about the operationof the world to the more culturally concrete beliefs, values and symbolswhich sociologists have traditionally studied .

Such a programme would raise profoundly the relationship betweensocial-scientific beliefs and the modes of cognition of the `real world' .It would undercut the position of those who have considered religiouscommitments as in some way distant from social-scientific beliefs. Itwould demand a systematic comparison of the ways in which socialscientists think and the ways in which `ordinary mortals' think. Such adevelopment would be in line with the broad conception of a criticalsociology which I briefly sketched earlier in this essay. Moreover, theworries which some socioligists have expressed about sociology beinginsufficiently transcending ought to be significantly attenuated by such aprogramme of analysis . For the prospectus which I have indicated here

124 THE SOCIOLOGY OF RELIGION :

would allow both for analysis of religion and yet would not be guilty ofexplaining away the beliefs and cognitive modes of either the religious orthe irreligious. Science would be preserved-but in a more open andflexible manner than has typically been the case in the last few decades .

NOTES AND REFERENCES

t . Roland Robertson, `Sociologists and Secularization', Sociology (forthcoming, 197o) .2 . Mary Douglas, Natural Symbols, London, 1 970-3. Robert N. Bellah, Beyond Belief, New York, 1970, esp . part 3 .4. See Mary Douglas, Purity and Danger, London, 1966 .5. Independently David Martin has characterized Douglas' approach in Natural Symbols

as being sociotheological . See his review of this book in British Journal of Sociology,vol. XXI, No . 3, September, 1970 .

6 . Compare Peter L. Berger, The Precarious Vision, Garden City, N.Y., 1861, with Berger,A Rumor of Angels, Garden City, N.Y. 1969 .

7 . In this connection see the debate surrounding Bellah's Position : Journal for the Scien-tific Study of Religion, vol. 9, No . 2, Summer, 1970.

8. See Devra Lee Davis, 'Religionswissenschaft and Religionssoziologie ; Two ideal types ofapproaches to the study of religion', Department of Sociology, University of Chicago,1967, (mimeo) .

g. See Robertson, `Sociologists and Secularization', op . cit . As I have noted there arereactions against this coming from both within and without the sociology of religion .For the external situation consider the implications of Neil J . Smelser and James A .Davis (eds.), Sociology, Englewood Cliffs, N. J ., i 969, pp . 101-1og . (This is a report of theAmerican Behavioral and Social Sciences Survey . The sociology of religion is charac-terized as a `soft' field.) On the theological side there is also a reaction against 'soft-ness' . See the arguments for conceiving of theology as a definite science in ThomasF. Torrance, God and Rationality, Oxford, 1970 .

to. Ian G . Barbour, Issues in Science and Religion, London, 1966, p . 124. Cf. John Dillen-berger, Protestant Thought and Natural Science, Garden City, N.Y., tg6o ; Harold K .Schilling, Science and Religion, London, New York, 1963 .

11 . cf. Roland Robertson, The Sociological Interpretation of Religion, Oxford, 1970, esp .ch . 7 . For a brilliant philosophical discussion of scientific procedures which `lets in'metaphysical and ultimate factors, see Romano Harre, The Principles of Scientific .Thinking, London, 1970.

12. The paragraphs which follow-up this statement are based upon Roland Robertson,`The Sociocultural Implications of sociology : A Reconnaissance', in A . H. Hansonet al. (eds.), Imagination and Precision in Political Analysis ; Essays in Memory of J. PeterNettl, London, 1971 .

13 . Cf. idem and Roland Robertson, review of Amitai Etzioni, The Active Society, NewYork, 1968, in British Journal of Sociology, vol. XX, No. 1, March, 1969, pp . 92-95-

14. Herminio Martins, `Sociology of Knowledge and Sociology of Science', 1970 (mimeo) .An abbreviated version of this appears in Hanson et al., op . cit ., entitled `The Kuhnian"Revolution" and its Implications for Sociology' .

