Zambezia (1981), IX (i).ESSAY REVIEWTHE STUDY OF RELIGION IN CENTRAL AFRICADURING THE LAST decade and in the social sciences the major publication outputwith respect to Central Africa has been in the areas of political science and history.From the anthropologists and sociologists there has been a declining output sincethe halcyon days when the Rhodes-Livingstone Institute initiated and brought tofruition an extensive and co-ordinated research programme on a regional basis forwhich there are few parallels. There has been, however, a continued output in twoareas, those of urbanization and religion. The six volumes reviewed here1 are allconcerned with the latter, and demonstrate the continued vitality of academicinterest in religion, which has such myriad inter-connections and ramifications forthe other institutions of society.While, however, they all have a common focus on religion, they represent awide spectrum of approach and subject matter. Like the other phenomena of humanculture and society, religion can be studied descriptively or analytically; it can beexamined in its past or contemporary forms; and it can be analysed as either adependent or independent variable, its forms being regarded either intrinsically asmanifestations of Ritual Man or as the epiphenomenal or surrogate activities ofEconomic or Political Man. In various degrees the authors represented in these sixvolumes reflect the entire spectrum of approaches suggested by these dichotomies.Gelfand and (largely) the writers of the van Binsbergen and Buijtenhuis volume areconcerned with contemporary religious data but while Gelfand is content withstraight ritual description the others are involved in interpretation, largely along thetheme of cultural adaptation. Fry, Bhebe and the authors of the Schoffeleers andWerbner publications are all concerned with historical data, but their analyticalperspectives are significantly different Fry sees his material primarily as areligious reflex to political contexts; the others are more interested in the way beliefsystems interrelate with a variety of ecological and societal requirements. All thisvariety of approach is reflective of both the diversity of the authors' training andinterest and also the multifaceted nature and import of religion in society. Takentogether, these six volumes constitute a valuable addition to the body of scholarshipon religion in Central Africa but their impact is not easily synthesized, nor do theyconsistently give us insights which constitute analytic advances on the state of theart.Professor Gelfand, in his The Spiritual Beliefs of the Shona, takes a straight-forward ethnographer's route. He is concerned only peripherally with analysis andseeks rather to record the results of field-work on Shona history and ritual carried1 N. Bhebe, Christianity and Traditional Religion In Western Zimbabwe 1859-1923 (London,Longman, 1979), 190 pp., £7.95; P. Fry, Spirits of Protest: Spirit Mediums and the Articulation ofConsensus among the Zezuru of Southern Rhodesia (Zimbabwe) (Cambridge, Cambridge Univ.Press, 1976), 145 pp., £4.50; M. Gelfand, The Spiritual Beliefs of the Shona (Gwelo, Mambo Press,1977), 216 pp., Z$5,4O; J. M. Schoffeleers (ed), Guardians of the Land, Essays on Central AfricanTerritorial Cults (Gwelo, Mambo Press, Zambeziana Series 5, 1978), 315 pp., Z$9,00; W. M. J. vanBinsbergen and R. Buijtenhuis (eds), Religious Innovation in Modern African Society (Leiden, Afrika-Studiecentrum, African Perspectives Series, 1976), 148 pp., Dfl.13; R. P. Werbner (ed.), RegionalCults (London, Academic Press, A.S.A. Monograph 16, 1977), 257 pp., £7.20.6768 ESSAY REVIEW*out between 1967 and 1976. He eschews'hypotheses' and sticks to description, inline with his previous work (p.5). Like some of his earlier published volumes, thisone is deceptively titled. The content relates not to 'the spiritual beliefs' of the *Shona in general, but rather to the ritual practice and belief of four Shona sub-groupings, the Manyika, the Vaunyama, the Hwesa and the Budjga. As such, the *book provides a useful complement to Gelfand's previous researches among theZezuru, the Korekore and the Karanga. For each of the four groups covered in thisvolume a similar order of presentation is utilized: a brief capsule history is followedby a discussion of the 'clan' spirits and their mediums; then follow sections ondiviners, family spirits, death rituals, marriage procedures, hostile and alien spirits >ŁŁand the rituals associated with them. In one respect the book presents the readerwith an advantage over previous publications in the series. By having the ritual *practice and belief of four different groups juxtaposed in the one volume both theoverall similarities and the detailed differences are made more ostensible, giving a ^clearer picture of the unity and diversity to be found in traditional Shona religion.While this book does not provide the wealth of detail on any one ritual centre that,for instance, Gelfand's earlier book, Shona Ritual,1 does, it breaks new ground ingiving us descriptive material on the ritual detail of the Manyika, the Vaunyamaand the Hwesa, groups which have not been extensively covered in this respect by cthe earlier literature.In his preface Gelfand states that this is the concluding volume on his -researches into the religion of the