UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Anthropological approaches to the understanding of witchcraft and sorcery : an historical and critical.. Campbell, Alastair Fraser 1973

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


UBC_1973_A8 C34_6.pdf [ 15.38MB ]
JSON: 1.0101396.json
JSON-LD: 1.0101396+ld.json
RDF/XML (Pretty): 1.0101396.xml
RDF/JSON: 1.0101396+rdf.json
Turtle: 1.0101396+rdf-turtle.txt
N-Triples: 1.0101396+rdf-ntriples.txt
Original Record: 1.0101396 +original-record.json
Full Text

Full Text

r ANTHROPOLOGICAL APPROACHES TO THE UNDERSTANDING OP WITCHCRAFT AND SORCERY AN HISTORICAL AND CRITICAL STUDY WITH SPECIAL REFERENCE TO THE WORK OF E.E. EVANS-PRITCHARD AND CLYDE KLUCKHOHN by ALASTAIR FRASER CAMPBELL B.A., University of Otago, 1967 Research Associate in Semiotics, International Centre for Semiotics and Linguistics, Urbino, 1972 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in the Department of Anthropology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June, 1973 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. (Alastair Fraser Campbell) Department of Anthropology The University of British Columbia Vancouver 8, Canada Date 25 - 5 - 197^ ABSTRACT Attempts to establish cross-culturally valid defini tions of witchcraft, sorcery and destructive magic are mis leading, since these phenomena do not constitute true classes, but bear only a family resemblance to each other. Moreover, the attempt to establish such definitions violates the integrity of native categories of thought, and thus ob scures the understanding of the way in which thought is manifested in actions taken in specific behavioural con texts. The understanding of native categories of thought, and of the way in which these are translated into overt be haviour in specific contexts of action, is conditioned by our prior experience as the members of a particular culture and social system. Our culturally acquired notions of the nature of human society, and of reality more generally, enter into our perception of the characteristics of primi tive societies. Particularly difficult for us, coming from a culture in which our notions of rationality are deeply influenced by the subject matter and methods of the natural sciences, is the understanding of behaviour associated with ideas of magic and witchcraft. A review of the history of anthropological theory in dicates a wide variety in approaches towards the understand ing of these phenomena. Thus magic and witchcraft have been variously interpreted as historical survivals from an earlier phase of human social evolution, as manifestations -ii-of a particular mentality peculiar to primitives, as an af fective response to situations of anxiety, as a mechanism providing for the release of tensions consequent upon life in society, and as a cosmology in terms of which natural and social relationships are ordered. The scope of such interpretations has ranged from generalizations made on the basis of a wide range of pheno mena, and aiming at cross-cultural validity, to interpreta tions of a restricted set of data from only one culture. It is with interpretations of the latter type that witch craft and sorcery become subjects of study in their own right, instead of being subsumed under some theory purport ing to hold true for the entire domain of magic and religion, or even primitive mentality as such. Tylor, Frazer and Levy-Bruhl may all be regarded as having offered theories of general applicability, in con trast to KLuckhohn and Evans-Pritchard. (Malinowski stands as an intermediate figure in this respect). But while, from this point of view, KLuckhohn and Evans-Pritchard may be grouped together, their work may nevertheless be contrasted in other respects. Thus, Evans-Pritchard emphasizes the logical coherence and rationality of Zande witchcraft, of which he tries to present the sense, and which he analyses within the framework of a sociologistic and structuralist approach. KLuckhohn, on the other hand, presents Navaho witchcraft as essentially irrational, and as standing in need of an explanation which he provides in terms of a psychologistic and functionalist theory. Implicit in these anthropological approaches are de--iii-finite assumptions about the nature of Western science, on the basis of which a number of oppositions have been posed between scientific thought and beliefs of a magico-religious order. An examination of the nature of scientific activity suggests that most of these assumptions are mistaken. By focusing upon the content of scientific thought, and the imagined psychology of the individual scientist, anthropo logists have overlooked the structural similarities between scientific beliefs and activities, and the beliefs and ac tivities characteristic of magic and witchcraft. As a re sult, they have failed to understand the most important de termining characteristic of each - the social context in which such thought operates. CONTENTS Abstract ........ i I. PROBLEMS OP TERMINOLOGY AND DEFINITION 1 A. Introduction 1 B. How Terminologies Develop 3 C. Magicf Witchcraft and Sorcery. . . 6 D. Terminological Usages 11 E. Resolution of Terminological Dif ficulties 34 Notes and References. .......... 44 II. FURTHER METHODOLOGICAL CONSIDERATIONS 50 A. Introduction 50 B. Cultural Knowledge as a Cultural Phenomenon 52 C. Levels of Cultural Understanding . 53 D. Their Ways of Thinking and Ours. . 59 E. The Importance of History 64 F. The Development of Anthropological Theory 67 G. Evans-Pritchard and Kluckhohn. . . 73 Notes and References 77 III. THE PROGRESSIONISTS 82 A. Introduction 2 -V-B. Magic and the Association of Ideas 86 C. Magic, Science and Religion. ... 91 D. The Stability of Magical Belief. . 95 E. A Critique of Tylor and Prazer . . 96 Notes and References. . 102 IV. LUCIEN LEVY-BRUHL . . . 107 A. Introduction 107 B. The Nature of Primitive Mentality. 113 C. A Critique of Levy-Bruhl 117 Notes and References 123 V. W.H.R. RIVERS 129 A. Introduction 129 B. Medicine, Magic and Religion ... 129 C. A Critique of Rivers 133 Notes and References 137 VI. AN AFFECTIVE THEORY OF MAGIC 138 A. Introduction • 138 B. Malinowski • s Psychologism 140 C. The Uniformity of Primitive Magic. 142 D. Anxiety and Magic 145 E. Black Magic. . . 152 F. The Politico-Legal Functions of Sorcery . 154 G. A Critique of Malinowski 156 -vi-Notes and References 166 VII. EVANS-PRITCHARD'S ANALYSIS OP ZANDE WITCHCRAFT. 170 A. Introduction 170 B. Evans-Pritchard•s Viewpoint. ... 172 C. Witchcraft Explains Misfortune . . 173 D. Action Against Witchcraft 178 E. The Direction of Witchcraft Accusa tions 18F. Scepticism and the Variability of Zande Beliefs 192 G. The Persistence and Stability of Zande Belief 196 H. A Critique of Evans-Pritchard. . . 207 Notes and References 209 VIII. KLUCKHOHN*S ANALYSIS OF NAVAHO WITCHCRAFT ... 214 A. Introduction 214 B. The Socialization of Witch Beliefs 216 C. „uckhohn*s Functionalism. . ... 219 D. The Individual Functions of Navaho Witchcraft 220 E. The Social Functions of Navaho Witchcraft 236 F. The Cost of Navaho Witchcraft. . . 239 G. Historical Fluctuations in Witch craft Activity 240 H. A Critique of KLuckhohn 244 Notes and References 248 -vii-IX. THE IMAGE OF SCIENCE AND THE UNDERSTANDING OF WITCHCRAFT 253 A. Introduction 253 B. The Progressionists 25C. Malinowski 257 D. E.E. Evans-Pritchard 263 E. Witchcraft and Science as Para digmatic Activies 269 Notes and References 274 Bibliography . . 278 DIAGRAMS 1. Witchcraft and Sorcery as Polar Ideal Types. . . P. 36 2. Family Resemblance and Universal Class P. 40 3* Oppositions Between Science and Magic. ..... P. 60 4. Homeopathic and Contagious Magic P. 88 5. Magic, Sorcery and Taboo ............ P. 90 6. Tylor's View of Magic, Science and Religion. . . P. 93 7. Frazer's View of Magic, Science and Religion . . P. 93 CHAPTER ONE PROBLEMS OF TERMINOLOGY AND DEFINITION A. Introduction In this thesis, I propose to examine the gradual de velopment of some of the approaches which anthropologists have adopted towards the phenomena of witchcraft and sor cery in primitive societies.1 In doing so, I will place a special emphasis on the work of Clyde Kluckhohn and of E.E. 2 Evans-Pritchard. It is with these two writers that witch craft and sorcery become subjects of study in their own right, instead of being regarded as merely particular in stances of "sympathetic magic", the "principle of parti cipation" or primitive man's need for a rationality-sub stitute. Kluckhohn*s study of the Navaho develops a psycho-functionalist theory of witchcraft and sorcery, partly deriving from Malinowski*s theory of the affective nature of magic, but also incorporating certain elements of psycho-analytic theory. Evans-Pritchard*s study of the Azande, on the other hand, is representative of a strict ly sociologistic approach which we might call structural ist.