15 . See Jan Loubser, `Calvinism, Equality and Inclusion : The Case of AfrikaanerCalvinism', in S .N. Eisenstadt (ed.), The Protestant Ethic and Modernization, Boston,1968, PP. 367-383 .

16 . See Robin Horton, `Neo-Tylorianisn : sound sense or sinister prejudice?' Man, N .S .,vol . 3, No. 4, December, 1968, pp. 625-634-

17 . Talcott Parsons, The Structure of Social Action, Glencoe, 1937 .t8 . L . Levy-Bruhl, How Natives Think, (Trans . L . Clare), London, 1926 .19 . See in particular Claude Levi-Strauss, The Savage Mind, London, 1965 .20 . Cf. Clifford Geertz, `Man and Ideas : The Cerebral Savage', Encounter, vol . XXVIII,

April, 1967, pp. 25-32-21 . Cf. Robin Horton, `African Traditional Thought and Western Science', in Bryan

R. Wilson (ed.), Rationality, Oxford, 1970, PP. 131-171 .

PROBLEMS AND DESIDERATA 125

22. See in particular J. D . Y. Peel, 'Understanding Alien Belief Systems', British Journalof Sociology, vol . XX, No. i, March, 1969, pp. 69-84. See also, more generally,Wilson (ed .), Rationality, op . cit.

23 . See Horton, 'African Traditional Thought and Western Science', op . cit . See alsoS. B. Barnes, 'Paradigms-Scientific and Social', Man N.S ., vol. 4, No. i, March,1969, pp. 94-102 .

24. Notably Rodney Stark and Charles Y . Glock, American Piety, Berkeley, 1968.25. One of the outstanding examples is Talcott Parsons, 'Christianity and Modern

Industrial Society', in E . Tiryakian (ed.), Sociological Theory, Values and SocioculturalChange, New York, 1 963, PP . 33-70 .

26. Roland Robertson and Colin Campbell, 'Religion in Britain ; the Need for NewResearch Strategies', paper given to University Association for the Sociology ofReligion, March, 1971 (mimeo) .

27. See inter alia, Parsons, 'Christianity and Modern Industrial Society', op . cit . ; RobertN. Bellah, 'Epilogue', in Bellah (ed .), Religion and Progress in Modem Asia, New York,1965, pp. 168-229 ; J . Milton Yinger, The Scientific Study of Religion, New York,1970; Gerhard Lenski, The Religious Factor, Garden City, N.Y ., 1961 .

28. Clifford Geertz, 'Religion as a Cultural System', in Michael Banton (ed .), Anthro-pological Approaches to the Study of Religion, London, 1966, pp. 1-46 .

29 . Bellah, Beyond Belief, op. cit .30. Talcott Parsons, 'Commentary' (on Geertz, op . cit .), in Donald R . Cutler, (ed .),

The Religious Situation i968, Boston, 1968, pp . 688-694.31 . Talcott Parsons, Societies : Evolutionary and Comparative Perspectives, Englewood Cliffs,

1966.32. Robertson, The Sociological Interpretation of Religion, op. cit ., esp. chs . 3, 7 and 8 .33 . Helmet Schelsky, 'Can Continual Questioning be Institutionalized?' in Norman

Birnbaum and Gertrud Lenzer (eds .), Sociology and Religion, Englewood Cliffs, 1969,PP- 418 -423-

34- This, it should be emphasized, is not an expression of adherence to those who wish todefine religion very broadly-i.e. in an inclusive manner . It is simply to argue thatmany of the questions raised by sociological analysis of religion go well beyond theboundaries of what I would claim to be the domain of religiosity. Cf. Robertson,The Sociological Interpretation of Religion, op. cit ., esp . ch. 3.

35. Thomas Luckmann (The Invisible Religion, New York, 1967) has argued very stronglyagainst focus upon 'official' religion-but I would wish to dissociate myself from theprecise terms in which he thereby includes as being 'religious' all forms of overcomingof man's biological givenness. Cf. Robertson, The Sociological Interpretation of Religion,op. cit ., passim. See also Colin Campbell and Robert Coles, 'Religiosity, ReligiousAffiliation and Religious Belief', Department of Sociology, University of York,(mimeo) .