Shona, commenced amongst the VaZezurutwenty-five years earlier. During that period this indefatigable scholar, in additionto producing a prodigious literature in his own field of medicine and significantpublications in the areas of law and history, has faithfully recorded a wealth of *ethnographic and ritual detail from assiduous fieldwork amongst the Korekore,Zezuru, Karanga, Manyika and other Shona groups. His writings have frequently **been criticized by social anthropologists and historians on a number of grounds,sometimes with justification. Gelfand's lack of formal training in the socialsciences often reveals itself in a neglect of social context and as a result the *questions evoking analytically important detail are not asked. This is unfortunate,for, to cite one instance, had the social location of the actors in the ritual dramas hedescribes been more fully analysed the sociological significance of his works wouldhave been immeasurably enhanced. But this is to cavil at the enormously valuable *""ethnographic gift to our scholarship provided by a professional physician who hasgenerously given of his 'spare' time over much of a lifetime to record the detail wenow possess in the corpus of his writings. I sometimes speculate that, had Gelfand vreceived an extended anthropological training at the beginning of his career, Shonaethnography could have had its counterpart to Schapera's lifetime of scholarshipon the Tswana. Of course, had this been so, Gelfand's other notable professionalaccomplishments would have been attenuated and, not least, thousands of v-Zimbabweans (of whom the writer is one) might not have benefited as they did fromthe healing touch of his physicianship. But this is all hypothetical, and students of Łthe Shona must now be content with an ethnographic record which, particularly inrespect of its representation of local and regional variation, they can ignore only atthe peril of their scholarship.Dr Bhebe's book, Christianity and Traditional Religion in Western Zimbabwe2M. Gelfand, Shona Ritual (Cape Town, Juta, 1959).M. W. MURPHREE691859-1923, is a beautifully crafted volume, tightly structured and presented inattractive format Written with economy of focus and clarity of expression, thebook is arguably the most readable scholarly book on Zimbabwean history that Ihave encountered. At the same time, its scholarship is impressive, the text beingcarefully documented by Bhebe's extensive archival and field research. The book,in effect, sets itself two objectives. The first is to trace the establishment of theChristian Church in Western Zimbabwe during the period covered. The second,and more important objective, is to identify and analyse this establishment in itssocio-cultural context, leading to a greater understanding of the religious dynamicswhich have shaped subsequent cultural and socio-political developments in thearea. The book incorporates sound historical scholarship with a pervasive socio-logical perspective which makes a significant contribution to social scienceliterature on Zimbabwe.Bhebe's starting point contains two fundamental insights, one explicit and oneimplied. The first is that African religions are to be seen as dynamic, mutatingsystems in constant and continuing change, not only during but also before theonset of the colonial phase. The second is that these religious systems are to be seenas existing in dynamic and reciprocal interaction with the other structures ofsociety, a relationship which is both reactive and generative. These two insights areof critical importance. The first rescues Bhebe's analysis from the exclusivelystatic view of African traditional religions so often implied by ethnographicliterature of the type exemplified by Gelfand's work mentioned earlier. The secondtakes Bhebe's analysis beyond the simplistic and sterile 'reaction to conquest'paradigm which is implicit in most historical and anthropological treatments of thesubject in Zimbabwe, and which characterizes traditional religion, often withRousseauian romanticism, as being a particularistic mode bereft of context,gallantly fighting a rearguard action in the face of a rapacious onslaught by theideological predators of a universalistic religion sustained in every respect by thecolonial situation. Such a paradigm ignores the complexities of motivation andcontext shaping the Christian expansion in Zimbabwe (which, as Bhebe shows,rapidly became an African as well as a missionary enterprise) and is a gross insultto the vitality of the indigenous religious traditions which (again as Bhebe pointsout) have amply demonstrated their ability to respond not only reactively but alsocreatively to changing contexts, both before and subsequent to the establishment ofcolonial domination.Bhebe's treatment is organized chronologically into five sections: (a) Ndebele-land: society and religion before 1859; (b) The Ndebele and the first missionaries,1859-68; (c) Christianity and Western Zimbabwe under Lobengula, 1870-89;(d) Religious institutions and the fail of the Ndebele kingdom, 1888-97; (e)Missions and the traditional societies, 1897-1923. The book ends with a chapteron Christianity and education and a concluding analytic overview which seeks toplace Bhebe's material within the context of other studies of religious change inAfrica. In this concluding