^ it is the latter work which has had the deepest impact on the anthropological profession and especially on British anthropologists, most of whose work on witch craft and sorcery since the publication of Witchcraft,  Oracles and Magic among the Azande must be regarded as -2-an elaboration - and even simple repetition - of themes already present in Evans-Pritchard*s analysis. In this respect, the Zande study has come to play a role in (British) social anthropology analogous to that ascribed by Thomas S. Kuhn to paradigms in the natural sciences. Evans-Pritchard*s study, in other words, has provided a community of researchers with a universally recognized achievement providing model problems and solutions; for what has been perceived as constituting a particular con-4 stellation of phenomena. It is significant that neither Witchcraft, Oracles  and Magic among the Azande nor Navaho Witchcraft attempts a systematic treatment of witchcraft or sorcery in gener al. Both exhibit a strictly ethnographic concern, and in this respect are typical of post-Malinowskian and pre-Levi-Straussian anthropology. Indeed, despite the return in recent years of anthropological interest in problems of an explicitly comparative nature, the analysis of sor cery and witchcraft has largely continued to remain on the ethnographic level. There have been a few recent attempts at conscious cross-cultural comparison, but these have been more in the nature of probes than anything else, and 5 in any case have been atypical. Indeed, the only serious attempt to assemble all the available material on witchcraft and sorcery in tribal societies has been that of Lucy Mair.^ Moreover, even Mair*s survey is mainly in tended to serve as a popular introduction to the subject. Consequently, she makes no attempt to impose a general in terpretation on her work, and mainly limits herself to a -3-criticism of existing theories. Yet if we thus have no general theory of witchcraft and sorcery in our possession, what sense can it make to speak of witchcraft and sorcery in general terms? And if we say that Evans-Pritchard's study of the Azande has come to serve as something like a paradigm in directing later studies of witchcraft and sor cery, what is there about the phenomena of witchcraft and sorcery which have been investigated in these later stud ies which makes them amenable to analysis in terms of a model originally devised to fit the facts of a different ethnographic context? B. How Terminologies Develop The principal aim of the anthropologist lies in try ing to render intelligible to the audience for whom he writes - primarily his professional colleagues - the mode of life of the members of another cultural group. This task of "rendering intelligible" involves two essential components: i. an explication of the conceptual categories in terms of which the people whom he is studying render explicit their views on the natural and social order, and ii* the determination of the relationship of this concep-7 tual system to manifest behavior. Thus the anthropologist is necessarily required to attempt the translation of concepts from one idea system (that of the culture he studies) into another (that of his colleagues). The difficulty posed by this necessity of translating arises, of course, from the fact that the con cepts which must be translated are often not referential 0 -4-(for example, in referring to notions like goodness, in finity or time), while those that are referential often, divide the world of experience in different ways. Thus it is that we find Mair, in trying to render a West Af rican concept in English, referring to the shrine of a g "talisman - or fetish, or god...". Yet, despite difficulties of this sort, anthropolof-gists have in fact succeeded in developing a common set of terms for communicating information among themselves concerning the institutions and beliefs of other peoples. This terminology has been developed roughly as follows. In his studies of other cultures, the anthropologist has been confronted with institutions and conceptual struc tures which seem to bear some resemblance - perhaps, fol-9 lowing Wittgenstein, we might say family resemblance -to institutions and conceptual structures with which he is familiar from his own culture or from other cultures his colleagues have studied. In this way, he is led into describing these phenomena in terms that are derived from other cultural contexts, labelling one aspect of social life "marriage", another "taboo", "animism1*, "gift ex change", "initiation ceremony" or Chieftainship"• How ever, we must keep our guard in relation to this prac tice. For while it seems inevitable if cross-cultural comparison is to proceed - the logical alternative being ethnographies written entirely in the native language -we must nevertheless eventually confront the question of whether or not these phenomena, apparently similar, are really sufficiently alike to warrant designation by the -5-same term. Leach's discussion of the difficulties of elucidat ing any universally valid definition of marriage seems worth mentioning at this point in order to illustrate the kinds of terminological difficulty anthropologists face. Starting from the premise that marriage represents a "bundle of rights", Leach shows that any attempt to go beyond this formula and seek a universally valid defini tion of marriage is in vain. This is for the reason that marriage may serve, in differing societies, to establish widely differing relationships of right and obligation. Leach, for example, lists ten examples of such rights and obligations, and the list could easily be extended. The important point is that in no society does marriage estab lish all these rights and obligations simultaneously: nor, on the other hand, is there any single one of these rights and obligations which is invariably established in every known society by marriage. Thus the institutions commonly described as marriage do not by any means all have the same legal and social concomitants. Hence the anthropolo gist's dilemma: if, in order to compare the marital insti tutions of different cultures, he frames a definition of marriage drawn from one culture, then the marital institu tions of other cultures will be misdescribed in terms of that definition. On the other hand, the attempt to formu late a definition of marriage which would fit the facts from every culture equally well would result in a concept so neutral and bare as to be devoid of content and mean-10 mg. -6-leach has also attacked anthropologists for tending to treat words like "sibling", "filiation*', "descent." and "affinity" as absolute technical terms which can be dis tinguished from one another by a priori reasoning without regard to ethnographic evidence.11 In a similar vein, L6vi-Strauss has called into question the validity of the concept of totemism. He has argued that this concept rep resents an improperly constructed semantic field, illegi timately grouping together a complex of customs and be liefs "actually extremely heterogenous and difficult to 12 isolate". C. Magic. Witchcraft And Sorcery When we consider the different customs and beliefs which have "been described as "destructive magic", "sor cery" and "witchcraft", we may note also with these pheno mena a wide diversity in belief and practice from one so ciety to another. This diversity has been recognized by anthropologists themselves. Thus Kluckhohn writes that "Navaho: •witchcraft* ... must immediately be recognized as a horse of a different colour from most Melanesian •witch crafts in that Navaho Switches1: seldom boast openly of 13 their power and are not available as hired agents". Si milarly, Evans-Pritchard has drawn attention to the dif ferences between Zande and European conceptions of witch craft : When a Zande speaks of witchcraft he does not speak of it as we speak of the weird witchcraft in our own history. Witchcraft to him is a commonplace happening and he seldom -7-passes a day without mentioning it. ... Unless the reader appreciates that witchcraft is quite a normal factor in the life of the Azande, one to which almost any and every happening may be referred, he will entirely misunderstand their atti tude towards it. To us witchcraft is something which haunted and dis gusted our credulous forefathers. But the Zande expects to come across witchcraft at any time of the day or night. He would be just as surpris ed if he were not brought into daily contact with it as we would be if confronted by its appearance. To him there is nothing miraculous about it.14 In view of this, it is important to distinguish the magical beliefs and practices of the Azande from those that might be held and perhaps practiced by a person be longing to our own culture. Moreover, these differences involve far more than a matter of mere familiarity, al though it would be wrong to underestimate the importance even of this. lor our own culture bestows a different meaning on witchcraft and magic from the Zande. As Peter Winch explains, Concepts of witchcraft and magic in our culture, at least since the ad vent of Christianity, have been pa rasitic on, and a perversion of other orthodox concepts, both reli gious and, increasingly, scientific. To take an obvious example, you could not understand what was in volved in conducting a Black Mass, unless you were familiar with the conduct of a proper Mass.and, there--8-fore, with the whole complex of reli gious ideas from which the Mass draws its sense. Neither would you under stand the relation between these without taking account of the fact that the Black practices are rejected as irrational (in the sense proper to religion) in the system of beliefs on which these practices are parasitic.15 In the light of these observations it becomes signifi cant that, in searching for a notion from our own culture with something of the same meaning for us that of witch craft has for the Azande, Evans-Pritchard did not select some idea drawn from a ritual or ceremonial context, nor some belief associated with the lore of Satanism, but in stead decided on the familiar everyday notion of "bad