H tion by an important anthropologist who is still developing and rethinking

Transcription

1 Review Article1 Order and Rebellion in Tribal Africa. MAX GLUCKMAN. New York: The Free Press of Glencoe (Macmillan), xii, 273 pp., 1 figure, notes, index. $6.00. Reviewed by RONALD COHEN, Northwestern University ERE is a record of the theoretical work deemed most worthy of republica- H tion by an important anthropologist who is still developing and rethinking his own position. Independently of any idiosyncratic value these papers may have as guideposts to Gluckman s thinking, they also reflect his training in, and acceptance of, many of the basic assumptions and concepts used by social anthropologists in Britain over the past 20 years. Certainly, this collection very clearly demonstrates Gluckman s unique development as an independent thinker within this particular sector of anthropology. The presence of both these interrelated and overlapping sets of concepts and theories in one book makes of it an extremely thought-provoking doorway to a number of the most influential epistemological problems in the discipline. The book includes a number of papers written from 1945 to the present. They deal with Gluckman s views on the political and legal life of traditional African societies, as we11 as his widely known views on Malinowski s work, especially as it pertained to Africa. There is also an extended introduction in which the author attempts to explain the development and present status of his own theoretical position within British social anthropology. The first chapter is Gluckman s (1945) review article of Fortes The Dynamics of Clanship Among the Tallensi. In it he recognized the major significance of the Fortes book. By isolating the Tale segmentary lineage system and describing in great detail the processes that set the system in motion, Fortes was in fact producing a causal theory to explain how a very large range of non-centrally controlled societies organized and maintained their social life. Differences in the operation of the lineage system would produce differences in behavior, ecology, and culture, since Breach of the rules precipitates not only material disaster but may so overwhelm a man that his line itself becomes extinct (p. 72). Territorial segmentation as well as ritual life are a function of (p. 76) or in accordance with (p. 73) lineage segmentation. Gluckman added to this by claiming that there was an important underlying force, i.e., there was a cause of the cause. He did this by taking part of Fortes interpretation of lineage process, that of complementary opposition, and giving it greater determinative power than was attributed to it by the author, or, as he says, The structure of Tale society is therefore a product of tensions between corporate units in a system of balanced ties and cleavages (p. 77). The operation of opposing forces that maintain themselves in equilibrium 950

2 [COHEN] Review Article 95 1 is seen in greater detail in chapter two on Succession and Civil War Among the Bemba. Here the author notes that the idea of equilibrium as a consequence of opposing forces goes back to the editors of African Political Systems who presented it in their introduction to that book, and to his own particular contribution to that early collection. Proposing to take the idea further he states that in societies with stagnant techno-economies conflicts take place between persons vying for office, but these conflicts are resolved by changing officeholders without changing the pattern of offices or relationships (pp. 86-7). He then applies this idea brilliantly to discrepancies in the data on Bemba succession, working with the additional hypothesis (assumption?) that each field worker had been given a biased account by opposite branches of the royal family. Underlying the succession is the fact that every three or four generations a branch of the royal line drifts away and loses its traditional rights of succession. This group then asserts its rights to the throne on some other allegedly valid grounds from those constitutional grounds advanced by the other contending party. A fight ensues over the office of the king and the winner takes all by distributing titles to all of his loyal followers. Because segmenting branches of the royal line struggle for the kingship, the state does not break up (p. 104) asit would if each faction under its own royal leader went their own separate ways claiming equal rank, rather than competing for the one ranking office that all recognize as supreme. It is assumed that if the cadet branch that drifted wins the struggle using other than constitutional arguments, it will subsequently always return to the traditional rules rather than incorporating those extra-legal ones used as the basis of its claim, thus changing the structure of the succession, although why this should always be so is unclear. In chapter three on Rituals of Rebellion in South East Africa published first as a Frazer Lecture in 1954, Gluckman set out to demonstrate that many ceremonies whatever [their] ostensible purpose... openly express social tensions (p. 112). There is little doubt that such hostility is indeed an integral part of many ceremonies, and Gluckman has presented convincing data on this point. He interprets these from a social-psychological point of view, although he takes great pains in the introduction to deny that his analysis does anything more than open the way for psychological interpretations (p. 27). Intentions notwithstanding, he does state very clearly that the Nomkubuwana ceremony of Zulu women in which the women act as temporarily lewd viragoes expresses the severe psychical pressure in their social subordination and their transference by marriage to stranger groups (p. 127). The conflict is social, its results are psychological, and the product of these two is cultural, namely the ritual. The method is inductive and logical (i.e. theoretical not empirical) ; working from the ritual, Gluckman relates it logically to the social conflict or tension, and from that point he infers the psychological tensions that provide the other half of the causal engine.2 This same argument seems to apply in a more complex way to rituals of rebellion in political life. Swazi rituals express antagonism. The ceremony states that in virtue of their social position [Swazi] princes and people hate the king, but nevertheless support him (p. 128). They may hate