section Bhebe relies heavily on two typologies, RaymondOliver's typology of African societies and Humphrey Fisher's typology of Africanreligious transition. The integration of this data in terms of these typologies is usefuland effective but not extended. I suspect that these typologies (and indeed theothers he mentions, developed by Horton and Murphree) are not adequate for thescope and variety of materials Bhebe produces. He is dealing with a multiplex set ofdata manifested in a variety of historical, social and cultural dimensions, all70 ESSAY REVIEWsuperimposed on each other. I am reminded of Horowitz's observation that anyanalysis of change must grapple with the problems of consistency (different rates ofchange), sequence (order of change) and congruence (fit with other aspects ofsocial structure), Bhebe in his concluding analysis deals largely with the issue ofsequence; the problems of consistency and congruence are touched on elsewherethroughout the text but not effectively treated in his analytic summation. Onesenses that the material is there, and that this is the proper subject for Bhebe's nextbook. In this volume he has approached his subject, as one would expect of anhistorian, inductively. The stage is now set for this scholar to take a more deductiveapproach, setting up his own typology and analytic framework on the basis of thiswork for a conceptual approach which could well lead to a considerable advance inour understanding of the religious dynamics of Africa, past and presentThe entire volume is replete with valuable insights too numerous to mentionhere. I would, however, particularly draw the attention of the reader to Bhebe'streatment of pre-Christian religious synthesis and innovation (Preface and Chapter1), Lobengula's intercalary role (pp. 41-66), the role of Black converts (pp. 34-36,82-83, 99-100, 129-132) and the heterodoxy and heteropraxy of the Christiancommunity, which developed during the period covered. If the book has a weaknessin the range of its analysis it lies in Bhebe's failure to explore extensively the socio-political location of the various missions and the significance of this factor for theirrespective policies and impact. But this is perhaps an issue more important for theperiod succeeding Bhebe's cut-off point of 1923 and does not significantly detractfrom a study which is otherwise thorough, concise and lucid.Dr Fry's Spirits of Protest: Spirit Mediums and the Articulation of Consensusamong the Zezuru of Southern Rhodesia (Zimbabwe), stems from field-workdone by the author in the Chiota Tribal Trust Land for his doctoral thesis in themid-1960s. It is a slim volume (123 pages of text) based on material collectedduring a period when Fry was repeatedly frustrated by the inhibitions imposed bythe mounting political conflict of the time. The district administration of theRhodesian Front Government was hostile and restrictive and the indigenoussubjects of Fry's research were suspicious of motive and cautious in response.Given these conditions, Fry was forced to abandon the accepted anthropologicalapproach of establishing a thorough sociographic context for his material throughan extended social survey, and his primary material consists largely of a small set ofextended case-studies of individuals interviewed and observed over time, coupledwith situational analysis. As a result we are given no quantitative dimensions to thesituational contexts Fry analyses, and no indices by which to evaluate Fry'sassertion of'the decline of Christianity in Chiota and the rise of spirit-mediumship'(p. 4) at this point in time. Given the circumstances, the lack of socio-demographicsurvey material is understandable; what is inexplicable is the absence of anystatistics on religious-group attendance and membership which must surely havebeen available. Fry's ethnographic homework is similarly scanty. Dr Bourdillon, ina review of this book in Africa,3 has commented that it 'fills a notable gap in theethnography of the Zezuru'. This is over-generous. While there is nothingexceptionable about Fry's short background ethnographic sections (pp. 5-29)there is nothing new about them either, nothing that has not already appeared in theworks of Bemardi, Broderick, Chavunduka, Garbett, Gelfand and Seed.}M. Bourdiilon, Africa (1978), XLVIII, 410-11.M. W. MURPHREE71On this narrow data base Fry erects an ambitious analytic superstructure, aninverted pyramid of generalizations teetering on its head. He sets himself the task ofverifying three hypotheses: (a) that the authority of spirit mediums rests primarilyon their charismatic gifts and their ability to perceive and articulate consensusrather than on the ascribed status of the spirits they purport to represent; (b) that asharp contrast is to be perceived between the achievement basis of Zezururaediumship and the alleged hierocratic basis of authority among Korekoremediums; and (c) that the revival of spirit mediumship among the Zezuru during theperiod of Fry's fieldwork represents an oscillation away from Christianity andtowards traditional religion corresponding to a parallel politico-cultural tendencyto reject Western values and revive traditional ones. Fry, within the confines of thisdata, succeeds brilliantly in the first task, fails completely in the second, and gives,in the third, an argument