luck".16 Witchcraft, magic and sorcery may therefore vary con siderably from one society to another, and what is called witchcraft in one culture may not resemble witchcraft in another culture so much as some other institutionalized be lief. The main dimensions along which beliefs and actions relating to witchcraft, magic and sorcery might vary would seem to be the following: i. Content - variation will occur in the constituent ele ments constituting particular traditions of witchcraft, magic and sorcery. To take some simple examples, among the Azande it is believed that witches inherit their harmful powers unilineally, in the form of a physical substance; in Europe this belief is absent and notions of witchcraft, are firmly linked to those concerning a pact with the Devil; -9-among the Navaho, on the other hand, both of these beliefs are absent and witchcraft ideas place a heavy emphasis on such practices as fratricide and were-animalism. ii. Meaning - even where identical or similar constituent elements are discovered in the belief systems of two dif ferent cultures, these can still not necessarily be equat ed, since each element derives its sense from the sum of its relations with the other elements of the total con ceptual system to which it belongs. Certain types of Navaho witches, for example, are believed to participate in secret nocturnal gatherings. At these gatherings, concerted ac tion against victims is planned, new members are initiated, cannibalism and sexual intercourse with dead women are practised, and victims are killed at a distance by ritual 17 means. These gatherings bear a striking resemblance to the European Witches1 Sabbath, but it would be wrong to re gard them as equivalent on that count. This would be to make a mistake of the kind that Prazer made. For the European Witches* Sabbath has no meaning in itself, but only in the context of the total demonological ideology and Christian belief. In the same way, the nocturnal gatherings ascribed to Navaho witches cannot be understood if they are abstracted from their cultural context, but only if they are related to the totality of Navaho witch beliefs and, indeed, ,the entire Navaho world-view. iii. Function - the implications of witchcraft, magic and sorcery for the social group considered as a perduring unit, and also for the human individual, will vary. For instance, a high proportion of Navaho witchcraft gossip re--10-fers to witches who live in distant localities and are thus rarely or never seen. Feuds involving these alleged witches are consequently unlikely to develop, and Navaho gossip about witches is therefore far less disruptive than in those societies, like the Zuni, where such gossip is centred in 18 the local group. To take a somewhat different example, Kluckhohn implies that a significant number of witchcraft accusations among the Navaho are made against the wealthy by those of a lower socio-economic status. In this case then, witchcraft belief operates as something in the nature of a levelling device, discouraging the undue accumulation of wealth by certain individuals, since the rich man knows that if he is stingy with his relatives and others, 19 he is likely to be spoken of as a witch. In contrast to this, no Zande commoner would dare accuse a prince of practising sorcery or of bewitching him, while only occasionally will a poor commoner accuse a wealthy commoner. Moreover, the whole structure of Zande oracle consultation serves to reinforce the status differentiations of Zande society. The King's is the final decisive oracle for pur poses of identifying witches and the victims of vengeance magic, and no appeal from it is recognized or permitted. Since the authority of the King upholds that of the oracle, any Zande who consults the oracle implicitly recognizes the authority of the King. Thus there is a relationship of re-, 20 ciprocal support between the authority of King and oracle. Similar variations occur on the individual level. Por example, Kluckhohn's argument, that Navaho accounts of witches copulating with dead women provide a channel for the release in fantasy of certain culturally disallowed aber-rations of a sexual nature, seems plausible. Probably a similar argument could be made in relation to certain as pects of European demonology, but would seem difficult to maintain with regard to the facts of Trobriand or Zande witchcraft. Considering the extensiveness of this range of pos sible variations, it becomes obvious that; there are grave difficulties confronting any anthropologist who would at tempt to develop a suitable terminology for the purposes of cross-cultural analysis. It remains, however, worthwhile examining the extent to which anthropologists have attempt ed this task, and degree of their success in this venture. Terminological Usages Contemporary researchers in this field employ a stock of terms ultimately derived from standard English usage. In standard English, these terms have a range of meanings that are primarily descriptive of the facts of English witchcraft and magical beliefs. Most of them entered the English language with their current meanings in the six teenth century. The most important of them are listed be low, with their most common meanings: Magic - the believed art of influencing the course of events by means of compelling the intervention of spiritual beings or by using some other occult principle. A synonym for sorcery and witchcraft. Sorcery - the practice of magic or enchantment; witchcraft. Sorcerer - the practitioner of sorcery; a magician or wi zard. Witch - a female magician; a sorceress; a term especially -12-used to describe a woman in league with the Devil or evil spirits. Wizard - a male who practises witchcraft. 22 Warlock - the male equivalent of a witch. It was from this usage that these terms were at first incorporated into anthropology. Thus Prazer, while modi fying the concept of magic in his attempt to define its logic (as he understood it), continued to employ the term "sorcerer" as equivalent to "magician", and to follow the practice of using "sorcerer" and "witch" as sexually speci fic terms. Prazer's use of the terms "magic", "science" and "religion" is somewhat more complex, since he attmpted a differentiation of universal validity between these three phenomena. Unfortunately his efforts in this direction re sulted in an unacceptable a prior ism. Thus, for Prazer, any technique not based on a knowledge of objective condi tions, and not seeking to operate through the goodwill of spiritual intermediaries, was ipso facto magical. The trouble with such an approach is that it leads to a group ing of phenomena in a manner very different from the way in which the people being studied group them. To take a sim ple example, among the Navaho menstrual blood is believed to be sometimes administered in food as a malicious act. If we were to define witchcraft and sorcery in an a priori fashion along the lines of "the belief in the power of hu man agents to cause harm and misfortune by the use of mys tical (non-empirical) means", then it is obvious that we would have to class this act as "witchcraft" or "sorcery". -13-For while the Navaho consider this action to "be harmful in that it consists in the administration of a dangerous sub stance, our own scientific knowledge leads us to view this belief as mistaken and therefore as non-objective, mystic and magical. On this basis we would group it together with such other Navaho beliefs as Frenzy Witchcraft, Disease Witchcraft and Eagle Pit Sorcery, The problem is that the Navaho do not. As EZLuckhohn explains, Observations of this kind were fre quent: "Women just do that to be mean. It hurts you all right, but it isn*t a witch way," My impression is that my informants felt that menstrual blood was intrinsically dangerous - there was no need to add "magical" proce dures ,23 It might therefore appear that the Navaho think of menstrual blood as something in the nature of what we would call a natural poison. But even here we must beware of misdescribing their categories. For some of what we des cribe as natural poisons are assimilated by the Navaho to the category of witchcraft. Thus, according to the Navaho, when a man is bitten by a snake, he sickens and dies be cause of the witchcraft the snake holds in its mouth. As a Navaho legend explains it, Witchery started out under the ground. First Man, First Woman and Coyote - these three started it. After everybody got above ground First Woman gave it out. Snake wanted some too, but his mouth was the only place he could put it. And so his bite kills you,24 -14-Prazer seems to have been unaware of the problem posed by facts such as these, and it is possible for us: to apply his; terminology only by distorting the world-views that this terminology was: intended to help us understand. Perhaps the force of this objection will become clear er if we briefly consider Vilfredo Pareto*s distinction be tween logical and non-logical behaviour. Pareto's attempt to clearly distinguish the logical from the non-logical -as cross-culturally valid categories - parallels Prazer1s differentiation between magic, science and religion. It differs, however, in being more systematic and in being based on more explicit criteria than those Prazer elucidat-25 ed. But despite this more systematic nature, Pareto*s differentiation encounters similar- difficulties to> that of Prazer* s, and it is useful to consider these here-. This- is not a matter of knocking