36. See, for example, J . Milton Yinger, 'Pluralism, Religion and Secularism', Journalfor the Scientific Study of Religion, vol . IV, Spring, 1967, pp . 17-28 .

37. See in particular Harold Garfinkel, Studies in Ethnomethodology, Englewood Cliffs, 1967 .See also Harold Garfinkel and Harvey Sachs, 'On Formal Structures of PracticalAction', in John C. McKinney and Edward A. Tiryakian (eds .), Theoretical Sociology,New York, 1970, PP- 337-366 .

38. A prime example is Alvin W. Gouldner, The Coming Crisis of Western Sociology, NewYork, 1970, PP- 39-395-

39. Focus on 'everydayness' has its limitations, as will be seen . See Robert W. Friedrichs,A Sociology of Sociology, New York, I97o, esp . ch . 12 .

40 . See, inter alia, Alfred Schutz, The Phenomenology of the Social World (trans . GeorgeWalsh and Frederich Lehnert), Chicago, 1967 . See also Peter Berger and ThomasLuckmann, The Social Construction of Reality, Garden City, N.Y., 1966 .

41. Burkart Holzner, Reality Construction in Society, Cambridge, Mass., x968. Seethe characterization of Berger and Luckmann as 'pan-cognitivists' in Martins,op. cit.

42. See Garfinkel and Sachs, op . cit .

126

43-

44-45-46.47.48-49-

50.51 .

Recall, for example, the differences between Simmel and Durkheim on this issue :Emile Durkheim et al., Essays on Sociology and Philosophy, New York, i g6o, pp . 354-375Loubser, op . cit.Peter Berger, The Sacred Canopy, Garden City, N.Y., p. 125.Max Weber, The Sociology of Religion (trans. Ephraim Fischoff), Boston, 1963, p. 206 .George C. Homans, The Human Group, New York, ig6o, pp. 313-333-Neil J. Smelser, Theory of Collective Behavior, London, 1962 .See, inter alia, Basil Bernstein, `A Socio-Linguistic Approach to Social Learning', inJulius Gould (ed .), Penguin Survey of the Social Sciences, London, 1965, pp . 144--168 .Of. the uses to which Bernstein's distinction and others associated with it are put inDouglas, Natural Symbols, op . cit . It would seem that the kinds of approach suggestedhere could be fruitfully related to the work of Piaget-in particular the latter'sdistinctions between thought processes of induction, transduction and deduction .Mass Observation, Puzzled People, London, 1948 .See, for example, E . R . Leach (ed) ., Dialectic in Practical Religion, Cambridge, England,1968 . Such arguments have in recent years been advanced in particular reference toBuddhist societies. Cf. Melford E. Spiro, Burmese Supernaturalism, Englewood Cliffs,196 7-

52 . For a brilliant analysis of the ambiguities in Weber's use of the term 'theodicy', seeGananath Obeyesekere, 'Theodicy, Sin and Salvation in a Sociology of Buddhism',in Leach (ed.), op. cit., pp. 7-40 .

53 . Berger, The Sacred Canopy, op. cit., pp. 67-68 .54. Ibid, p . 68. See M. M. Ames, 'Magical-animism and Buddhism : A Structural

Analysis of the Sinhalese Religious System', Journal of Asian Studies, vol . 23, 1964,pp. 21-52 . See also Spiro, Burmese Supernaturalism, op. cit .

55- S. J . Tambiah, `The Ideology of Merit and the Social Correlates of Buddhism in aThai Village', in Leach (ed.), op . cit., PP. 41-121 . See also Tambiah, `Animals areGood to Think and Good to Prohibit', Ethnology, vol . XIII, No. ,, January, 1969,PP- 423-459-

56. For this and associated points, see Robertson and Campbell, op. cit .

THE SOCIOLOGY OF RELIGION