that can only yield the verdict 'ueproven'.Fry's failure to sustain his argument in his comparison between Zezuru andKorekore mediumship is the result of two critical analytic weaknesses, thecomparison of two different types of data on the assumption that they are the same,and the tendency to construct facile dichotomies for the purposes of generalization.Garbett's material on the Korekore, on which Fry relies, is concerned with aparticularly stable instance of Korekore mediumship, existing in a differenthistorico-political context. One cannot take this data, therefore, contrast it withFry's and conclude, ergo, that Korekore and Zezuru mediumship are differentThey are in certain respects; but the evidence for this conclusion is not found inFry's book. Indeed the record, as shown in Garbett's chapter in the Werbnervolume reviewed below, indicates that many of the entrepreneurial aspects ofmediumship found by Fry among the Zezuru are also to be found among theKorekore. The same analytic weaknesses vitiate Fry's depiction of a decline inChristianity and a corresponding resurgence of traditional religion in Chiota. It isclear that Fry's case-studies represent an important strand in the warp and woof ofthe dynamic religious system in Chiota, but he does not give us the data required forrigid analytic contrast and comparison. Fry sees the spirit mediums as being'unequivocally opposed structurally to Christianity' (p. 120) on a line delineatedby opposed views on African nationalism. This is not convincing in the light ofstrong support for the nationalistic cause by many Christians. Fry's rigiddichotomization here is as misleading as when he says that the authority of thechiefs is coercive and that of the mediums is consensual (p. 9). Here a tendency iselevated to a rale; the facts are that many Shoea chiefs opt for a consensual basis fortheir authority, as Weinrich has shown.4In one respect Fry's analysis is convincing. This is his demonstration of theentrepreneurial dimension of mediumship which effectively discredits the con-ventional stereotype, often naively accepted by Shona ethnographers, of a rigid andgiven hierarchy of spirits and mediums. The spirit medium role has considerable'room for manoeuvre', to use Lucy Mair's phrase, and Fry skilfully demonstrates itsfunction of consensual summation. It is a pity that field-work conditions did notpermit the author to range more widely in pursuit of the analysis of this role andfunction. A book with this kind of focus would have been a far better one.I turn now to the three volumes edited, respectively, by van Binsbergen andBuijtenhuis, Werbner, and Schoffeleers. The van Binsbergen and Buijtenhuis4 A, K. H. Weinrich, Chiefs and Councils in Rhodesia (London, Heinemann, 1971).72ESSAY REVIEWvolume is of the least interest since it has little internal cohesion and consists of aseries of working papers by authors who have published more scholarly workelsewhere. The idea behind the publication was a good one: to re-examine standardanalyses of African religious experience in the light of recent socio-political events,with special reference to the importance attached to the colonial situation andemergent nationalism as being the causal nexus of these analyses. Unfortunatelythe contributions fail to contribute significantly to this goal. Daneel gives arambling pot-pourri of field notes with little analytic progression, and HaroldTurner ties himself up in knots trying, unsuccessfully, to improve on Sundkler'stypology of independence. The one useful chapter is van Binsbergen's case study ofthe Lumpa church in Zambia which stands on its own as an insightful analysis ofone variety of religious independency within the context of a newly independentAfrican state. But taking the volume as a whole, one can only agree with the editorsthemselves when they admit that it 'is fair representation of the lack of consensusand the state of flux currently dominating this field of inquiry' (p. 7).The last two volumes, edited, respectively, by Schoffeleers and Werbner, are adifferent matter. Both are seminal, and while containing admittedly exploratoryanalyses are informed by an analytic rigour and correspondence of focus thatprovide a cohesiveness lacking in the van Binsbergen and Buijtenhuis volume.Chronology is an important consideration in the evaluation of these two books. Inthe Werbner volume, published in 1977, the editor states, 'These essays begin anew phase in the anthropological study of regional cults. It is the first time that thecults are treated as a major focus of comparative analysis' (p. xxxvi). Strictlyspeaking this is incorrect, for although the Schoffeleers volume was only publishedin 1979, it represents papers produced on the theme of territorial cults in 1972, withan introduction written by the editor in 1974. Although the distinction betweenterritorial and regional cults is of significance, the two volumes present cognatematerial. Both chronologically and conceptually the two publications represent aprogression and I shall take up the Schoffeleers volume firstThe book is divided into three sections containing articles respectively onZambia, Malawi and Zimbabwe, a total of ten essays. Some of the essays are oflimited impact, bearing the marks of hastily constructed papers prepared forconference presentation. But a