down straw men, since both Prazer* s differentiation between magic;, science and reli gion and Pareto's distinction between logical, and non-logi cal behaviour, are representative of methodological presup positions? which continue to: exercise a deep influence on anthropological thought. Levy-Bruhl's opposition betweem mystic thought and scientific thought, accepted by Evans-Pritchard, is an example of this. In Tha Mind and Society, Pareto outlines four eriteria, all of which must be satisfied in order for an action to be considered logical. These are: i. the action must be end-directed - that is to say, it must be performed: by the actor with a. goal in mind, and with the intention of achieving that goal; 15-ii, the goal which the actor is seeking to achieve must he empirically identifiable; iii, the action performed must actually tend to produce the result which the actor envisages; iv, the actor must have "good" (what Pareto describes as "logico-experimental") grounds for his belief. Conversely, an action may be adjudged non-logical according to a number of criteria: i, if the actor performs the action without intending to achieve any result by doing so; or ii, if the end which the actor hopes to achieve by his ac tion lies outside the field of observation and experiment and is therefore "imaginary" (Pareto assigns actions aiming at the salvation of the soul to this category); or iii, if the end sought is real, but is not gained in the way in which the actor thinks it is (in this category, Pareto places magic as well as certain activities like wage-cutting by businessmen under conditions of free compe tition); or iv, if the action performed actually does tend to produce the result the actor is seeking, but he nevertheless lacks logico-experimental grounds for thinking so. It is important to recognize that Pareto does not re gard non-logical conduct as being the same thing as illogi cal conduct, and in this respect it is interesting to note his remark to the effect that a mistake in engineering could not be regarded as a non-logical action. But what is the difference between a mistake in engineering and the er ror of a businessman, under conditions of free competition, -16-who thinks that by cutting his employees' wages he will in crease his profits? Moreover, is the businessman's mistake really comparable to the performance of a magical rite? Or ought it rather, as Peter Winch suggests, be compared to a mistake in a magical rite? For while the businessman's mistake is a particular act within the category of business behaviour, magical performances themselves constitute a category of behaviour. As Winch explains, Magic, in a society in which it occurs, plays a peculiar role of its own and is conducted according to considerations of its own. The same is true of busi ness activity; but it is not true of the kind of misguided business activity to which Pareto refers, for that can only be understood by reference to the aims and nature of business activity in general.26 It is because of Pareto*s failure to distinguish gen eral categories of action from particular acts within such categories, that his attempt to differentiate non-logical from illogical behaviour is so unclear. Illogical acts surely involve mistakes in logic; non-logical acts fall outside the criteria of logic. Moreover, this failure is compounded by the fact that Pareto was unaware of the fact that criteria of logic are not given, but arise out of, and are intelligible only in terms of, modes of social life. From this it follows that criteria of logic cannot be ap plied to modes of social life as such. Science, for exam ple, is one such mode and religion is another. Within each mode, actions may be either logical or illogical, but nei ther science nor religion as such is either logical or il--17-logical. Each is non-logical, and each has criteria of in telligibility peculiar to itself. It follows that to try to use the aims and activities of one as a means to under stand the aims and activities of the other can only lead to misunderstandings. But it is precisely this that Pareto -27 together with Prazer - is guilty of. It is to Malinowski, more than to any other single writer, that we owe the insight that the crucial criteria for differentiating the technical from the magical must be those of the people themselves. Por although Malinowski did not explicitly discuss this criterion, he made use of it, and that, as Nadel comments, "is still a lesson worth 28 learning". Indeed, recognition of the necessity of pro viding an account of native categories of thought - and of not imposing his own in an a priori fashion - permeates Malinowski*s writings. Thus, in his discussion of the Trobriand outlook on myth, Malinowski asks, ...what is myth to the natives? How do they conceive and define it? Have they any line of demarcation between the mythical and actual re ality, and if so, how do they draw this line?29 Hence, instead of being presented with a set of preconceiv ed categories - like "myth", "legend", "fairy tale" or "fable", we are given Trobriand categories, with the near est English equivalent to them offered as a translation, or an explanation of their meaning when no near English equi valent exists.^ Similarly in his discussion of Trobriand magic, Malinowski did not try to provide a series of -18-a priori categories of supposedly universal value, but at tempted instead to select a set of words which best fitted 31 the facts of Trobriand belief. What, then, were these facts, and what terms did Malinowski select to describe them? The Trobrianders conceive of two kinds of practitioner of "black" magic. These are the bwaga*u, always a male practitioner, and the .yoyova, always a female. Of these, the bwaga'u is the most common and there are usually one or two men in each village who are known and feared as such. But although less common, the yoyova is considered far more deadly than the bwaga'u. For whereas the bwaga'u is merely the possessor of a powerful form of magic, the yoyova possess a mulukwausi. or disembodied second self, which can fly through the air at will. The power of the bwaga1u lies in his knowledge of spells. These he may learn from his father (without pay ment) or from his maternal uncle (if a high fee is paid). An aristocrat may also learn these spells from an unre lated commoner, on payment of a suitable fee. The yoyova, on the other hand, can only very gradually be initiated into her powers. Indeed, since the process of initiation begins with the cutting of the umbilical cord at birth, only a small child whose mother herself is a yoyova can become one. This is not, however, to suggest that the powers of the yoyova are in any way inherent or inherited. All of them derive from magic, which must be spoken at 32 every stage in the training of a young yoyova. Special spells must also be uttered by the yoyova every time she .19 wants to become invisible, to fly, or to penetrate the darkness and see if an accident is happening. The bwaga'u hunts down his quarry by placing a spell on those places the intended victim frequents. In this way, the victim is confined to bed and immobilized. This gives the bwaga'u the opportunity to creep out to the vic tim^ hut at night, which he does equipped with herbs over which a suitable spell has been uttered. These herbs the bwaga'u attaches to a long stick and attempts to thrust through the thatch wall of the hut, and into the fire over which the victim will be lying in an attempt to keep him self warm. It is believed that, should the victim inhale the fumes of these burning leaves, he will contract a dead ly disease. In another rite the bwaga'u carries out, some coconut oil is first boiled in a small pot. Leaves of various herbs are then soaked in this oil and later are wrapped around a stingaree spine or some other pointed object. An incantation is chanted over the spine with attached leaves and the bwaga'u hides with it behind a shrub or house. Then, on sighting his victim, he thrusts the "dagger" he has made in his direction and violently turns it in the air, as if to stab the victim and twist the spine in his wound. This rite is thought never to fail in killing a man, if 33 properly carried out and not countered by another magician. When the yoyova - in her mulukwausi form - attacks a victim, it is believed that she may do so simply by kicking or hitting him or her: illness results. More dangerously, the mulukwausi may pounce on a victim and remove his or her 20 lungs, heart, brain or tongue. These may be devoured on the spot, or carried away to be consumed at some future date. The victim is thought certain to die in a relatively short period of time, unless another yoyova is paid to search for and return the missing parts before they are eaten. The bwaga'u practises his art on his own behalf, or for a fee. In purchasing his services, chiefs and men of rank have first claim and he would not sell his services to lesser men for unjust causes. Among his main functions are the safeguarding of the rights and privileges of the chief. Thus the bwaga'u will punish direct breaches of etiquette and ceremony, as well as flagrant offences like adultery with one of a chief's wives. The bwaga'u. therefore, al though he may sometimes abuse his powers, cannot be thought 34 of as a criminal but is a mainstay of the social order. In contrast to the bwaga'u. who often treats his spe ciality quite openly in conversations, the role of the yoyova is never publicly