3 952 American Anthropologist [67, 1965 or love a particular king, but their antagonism to the office is offset by their loyalty to the social order. Therefore they express all of these psychological difficulties, again related to ambivalence over subordination, in the product, namely the ritual. This does not mean that the ritual expresses nothing else, but it does mean that, to Gluckman, certain invariant laws are expressed in it, or any socio-cultural phenomena, and these can be induced by logical analysis, which in this case becomes a variety of semantic analysis. In a short essay on Mau Mau (chapter four) given first as a radio broadcast, and then published in The Listener in 1954, Gluckman carries these ideas further. In order to counter the claim that Mau Mau was an atavistic surge of pre-colonial savagery (an all too common interpretation of African violence) he suggests instead that it was a movement aimed at and stimulated by the overwhelming strength of the Europeans. Fundamental sociological causes are mentioned such as urbanism, land shortage, lack of leadership, and an internecine struggle for power among the Kikuyu themselves. However, in this analysis the sociological features are discussed only in passing, although Gluckman certainly does not belie their importance. Instead he asks why the movement involved obscene oaths and why it became so violent. He answers these questions by citing an earlier paper of Evans-Pritchard which argued that obscenity in African religion is correlated with occasions of heavy work or hardship, and not cults of rebellion. Contrarily, Mau Mau was a product of the new situation and was in Gluckman s words a movement of despair (p. 145) using a magic of despair that led in turn to a nihilistic doctrine of killing and being killed as the important goal of membership. Mau Mau obscenity stemmed therefore not from the African past but from a regression to instinctual desires which we know are universal.., [The] oaths use, as all secret oaths must, a few limited themes: blood, sex, excreta, bestiality, the threat to murder near kin (p. 145). Since the reader is given very little information on the nature of these universal instincts or of the lack of relation of the form and content of the Mau Mau rituals to traditional Kikuyu activities, it is difficult to evaluate these statements objectively, except to comment that Gluckman seems to be relying in his analysis upon psychological processes as the driving force behind the form of the Mau Mau rituals, as opposed to explanations based upon social, cultural, and historical features. In chapter five, The Village Headman in British Central Africa reprinted from Africa (1949), the author, together with J. C. Mitchell and J. A. Barnes, discusses the problem of the indigenous leader in the modern world. They suggest the hypothesis that the delicacy of the headman s position arises from conflicting principles. First it arises from his position within the village group as such... But the main source of the ambivalence of his position is that he is the personality in whom the domestic-kinship and political systems intersect (p. 151). This descriptive generalization is illustrated in the discussion that follows and in the following chapter on the Zulu chief and the native commissioner. It indicates that Gluckman and his colleagues viewed the role conflicts of the African chief as a result of the presence of different social systems that