number of others are workmanlike analyses whichexpand on the field of work of their authors and advance our knowledge of thevariety and complexity of the territorial cults in this part of Africa, a study whichhas not received systematic attention heretofore. In particular I would single out inthis respect the chapters by van Biesbergen on Zambia, Schoffeleers on theChisumphi and Mbona cults in Malawi, Bourdillon on the Korekore and Tavara,and Rennie on the Musikavanhu cult of Zimbabwe and Mozambique, Schoffeleersalso presents, together with Mwanza, an interesting piece on the Mwari shrines ofsouth-western Zimbabwe which is significant in the light of Werbner's treatment ofthe same subject in the book reviewed below.But the real jewel in this book Is Schoffeleers's introductory chapter, taking up46 pages of text and providing an overview of the subject. SchofFeleers delnes theterritorial cult as one 'whose constituency is a territorial group identified bycommon occupation of a particular land area, so that membership of the cult is...a consequence of residence and not kinship or ethnic designation' (p. 1). Territorialcults are therefore a specific mode of religious organization to be distinguishedfrom, for instance, cults of affliction or the ancestral cults. Unlike the latter theyM. W. MURPHREE73have a specific geographic base and, concomitantly, are inclusive in nature, all theinhabitants of the territory being their constituency. This spatially determined andinclusive character also implies that they closely parallel the institutions of secularauthority, although the correspondence does not always hold when the cultsexpand to become what SchofFeleers calls 'federative' cults. 'One can... state thatterritorial cults are by nature political since they are the religious representatioii ofwhat are basically and primarily territorial and political groups' (p. 6).From this definitional base Schoffeleers proceeds to the difficult task ofsynthesizing the material at Ms disposal into a coherent whole of conceptual valuefor further analysis. He succeeds remarkably well. Modestly, Schoffeleerscomments in Ms conclusion that 'Although more useful ways of classifying may beproposed in the future, the one that has been proposed allows us at least to create abeginning of order in an otherwise bewildering variety of cults' (p. 42). Schoffeleers'sessay does provide this 'beginning of order5 in an analysis that is innovative,concise and convincing.In his first, and perhaps most original, section Schoffeleers argues that theterritorial cults are, in essence, profoundly ecological in nature. They are theinstitutionalized manifestations of the earth philosophies of the region, thearticulation of communal and ecological concerns. They have a prescriptivefunction with respect to food production and distribution, the protection of naturalresources and the control of human settlement and migration. Such is their impacton the ecological system that Schoffeleers borrows Rappaporfs phrase and speaksof'a ritually directed eco-system' (p. 3). The coalescence of religious, communaland ecological concerns stems from a logic positing a causal connection betweenthe physical and moral orders. Or, as Schoffeleers puts it, 'management of naturedepends on the correct management and control of society' (p. 5),SchofFeleers then proceeds to an organizational typology of the territorial cults,a section which I suspect is more likely to be subject to revision than the rest of thischapter. He then turns to an historical overview of the territorial cults in a sectioncontaining seminal insights on a number of issues too numerous to mention here.He has important things to say regarding the dynamic and varied posture of theterritorial cults vis-a-vis other competing tendencies, and foresees their demiseconcomitantly with the erosion of their ecological and communal base as a result ofthe introduction of bureaucratic government with its secularizing tendencies, landalienation and wage earning.One issue of comparative importance escapes significant treatment inSchoffeleers's analytic net. He does not effectively contrast the inclusive principleof the territorial cults with the exclusive principle of other religious modes,including Christianity, or the implications of this contrast for innovation andadaptability at individual and group levels. These facts are critical in the process of'modernization', and both inclusiveness and exciusiveness would appear to havetheir evolutionary advantages and disadvantages in this respect I find it curious, infact, that Schoffeleers is so pessimistic about the future of the territorial cults. In thelight of their manifest adaptability and vitality in the past, and the continuingcentrality of ecological concerns, one wonders if they will not, Phoenix-like,emerge from the ashes of the present into a transmogrified future bearing anessential continuity with the past, a continuing component of the rich mosaic of thereligion of Central Africa.The Werbner volume is a collection of eight papers presented at a 197674 ESSAY REVIEWconference on regional cults sponsored by the Association of Social Anthropolo-gists of the Commonwealth, together with an introduction by the editor. Six of theeight essays deal with Central African materials, the other two drawing on datafrom Western