donned. Ho woman would ever di rectly confess to being a yoyova, even to her husband. But, at the same time, some women are marked as such and even play up to the role. Por supernatural powers are a good source of income and a reputed yoyova will receive gifts on the understanding that a particular person is to be injured in return for them, or in order to undo the da mage another yoyova has caused. Her role is therefore semi-public and the most important yoyova are known hy name. Moreover, according to Malinowski, "to have such a character would in no way spoil matrimonial chances or do anything hut enhance the social status of a woman"• Malinowski offered, as English equivalents to the terms bwagaftu and yoyova, the words "sorcerer" and "witch" respectively. He thus continued the popular English use of these terms in a sexually specific manner, a practice which is perfectly reasonable when it is recalled that the Tro-brianders themselves draw a distinction between the ma gical powers of the male bwaga*u and the female yoyova. This usage was also followed by Eeo Fortune in regard to the in some respects similar material from Dobu,^ but otherwise has not generally been adopted by anthropolo gists. Another aspect of Malinowski*s terminology has proved more enduring. It will be noted from the above discussions that Malinowski used the term "sorcery" to refer to the whole domain of destructive magic and not only to that part of it deemed illegitimate by the members of the social group concerned. Oceanianists have continued to employ his 37 terminology in this respect, although some writers (E.M. Berndt and Meggitt) have begun to speak of Retaliatory sor cery" or "legitimate sorcery" in order to differentiate that branch of destructive magic which constitutes the so cially sanctioned redress of a wrong from "illegitimate sorcery", or socially condemned destructive magic. In do ing so, these recent writers on Oceania have adopted a dis-tinction long accepted by African specialists. Standard Africanist usage derives, of course, from Evans-Pritchard1s classic study of the Azande. If Malinowski did not render explicit the criteria he was em-22-ploying in establishing the categories he did, Evans-Pritchard was explicit in emphasizing that he was not in»-terested in defining witchcraft, oracles and magic as "ideal types" of thought, hut was anxious rather to explain what the Azande themselves understood when they spoke of mangu, soroka and ngua. Thus he wrote, I am not greatly concerned with the question whether oracles should be classified as magic, nor whether the belief that children are unlucky who cut their upper teeth before their lower is a form of witchcraft; nor yet whether taboo is negative magic• My aim has been to make a number of English words stand for Zande notions and to use the same term only and al ways when the same notion is being discussed* For example, the Zande does not speak of oracles or taboos as ngua, and therefore I do not call them "magic".39 Unlike the Trobriand Islanders, the Azande do not make any differentiation between the occult powers of male and female. Thus in contrast to the Trobriand distinction be* tween male bwaga'u and female yoyova, the Azande make a differentiation of a quite different order: that between mangu. gbigbita ngua and bagbuduma. In translation of these Zande concepts, Evans-Pritchard offered the terms "witchcraft", "sorcery" and "vengeance magic" respectively. In doing so, he abandoned the convention - followed by Malinowski - of using the terms "sorcerer" and "witch" with a sexual referent, since this referent has no sense in the context of Zande society. Even more radically, in using the term "witch" to describe the Zande possessor of mangu, 23-Bvans-Pritchard abandoned the old link between the notions of witchcraft and magic. The Azande conceptualize manga as a physical substance found in the bellies of ira mangu (possessors of manguK It is usually described as an oval swelling or sack, of blackish or reddish colour, which sometimes contains the seeds of pumpkin and sesame consumed by the ira mangu in the gardens. Mangu is believed to be inherited unilineally from the parent of the same sex. In other words, a female ira mangu will pass on mangu to all of her daughters but to none of her sons, while a male ira mangu transmits mangu to all of 40 his sons but to none of his daughters. Ira mangu are believed to show a few external signs of their condition. Thus the possession of red eyes, or the issuing of maggots from a person*s body before burial, are 41 considered indicative of mangu. Ira mangu are also thought to have unpleasant personality traits: A spiteful disposition arouses sus picions of witchcraft. Glum and ill-tempered people, those who suffer from some physical deformity, and those who have been mutilated are suspected on account of their spite* fulness. Men whose habits are dirty, such as those who defecate in the gardens of others and urinate in pub lic, or who eat without washing their hands, and eat bad food like tortoise, toad, and house-rat, are the kind of people who might well bewitch others. The same is thought of unmannerly people who enter into a man's hut without first asking his permission; —24** who cannot disguise their greed in the presence of food or beer; who make offensive remarks to their wives and neighbours and fling insults and curses after them; and so on.* Ira mangu are held by the Azande to be responsible for misfortunes, which they deliberately cause to fall on people. It is important to note that their method of doing so involves no magical technique (there is no rite, no spell, no use of medicines and no necessary ritual condi tion on the part of the performer) but rather is in the na ture of a psychic act. The ira mangu is believed to dis patch his mbisimo mangu (the spirit of his mangu) to accom plish his ends. The mbisimo mangu is believed to remove part of the spirit of the victim's flesh to devour; by so«* doing, it causes illness and death. Ira mangu often combine to assist each other in crimes and to feast together. They are believed to be organized into a brotherhood presided over by the oldest and most ex perienced members. The training and tuition of the younger by the older ira mangu is thought necessary before a young witch can become strong enough to kill his neighbours Young children who are ira mangu are not considered by the Azande to be dangerous, since their mangu is too small to be able to injure others. It is even thought possible for a person's mangu to remain inoperative, or ncoolw, throughout the course of his or her life. In this way, it is considered possible for a man to be an ira mangu and at the same time a good citizen, meeting his obligations and 44 living on good terms with his neighbours. For mangu to -25-be dangerous, it must first be activated by hatred. As one Zande expressed it to Evans-Pritchard, hatred springs first in the breast and then goes down to the belly to rouse 45 witchcraft. Anybody except an aristocrat may be an ira mangu, and almost everyone is accused, by someone or other at some period of their lives, of having bewitched their neigh bours. Usually however, it is only those who make them selves disliked by many of their neighbours who are often 46 accused of witchcraft and earn reputations as ira mangu. Most suspicion is attached to the aged since it is believed that, generally speaking, the older a man grows the more potent his mangu becomes and the more violent and unscrupu lous he becomes in its use. Most ira mangu are believed to eventually fall victim to vengeance magic, sorcery, or the 47 malice of another ira mangu. Ira mangu are considered by the Azande to be like ira gbigbita ngua (sorcerers) who practise a variant of ngua (magic) that is considered both illicit and immoral. Ira mangu and ira gbigbita ngua are seen as being alike "the enemies of men", and are therefore placed in the same moral category. Mangu and gbigbita ngua are together op posed by and opposed to wene ngua (good magic). But gbigbita ngua is also conceived by the Azande in a very different way from mangu, in that it is not in any way de pendent on an inherent power in the ira gbigbita ngua. Rather, it employs a magical technique and is available to anyone who can secure the requisite medicines and who is familiar with the proper procedure for their utilization. -26-The most feared of all bad medi cine, and the one most often cited as cause of sickness is menzere. It is: probably derived from an arboreal parasite. The sorcerer goes by night, generally at full moon, to the home stead of his^ victim and places the medicine on its threshold, in the cen tre of his; homestead, or in the path leading to; it. As he does so he ut ters a spell over it. It is said that if he succeeds in slaying his enemy he will mourn him by wearing a girdle of bingba grass for several days after his death. If the sorcerer neglects; this rite he may fall sick. The gir dle would not lead too his detection "because men often mourn for a few. days after the death of distant relatives. Menzere is so potent a medicine that should any man for whom it is not intended step over it he will be ill for a while though he will not die. There are many antidotes to menzere and a man who knows these will be sent for immediately if a man suspects, he is; attacked by it. Menzere is regard ed with abhorrence by/ all. Azande have always; told me that in the past those who killed men with witchcraft