4 COHEN] Review Article 953 intersected in his office. Later Gluckman was to tie these systems together by the use of the concept of social field which referred to the one situation in which several systems of action take place. Work on African chiefs and their role conflicts has expanded rapidly since the 1940 s when these authors presented their views. However, except for Doob s recent work on attitudes to authority in Buganda (see chapter eight, The King s Men, edited by L. A. Fallers, Oxford University Press, 1964) almost all of this work infers attitudes rather than studying them. It is claimed that the headman is in two social systems, therefore the attitudes of his followers to him are fundamentally ambivalent (p. 152). This is logically a non-sequitur since there are a number of possibilities in this kind of situation of which this is only one. Followers could have the same attitudes, positive or negative to both systems, and then include the headman in this attitudinal framework; or followers could have different attitudes to both systems, and then include the headman in both of these. Given the fact that these different attitudes were positive and negative, then the followers would be ambivalent towards the headman as suggested. It might also be found that attitudes to the headman are not related to attitudes to each social system of which he is a part, but to something else, e.g. the behavior of a particular headman in office; or different groups of followers might hold differing views about the headman that could or could not change over time. Inferring attitudes from social organization is a good way to come upon one among a number of possible hypotheses, but the study of attitudes requires that every possibility be examined, not merely the one suggested by social anthropological description. Chapter seven is an essay delivered as four radio talks in 1955 that summarize the major findings presented in Gluckman s book The Judicia2 Process Among the Barotse (1955). In it he gives the details of four legal cases and analyzes the arguments used by plaintiffs, defendants, and judges. This leads him to posit the presence of a set of norms referred to in Barotse law, and often in Western law, as the reasonable man who is in native jurisprudence the means by which abstract legal rules are focused on to the varied circumstances of life (p. 179). This is even more important to the Barotse than Western judges, since Barotse do not have modern methods of testing the validity of evidence. Instead they must depend almost entirely on cross-examination. In doing so they try to assess whether the action being described is reasonable, given the situation, the changing times, and the flexibility of words and the rules they represent. The law can thus adapt to Barotse needs and, as Gluckman presents it, this legal system seems a highly adjustable, just, and effective instrument for maintaining social control. Chapters eight through ten deal with Gluckman s attempts to understand and criticize the work of Malinowski. He gives Malinowski ample credit for showing us all that rich results could come from extended field work, although Gluckman also claims that many of the most stimulating generalizations in anthropology have always come from the non-field working scholars. He then criticizes Malinowski on almost every other ground; for his lack of understand-

5 954 American Anthropologist [h7, 1965 ing of what history could do for comparative work, for his naivetc in creating theory, and for his lack of sophistication in the African data that led him to create an oversimplified paradigm of social change in Africa. Upon re-reading this section, I received a distinct impression that it was really an honest attempt by Gluckman to understand why a man who in the passage of time had been proven so wrong on so many counts, had during his lifetime been hailed as such an all-encompassing leader not only of his discipline, but of Western intellectual life as a whole. The part of Gluckman s work that represents the conceptual framework of British social anthropology of the last two decades is best reflected in such basic concepts as balance or equilibrium, process, and change. In Gluckman s papers and those of his colleagues the concept of process is utilized for descriptive generalizations, often called principles, that describe the operation of a set of rules governing the everyday lives of people in society. Thus fission, fusion, segmentation, complementary opposition and other similar concepts are all means by which these rules operate to produce certain consequences in action-if everyone, or a working majority, live in accordance with the rules. As noted above, breaking of the rules has been seen by Fortes to lead to possible extinction (and Gluckman did not criticize this statement in his review article), therefore almost by a Darwinian compulsion the rules become action-or elseand the study of society becomes the study of rules. In this sense it has always seemed to me that social anthropology in England is misnamed, and should in fact be called cultural anthropology, since rules are part of the traditional ideological values of the society, and form only one force (albeit a powerful one) affecting the interrelationship of individuals and groups. The logical development of such an approach is to give up empirical study of social action and study only the rules qzja rules such as Needham has done, hoping that their determinative power will make for worthwhile predictions when applied to behavior. Another difficult and key concept used in this work is that of equilibrium or balance. In discussing Fortes work on Tale society Gluckman says Tie balances cleavage bocause these are inherent aspects of all relations (p. 78) or again in summing up his own approach he speaks of the balances of conflict and collaboration (p. 49) and claims that this approach derives from the work of Bateson, Evans-Pritchard, and Fortes. I must admit that I have always found it hard to accept this concept of balance or equilibrium. A balance requires that one group or set of forces be equal and opposite to another set over time. Let us call the first set A, and the second set B. The concept of equilibrium expressed in Gluckman s work, and those of many other social anthropologists entails the thesis that all of the variable components of A produce an equal and opposing tendency to all those of B over time. This is possible, but there are logically three other outcomes to such a vectorial argument. A could be greater than B, producing a shift towards B, or vice versa, or all of these conditions could obtain at different times under different conditions. Such possibilities are not discussed. Instead the most improbable alternative seems to have been chosen without any explanation. This in turn explains my concern with arguments such as Gluck-