Morocco and South Sinai. All are conceptually sophisticated andthe volume exhibits the professionalism we have come to expect in the ASAMonograph Series.Curiously, nowhere in the book can I find a definition of a regional cult. Many ofthe authors are dealing effectively with what Schoffeleers has defined as 'territorialcults', but in some instances the cults dealt with do not conform to his criterion ofinclusive territoriality. Instead the focus of this book is on 'cults of the middlerangeŠmore far-reaching than any parochial cult of the little community, yet lessinclusive in belief and membership than a world religion in its most universal form'(p. ix). Contrasted with the Schoffeleers volume, this expansion in scope providesthis book with the comparative data on inclusive/exclusive principles which arelacking in the other work. At the same time this leads to a certain diffuseness offocus and the cohesiveness of the volume is only rescued by the synthesis providedby Werbner in his introduction.Werbner identifies as the central theoretical concern of the book the issue of'correspondence', the widely accepted theory in an anthropological traditionstemming from Robertson Smith via Durkheim that postulates a correspondence'between symbolic representations of the social world and patterns of conduct'(p. xviii). Or, as another contributor, Eickelman, puts it, this is the presumption of'an elegant one-to-one correlation between ideology and social action' (p. 4).Werbner, Eickelman and, implicitly, most of their co-authors suggest that theuncritical acceptance of this proposition leads to an attenuated and inadequateanalysis, disguising the ability of religious systems to generate their ownmoraentums which are not only modified by, but also modify, the direction of socialactionŠa perspective paralleling my earlier comment on Bhebe's insight on thereactive and generative aspects of religious institutions. They propose analternative conceptual stance, at one point presented under the rubric 'interaction':'The notion of correspondence masks the two-way interaction which occursbetween symbolic conceptions of the social order and patterns of social action. Theidea of interaction, as opposed to correspondence, necessarily implies a lack of fitbetween the two analytic levels at any given moment' (p. 4). A concomitant stanceto the 'interaction' perspective taken by the volume is the abjuration of rigid,opposed dichotomies. Commenting on Victor Turner's parallel dichotomization,Werbner states, 'Like other ideal typologies, however, it tends to represent asmutually exclusive alternatives which are, in fact, aspects which combine in asurprising variety of ways within a range of actual cases' (p. xiii).Two methodological approaches inform the authors' approach to their subject.They are concerned, first, to analyse their data over time; change, in both itsreactive and generative dimensions, is an essential component in their data.Secondly, they are concerned to ensure that their analysis arises inductively, withan emphasis on the perceptions and perspectives of the actors themselves. None ofthis is new to anthropological analysis, but the combined conceptual andmethodological approach taken in the book is refreshing and stimulating, marking,it is hoped, a new era in the anthropological study of religion. The book fails in anyattempt to construct a model yielding high-level generalizations. Yet it abounds increative analytic insights, and in its respective chapters presents a series of valuableM. W. MURPHREE75caveats against simplistic analyses of the interactive process between religion andthe other institutions of society based on older models of conventional anthropo-logical wisdom. For this alone it is more than worth the purchase price.In retrospect, my reading of these six volumes confirms the happy suspicion thatthe study of religion in Central Africa is alive and well. I find the Bhebe,Schoffeleers and Werbner volumes of far greater value than the others. Werbner isthe most stimulating and provocative book; the Bhebe and Schoffeleers works arethe most convincing. They are, in their respective disciplines,' studies of the middle-range', to paraphrase Merton, studies 'intermediate to the minor workinghypotheses evolved in abundance during the day-to-day routines of research, andthe all-inclusive speculations comprising a master conceptual scheme'.5 Satisfyingin their rounded, cohesive argument, they do not frustrate by wanton generalizationpatently beyond the limits of the evidence. That scholarship in Central Africanreligion has graduated from Merton's 'minor working hypotheses' to mid-rangetheory is a considerable achievement, and it is to this level of analytic enterprisethat work should now be directed during the next decade. The works reviewed inthis essay indicate that the capacity is there. And I would remind those tempted tomake the precipitous leap from 'minor working hypotheses' to grand theory that, asGoldkind has said,6 while any theory may be better than none at all, gooddescription is better than bad theorizing.University of ZimbabweM. W. MURPHREE5R. K, Merton, Social Theory and Social Structure (New York, The Free Press, 1949), 5-6.*V. Goldkind, 'Three studies of Mexican Indians: A methodological critique', AmericanAnthropologist (1979), LXXXI, 894.