were generally allowed to pay compen sation, but that those who killed men by sorcery were invariably put to g death, and probably their kinsmen also. Other bad medicines include the hairs of the ant-bear. A spell is uttered over these, and they are placed in a man's beer in order to slay him. They cause his neck and tongue to swell, and if an antidote is not quickly adminis tered, he will die. Gbigbita ngua may also affect the ver dict of the poison oracle, and can break up the family of a -27-49 man. It is important to understand that the Azande do not stigmatize gbigbita ngua as bad simply because it destroys the health and property of others, but because it flouts moral rules. Wene ngua (good magic) may also be lethal, but it strikes only those who have committed a crime. Thus bagbuduma, the vengeance magic used against witches, is the most destructive yet most honourable of all Zande magic. It acts only against a guilty witch, and if attempted to be used to kill an innocent man out of spite, it would not only prove ineffective, but would actually turn against and destroy the magician who sent it. Such magic operates re gularly and impartially in executing justice in accordance with the moral and legal sanctions of the community. Gbigbita ngua, on the other hand, is bad medicine, for it does not give judgements, but slays one of the parties to a dispute without regard to the merits of the case. It is a personal weapon aimed at some individual whom the ira  gbigbita ngua dislikes, but against whom he has no moral or legal case. It is used out of spite against men who have broken no law or moral convention. No Zande will confess himself to be an ira gbigbita ngua, and most do not even like to discuss the matter in case suspicions be aroused concerning the sources of their knowledge. Gbigbita ngua must be performed at the dead of night, for the ira gbigbita ngua would be slain if seen practicing his rites. In translating ngua as "magic", gbigbita ngua as "sorcery", bagbuduma as "vengeance magic" and mangu as -28-"witchcraft", Evans-Pritchard was not seeking to devise a terminology of general applicability to other cultures, but was searching instead only for some English terms which could be used consistently to refer to what are distinct Zande concepts. Nevertheless, despite his hesitation in transcending the facts of Zande ethnography, Evans-Pritchard^ terminology can be viewed as suggesting as a conventional usage the term "witchcraft" to refer to an in herited or inherent condition, "sorcery" to refer to the application of magical techniques in a criminal manner and "vengeance magic" or "destructive magic" to refer to the socially sanctioned use of magic in order to punish a cri minal or protect property. The main impact of Evans-Pritchard's work, however, was not to come until the period following the end of the Second World War, and thus the terminology he suggested did not find immediate acceptance. Kluckhohn, indeed, seems to have written Navaho Witchcraft without even having read Witchcraft, Oracles and Magic among the Azande. Perhaps because of this, Kluckhohn described a class of phenomena among the Navaho as "witchcraft" which stands far closer to "sorcery" in Evans-Pritchard's use of the term. "Sorcery" he also used in quite a different sense from Evans-Pritchard, to describe a sub-type of witchcraft. Yet des pite the differences in the terminology he adopted, Kluckhohn, like Malinowski and Evans-Pritchard, offered his terminology primarily to serve as a set of convenient la bels which could serve to approximate in translation of Navaho terms. He made no attempt to apply a set of precon--29-ceived categories in an a priori manner. Kluckhohn used the term "witchcraft" (uncapitalized) as a generic category to describe four sub-types of Navaho belief: those of Witchery, Sorcery, Wizardry and Frenzy Witchcraft (all capitalized). Of these beliefs, Kluckhohn considered the concepts of Witchery and Sorcery to be the most closely linked in native thought, and to form, to gether with Wizardry, a major pattern of attributes clearly 50 differentiated from Frenzy Witchcraft. We may briefly specify the content of each of these categories of Navaho thought as follows: i. Witchery is normally learnt from a grandparent, parent or spouse who is also a Witch. Initiation requires the killing of a close relative, usually a sibling. Witches are believed to be especially active at night, and to roam around at great speeds in the skins of coyotes, bears, owls, wolves, desert foxes and crows. They hold assemblies or "Sabbaths" at which they plan concerted actions against victims, initiate new members, kill victims from a distance by means of ritualized practices, and practise cannibalism and intercourse with dead women. At these meetings, songs are also sung and dry paintings are made. Some Navahos consider that these paintings represent intended victims and believe that the chief Witch shoots a turquoise bead into each painting by means of a small bow. Witches are also thought to make "poisons" from the flesh of corpses. These poisons may be dropped in through the smoke-holes of hogans, placed in the mouth or nose of a sleeping victim, blown from furrowed sticks into the face of a victim in the -30-midst of a large crowd, or administered by means of a ciga rette. This leads either to sudden fainting, loss of con sciousness, lockjaw and a swollen tongue, or else the gra-51 dual wasting away of the victim. ii. Sorcery is closely related to Witchery in Navaho thought, and Sorcerers are believed to take part in the same Sabbaths as those in which Witches participate. Sor cery, however, employs a number of techniques peculiar to itself. Of these, the casting of spells is the most cha racteristic. There is no need for the Sorcerer to person-* ally encounter his victim. All that is necessary is that some clothing or personal offal, belonging to the victim, be obtained. This is then buried together with corpse flesh or some other material from a grave, or alternatively is buried in a grave, or under a tree which has been struck by lightning. The Sorcerer then recites a spell, often specifying the number of days after which the victim is to die. This spell may be chanted, sung, or both chanted and sung. Sometimes a "good prayer" may be recited backwards as a part of the technique. Other procedures followed by Sorcerers include whispering a spell while stepping over someone, torturing the effigy of a victim, the use of evil*-* wishing sand-paintings similar to those Witches employ, scratching the image of a victim on a stone which is then concealed in the victim's home, car or saddle-bag, and the cutting-open of the belly of a toad and the placing inside it of a charm while repeating a spell. Each Sorcerer is also believed to have a special power which assists him. The earth, the sun, lightning, bears, owls and snakes are -31-all examples of these powers. Whirlwinds and some animals, 52 especially dogs, are also thought to practise Sorcery. iii. Wizardry, like Witchery and Sorcery, requires the killing of a sibling or close relative as a prerequisite to participation. In contrast to Witches and Sorcerers, how ever, Wizards do not participate in Sabbaths, nor do they become were-animals. Particular to Wizardry is the intro-^ection of foreign objects - particles of stone, bone, ash, charcoal or quill - into the body of the victim. This is generally believed to be accomplished by placing such ob jects in a basket and making it rise through the air by in cantation, although some Navahos consider that the Wizard does his "shooting" through a tube, and some consider that he must remove his clothes and rub ashes on his body before doing this. Emaciation, together with pain in the area where the missile is lodged, is usually considered diagnos-53 tic of Wizardry. iv. Frenzy Witchcraft remains within the general corpus of Navaho witchcraft, in that it is a malevolent activity, di rected against the rich in particular. It resembles other forms of Navaho witchcraft in requiring the killing of a sibling as the price of initiation. Characteristic of Frenzy Witchcraft is the use of special plants, of which datura is the most prominent. Each plant must be gathered in a prescribed manner, and each plant has its own song. It may be administered in food, in a cigarette, or by kiss ing. Frenzy Witchcraft is associated with love magic, trading and gambling, and may also be used for success in hunting and salt-gathering. Its techniques do not involve -32-the dead and its practitioners do not attend Sabbaths or transform themselves into animals. Pear is universally ex pressed in regard to Frenzy Witchcraft, but there is: no unanimity that it is unqualifiedly bad. Some Navahos con sider that it may be usedi in relatively respectable ways like self-protection, or for success: in trading or gambling 54 against out-groupers. These categories of Navaho witch belief are so obvious ly different from those of the Trobrianders and the Azande that it is only to be expected that the terminology employ ed by Kluckhohn in translation of them should differ mark edly from that used by Malinowski or Evans-Pritchard. Moreover, it is difficult to believe that these differences would have been any less substantial even if Kluckhohn had been familiar with Evans-Pritchard1 s work, given his? con cern to understand the Navaho world-view/ rather than at tempting to develop broad cross-cultural generalizations. Our survey