6 COHEN] Review Article 955 man s treatment of the Bemba succession. He simply assumes that the system will not change over time, even though unconstitutional means for gaining the throne have been and will continue to be used. This rather odd assumption of balance or equilibrium underlies much of the difficulty with the concept of social change inherent in British social anthropology. In these papers, Gluckman sees much social change in the phenomenon of personnel turn-over produced by competition for office, even though the rules of the system do not change. Change is a generic term that includes both cyclical change, or recurrence, and permanent change, or evolution. If a situation is different in time two from its previous state in time one, we say something about it has changed; if at time three it is the same as it was at time one, then we can suspect a cyclical change, in which a number of important determinants have remained constant. If at time three and succeeding times the phenomenon remains different from time one, then it may be thought of as having evolved. The difficulty with Gluckman and many of his colleagues is their desire to see one of these as a problem in maintenance and the other as a problem of another sort, when in fact they are the same problem, that of the determinative variables and their intensities that produce any social situation whether it is stable, changing in cycle, or evolving. Gluckman s unique place within this set of assumptions accepted quite widely in social anthropology derives from his Hegelianism and his use of semantic analysis. Gluckman has analyzed social situations in much the same way as the rest of his British colleagues. After recording the data and isolating the structure, i.e. the rules, analysis requires that invariable processes be isolated (cf., p. 236). Gluckman s own contribution to this approach is to look over the processes for a dialectic, a conflict that acts as a source or energizing font to maintain the system. For him the source of social process is conflict, while Radcliffe-Brown claimed simply that social process is the functioning of structure, i.e., the rules operating in action. Much of this conflict can be isolated by analyzing social behavior, and it can also be seen in the words used in rituals and ceremonial observances. Indeed, as I read him, Gluckman feels that the greater the acceptance of the established order as right and good, then the greater the possibility of rituals of rebellion (cf., p. 127). However, what the rituals actually mean is a problem of interpretation, and in this case of semantics, since we are dealing with the words used during the rituals. Alternative explanations to those given by Gluckman are available. Thus in the rituals associated with Swazi succession, he says that they reflect conflicts ir, the structure of the state over the kingly office. They could also be interpreted, as Barn& does for the Ngoni ritual of succession (p. 168), to mean that any man who stands out from his social surroundings is courting enemies and enhancing his susceptibility to malevolence. Thus in none of the Swazi songs do people say we hate you, to the king, instead it is always they who hate the king, or you who are his enemies, i.e., a king always has enemies. Which interpretation is right, or are they both correct and simply form part of the meaning set attributable to these ritual statements? This problem can be seen more clearly in a