of the terminologies employed in the period be tween the two world wars; therefore leads us to identify two major features of terminological evolution: i. The recognition, implicit or explicit, of the importance of elucidating native categories of thought and of the necessity of adapting anthropological terminology to this end. ii. The lack of any consistency in terminological conven tions adopted by different writers. In the period following the Second World War, the sit uation with regard to the last of; these facts: has- changed considerably, and of the three different sets of terminology -33-adopted by Malinowski, Kluckhohn and Evans-Pritchard, that of Evans-Pritchard has come closest to being accepted as standard. This is for a number of reasons. Firstly, Evans-Pritchard's study was of an African people, as have been the overwhelming majority of studies of witchcraft and sorcery in the period following the Second World War. Apart from the fact that his work would therefore tend to be read more carefully by Africanists than would be the case if the Azande inhabited Oceania or the Americas, a greater similarity in the ethnographic evidence being considered would render his terminological system more directly appli cable. (It is interesting, in this respect, to note the greater influence of Malinowski*s terminological conven tions among Oceanianists.) A second factor of some impor tance is probably the fact that Evans-Pritchard is a British anthropologist and that most subsequent research on witchcraft, sorcery and magic has been carried out by British anthropologists. One of the implications of this is that his sociologistic approach is far closer to the methodological presuppositions of later investigators than is Kluckhohn*s interest in personality and culture. More over, it is necessary to point out that Malinowski's theo retical writings on witchcraft and sorcery are slight (both in size and in substance) and have not exercised a profound influence for this reason. Among conscious attempts to standardize and generalize Evans-Pritchard*s terminology, that of Middleton and Winter is worth noting. Suggesting "wizardry" as a generic term to cover both witchcraft and sorcery, they have defined -34-witchcraft as being based on a "mystical and innate power", and sorcery as a magical (objectively fallacious) technique 55 potentially available to anyone. While few anthropolo gists have accepted this suggestion to use wizardry as a 56 generic term covering both witchcraft and sorcery, most have adopted the convention of distinguishing between witchcraft and sorcery along the lines suggested by 57 Middleton and Winter. Unfortunately, this attempt at a terminological stan dardization can hardly be described as an unqualified suc cess. Por few other societies have the same notions of witchcraft and sorcery as the Azande have. Consequently, the attempt to apply the terminology, originally devised to fit the Zande facts, to other cultural contexts has given rise to a considerable degree of confusion. Por, even for-ulated as broadly as by Middleton and Winter, this termino logy encounters considerable problems of application. Thus traits assigned by one writer to "witchcraft" are assigned by another to "sorcery", and Middleton himself writes that among the Lugbara, "the ability, and the wish, to poison people by sorcery may be inherited, especially from the mo-58 ther", E. Resolution of Terminological Difficulties It is not, of course, necessary to conceptualize witchcraft and sorcery as discreet categories admitting of no intermediate forms. Rather, one might follow the pro cedure widely used in the social sciences (whether con sciously or unconsciously) of attempting to construct ideal -35-types rather than that of attempting to establish empiri cally descriptive categories of direct application. First systematized by Max Weber, this methodological approach has been suggested as relevant for anthropology by W.J. Goode. Goode applies the method, in the form of oppositions of po lar ideal types, to try and make some sense of the old dis tinction between magic and religion. He explains that: In its application one accepts the idea that any given magical or reli gious system is concretely not to be found at either extreme, but some where between the two. This is, of course, always an approximation, as the application of any scientific concept to concrete situations will be: the unique situation or phenome non rarely, if ever, equates with the conceptual description or theoretical formulation of any science. Further more, the decision as toward which pole a supernatural system falls re-» quires several characteristics, each of which is a variable running be tween two opposing or antithetical forms•59 Utilizing Goode*s suggestion, we might construct polar ideal types, opposing sorcery to witchcraft, in some such manner as is illustrated in Diagram One (next page). If such a scheme were employed, most of the phenpmena so far described in this chapter could be grouped without difficulty as standing closer either to witchcraft or to sorcery. It would, however, be the sorcery pole which con formed more closely to the empirical data than the witch craft pole. For example, the Azande believe that witch--36-CHARACTERISTIC WITCHCRAFT SORCERY Acquisition inherited learnt Method innate power magical technique Psychology unconscious & unintentional conscious & intentional Potential Distribution restricted unrestricted Diagram 1. Witchcraft and Sorcery as Polar Ideal Types craft may be conscious and unintentional, but also consider that it is usually conscious and intentional. Zande witchcraft does not, therefore, exactly coincide with the witchcraft pole of our ideal type. Zande sorcery, on the other hand, exactly fits our designation of it in terms of the ideal type. But even although the attempt to construct ideal types might seem to offer advantages in comparison to that of trying to formulate categories of direct empirical applica bility, the construction of polar ideal types of witchcraft and sorcery along the lines suggested above nevertheless encounters serious difficulties. This is for the reason that any attempt to formulate cross-cultural categories on a content basis necessarily involves a departure from the categories of particular societies and may therefore impede our understanding of these. -37-Por instance, the set of criteria used to group a par ticular constellation of phenomena towards one end or the other of the continuum, may not really he equivalent. Among the Azande, to take one example, is the notion that bewitchment may constitute an unconscious and unintentional act as significant as the notions that witchcraft is in herited and constitutes an innate power? Perhaps one might try to circumvent this objection by arguing that the criteria selected for constructing an ideal type must be logically dependent on each other. Thus one might argue that the criteria selected for defining sorcery in Diagram 1 (preceding page) are significant in that it follows from the fact that sorcery employs a magical technique that it must be learned, that its application must be conscious, and that its potential distribution among the population is unlimited. Against this view, I would argue that logical dependencies are significant only in so far as they are actually perceived as being such within the contexts of the particular belief systems con cerned. Different peoples will draw different conclusions from the same premises, and the same conclusion from dif ferent premises. The problem of selecting criteria and of determining their equivalence therefore remains. In many respects, the attempt to construct ideal types of magic and sorcery may be compared to Pareto1s search for residues (recurring features in our observation of human society which provide a suitable subject for scientific generalization). Examples of residues are baptism (the use of water, blood or other substances for purposes of ritual -38-or moral purification) andi sexual asceticism (the notion that sexual relations are to be avoided! as being morally or physically debilitating, or for some other reason). The problem is that phenomena like sexual asceticism and bap tism - and, for that matter, the inheritance of witchcraft, or ensorcellment by magical techniques - arer not simply given to our observation, but can be distinguished only through a process of abstraction. The common features used to develop such categories as baptism, sexual asceticism, ensorcellment by magical techniques, and the inheritance of witchcraft, are derived by analysing them out of the total systems of ideas from which they derive their sense. But ideas cannot be torn out of their context in this way, since their meaning is determined by the role they play in the system of which they form a part. As Winch remarks, It is nonsensical to take several systems of ideas, find an element in each which can be expressed in the same verbal form, and then claim to have discovered an idea which is com mon to all systems. This would be like observing that both the Aristo telian and Galilean systems of me chanics use a notion of force, and concluding that they therefore make use of the same notion.60 In addition to the problem posed by the fact that the traits selected by the anthropologist for the purpose of constructing his ideal types will have a different meaning in each particular society, there is: also the problem that such classificatory systems