7 956 American Anthropologist [67, 1965 recent work by one of Gluckman s colleagues, V. W. Turner ( Some Symbols of Passage in Ndembu Circumcision Ritual, in Essays on the Rituals of Social Relations, edited by Max Gluckman, Manchester University Press, 1962) in which he uses terms from semantics such as significata, polysemous, positional meaning, and so on. The trouble with this very interesting approach is its objectivity, and even when that is taken care of, its relationship to cognition on the one hand and cultural tradition on the other. It is almost impossible to prove whose interpretation of meaning is correct especially across languages and cultures when only one meaning is given to key words. Thus Gluckman s use of the word hate for Swazi rituals of rebellion, or Barnes use of the word kill are suspect. For example, Kanuri, and some other African languages use the word eat to mean roughly comrume; i.e. eat food, cheat some one and thus consume his goods, take away some one s power in a real or a suprrnatural sense, or even simply to diminish something, e.g., the wind ate the tree, i.e. the wind blew down a branch. Thus we must have the meaning set for hate and kill to judge the validity of these interpretations. In American anthropology, especially linguistics, there is now an attempt to develop serious and objective methods for carrying out semantic analysis, and this parallel attempt. by the social anthropologists to carry out semantic analyses would probably benefit from a closer scrutiny of such work. However objective though it may be, even componential analysis gets at only one level, that of the cultural object itself, usually the word or words. It cannot without independent validation make reliable claims about the cognitive process of informants who use these words, since the word or words are part of the cultural tradition, and not necessarily the cognitive make-up of individuals. Parenthetically, it should be noticed that a particular thread of enquiry has led, not to sociological research, but to cultural products, making us wonder again why the term cultural anthropology is not more widely used in Britain. Finally it is necessary to comment on Gluckman s own evaluation of his theoretical contribution. In his introduction, the author very flatly rejects a major assumption upon which much of his work has been based. He says I now abandon altogether the type of organic analogy for a social system with which Radcliffe-Brown worked, and which led me to speak of civil war as being necessary to maintain the system (p. 38). This amazing recant, of which there are very few in anthropology, brings Gluckman much closer to contemporary social scientific thinking outside British social anthropology. In the last 15 pages of his introduction (pp ) Gluckman considers the operation of variables on social systems, and in so doing moves from the qualitative concept of society as an equilibrium to a probablistic view of social reality in which a number of factors coming from all levels of analysis are seen to play a part in the formation of observable social phenomena. Although this approach is presaged here and there in his papers, e.g. in the variables mentioned in the Mau Mau paper, it is not a major feature of his past work. Now after more than two decades of writing and thinking he is still fresh enough to change and think in terms of series of social processes, operating within an ecological setting, and the bio-

8 COHEN] Review Article 957 psychical framework of human life, as well as the restrictions and action of a technology and a culture (p. 38-9). In this one phrase Gluckman has brought his own ideas into line with that of emerging social science in the world at large. If he represents a current of thought among the rest of the British social anthropologists, we can look forward to a much greater degree of agreement on basic approaches to social reality on the part of all anthropologists. I have taken some pains to describe and critically comment on the theory and concepts presented in this collection because Gluckman himself did not analyze what is now apparent; British social anthropology, valuable as it is and has been in producing the most excellent descriptions in the history of the discipline, has for some time now been in a theoretical cul-de-sac. However, leaving aside historical research which Gluckman also encouraged and carried out, there are a number of other steps, exemplified in his work, that have been taken to get out of this difficulty. The acceptance of the organic analogy with its stress on equilibrium, such that dynamics became the study of recurrent processes, has led to the formal study of rules. By studying conflicts as these were manifest in the meaning of ritual, Gluckman opened up another thread, that of semantic analysis, which is currently being probed in America as well. Finally by rejecting the organic analogy and searching for factors that are associated concomitantly with, and in antecedence to, the social phenomena under study, Gluckman has indicated his acceptance of the idea that a general social science approach using statistical arguments is another way out of the problem. For those who would like to pause a moment and take a look behind to see where we have been and, hopefully, where we may go from here, I would recommend a careful reading of this book. NOTES The author would like 1.0 express his appreciation to Professor Elizabeth Colson who made some useful comments on this review article, and most especially to Professor Gluckman who has commented at great length on a number of major points as well as making some useful editorial corrections. Professor Gluckman has commented in detail (personal communication) on his use of psychological analysis. His position is that he utilizes what research results psychology has to offer as a set of givens in the analysis of social and cultural phenomena, and that in this sense he is not doing research at the psychological level. Furthermore Professor Gluckman stresses that he distinguishes sharply between the words psychological and psychical. Psychological in his terms refers to relations between events in individual mental systems, while psychical refers to the acts of behavior (including feelings, etc.) of individuals. He suggests that he deals only with the psychical and leaves the psychological level to those trained specifically to deal with it. I will not discuss this point further here, although I hope at some future date to make clear my own position on this important methodological issue.