tend to obscure the? relevance of situation and process. Victor Turner's discussion of the applicability of Middleton and Winter's; attempted dif ferentiation between witchcraft andl sorcery to the facts <s£ Lugbara ethnography brings out something of the signific ance of this. Even although Middleton and Winter here were not (explicitly) attempting to establish this dif ferentiation in terras of ideal types, Turner's remarks are nevertheless still relevant and to the point: •.. Lugbara themselves find it diffi cult to distinguish between deaths brought about by witchcraft and "by 'ghost invocation' by elders against habitually disobediant juniors?. Both are said to be motivated by the sentiment Lugbara call ole.... In a witchcraft context ole m&jr "be trans lated as 'jealousy* (an unrighteous sentiment) and in a ghost-invocation setting as 'righteous indignation*. Middleton's rich case material shows that the same death may be interpret ed by different factions as: one or the other, again according to the structural perspective of the inter preters. The fact, too, that both •witches* and •sorcerers*, in Middleton*s usage, may be called oleu ... a derivative of ole, makes it clear that what is regarded- as: ideo logically important by the Lugbara is belief in the existence of a broad class of persons who can in jure others by mystical means irres pective of motive. It is only in the action-field context that alle gations of this or that specific means are made by interested par ties. Almost every society recog nizes such a wide variety of mystic ally harmful techniques that it may -40-be positively misleading to impose upon them a dichotomous classifi cation. The solution to this problem seems to me to hinge on abandoning any idea that we can define exactly what witch craft, sorcery, or destructive magic are (in the sense of some shared essence), since there need be no set of shared essential attributes which can be abstracted. While we might be able to outline some of the attributes shared by some examples of witchcraft, sorcery and destructive magic, there is no set of characteristics simultaneously applic able to all examples and to them alone. Thus, instead of seeking to establish universal classes, we should recognize that we are dealing with concepts which bear only a family  resemblance to one another. The difference between a class and a family resemblance is brought out in Diagram 2, below. Family Resemblance Class 1 2 3 5 1 2 3 4 5 6 23456 1234 I T E 1 4 5 6 1 4 5 6 g 1 6 12 3 6 2 3 6 1 3 4 6 ATTRIBUTES Diagram 2: Family Resemblance and Universal Class If, in the above diagram, each horizontal line is interpret ed as representing an item, and each number within each line as representing an element of an item, it will be seen that each vertical line represents elements shared in com--41 mon by different items. All the members of the same class share one element in common(1). With the group of items which have been described as bearing a family resemblance to each other, however, there is no single element which is shared by all items, which nevertheless have a good deal in common. A family resemblance, in other words, is constitu ted by a network of overlapping and cross-cutting resenh-blances.^ It is my contention that we will be better served if we conceptualize witchcraft and sorcery as constituting "natural families" rather than true classes, since the at tempt to establish classes to group such beliefs ends up in a priorism and a violation of indigenous structures of thought. While this is less true of the ideal-type mode of classification than of the attempt to establish strict categories, it remains also true of this. There is, indeed, good reason to think that intellectual structures should be little suited to the development of cross-cultural cate gories, in terms of which their contents may be described. This is because the component elements of idea systems bear an internal relationship to each other, and each idea sys tem must therefore be understood in and for itself. It might be argued, against this view, that every science must develop a precise terminology if confusion is to be avoided. Yet even if we accept the dubious methodo logical postulate on which this argument is based - the unity of method of the natural and social sciences - it is invalid. Contrasting nominalist and realist (or "essen-tialist") approaches towards the importance of definitions, -42-Sir Karl Popper has persuasively argued that "in science all the terms really needed must be undefined terms."^ He contrasts the situation, as he sees it, between philosophy and physics. Philosophers, for twenty centuries, have wor ried over the meaning of terms, but their discipline con tinues to be vague, ambiguous and full of verbalisms. Phy sics, on the other hand, hardly concerns itself about terms and meanings, but about facts instead, yet has achieved great precision in its discourse. The conclusion which Popper draws from this is that it is a mistake to .try and make the statements made in a science depend on the meaning of terms. No argument should be based on a definition, for this merely shifts the ques tion of its validity back to its defining terms. The cor rect solution in trying to avoid the imprecision inherent in all language does not therefore lie in trying to specify terms more exactly. Rather, it consists in trying to re main within the limits of vagueness, phrasing sentences in such a way that possible shades of meaning of terms are not important. In Popper's view, the precision of language depends in fact on not trying to make terms too precise. Terms like "sand-dune" and "wind", for example, are very difficult to define exactly (how do we distinguish a dune from a hill, mound or heap; a wind from a gale, breeze or hurricane?). Yet these terms have proved sufficiently precise for many of the purposes of the meteorologist and the geologist, and where a more exact specification is necessary, it is suffi cient to say "dunes, forty to fifty feet high" or "winds of -43-64 a velocity of forty to fifty miles per hour". Anthropologists may also adopt the procedure of the geologist and meteorologist. Instead of trying to formu late terms which exactly describe all cases of witchcraft, sorcery and destructive magic, not confusing these cate gories and also distinguishing them clearly from such other notions as ghost-invocation or cursing, where precision is necessary we may adopt the practice of speaking of "des tructive magic with the following features...", or "witch craft, where the witch is attributed the following charac-65 teristics..."• Notes and References 1. I employ the term "primitive" in reference to societies with the following characteristics: low-level techno logy, non-literacy, low population density, "multiplex" role systems, social relations organized on a face-to-face basis. No implications of inferiority or of "backwardness" are intended, 2. Kluckhohn, C, Navaho Witchcraft, (Beacon Press, Boston, 1967); Evans-Pritchard, E.E., Witchcraft,  Oracles and Magic among the Azande, (Oxford University Press, London, 1937). 3* Or perhaps we should say "socio-structuralism", in order to differentiate Evans-Pritchard*s approach from the linguistico-logical structuralism of writers like Claude Levi-Strauss. 4. Kuhn, T.S., The Structure of Scientific Revolutions, (University of Chicago Press, Chicago, 1962), P. x. 5. See Douglas, M., Natural Symbols: Explorations in  Cosmology, (Barrie & Rocklife, London, 1970); "Intro duction: Thirty Years After Witchcraft, Oracles and  Magic", in Douglas, M., Witchcraft: Confessions and  Accusations, (Tavistock Publications Ltd., London, 1970); Leach, E.R., Rethinking Anthropology, (Athlone Press, London, 1961), Ch. I; Nadel, S.F., "Witchcraft in Pour African Societies: An Essay in Comparison", American Anthropologist, Vol. LIV, (1952), Pp. 18-29, reprinted in Marwick, M. (ed.), Witchcraft and Sorcery, (Penguin Books Ltd., Harmondsworth, 1970), Ch. XXIII; Swanson, G.R., The Birth of the Gods. (University of Michigan Press, Ann Arbor, 1964), Ch. VII. 6. Mair, L., Witchcraft, (McGraw-Hill Book Co., New York, 1969). 7. That the relationship between conceptual categories and manifest behaviour must be posed as problematic, not taken as given, is shown by the phenomena which social -45-scientists have characterized "by such terms as deviance, ideology, rationalization and false consciousness. The nature of this relationship is discussed further in Ch. II of this thesis. 8. Mair, L., Witchcraft, op. cit., P. 165. 9. Wittgenstein, L., Philosophical Investigations, (Basil Blackwell, Oxford, 1953). 10. Leach, E;R., Rethinking Anthropology, op. cit., Pp. 105-8. 11. ibid., P. 27. 12. Le'vi-Strauss, C., Totemism. (Merlin Press, London, 1964), P. 3. 13. Kluckhohn, C, Navaho Witchcraft, op. cit., P. 67. 14. Evans-Pritchard, E.E., Witchcraft..op. cit., P. 64. 15. Winch, P., "Understanding a Primitive Society", Ameri-* can Philosophical Quarterly, Vol. I, No. 4, (Oct. 1964), P. 310. Emphasis in original. 16. Cf. Ch. VIII of this thesis, Section D, "Action Against Witchcraft". 17. Kluckhohn, C, Navaho Witchcraft, op. cit., P. 27. 18. Ibid.. P. 96. I follow here the interpretation offered by Kluckhohn in Navaho Witchcraft. Max Marwick, how ever, reports receiving a personal communication from Kluckhohn containing information not supportive of this interpretation. According to Marwick, Kluckhohn stated in this communication "...that his more recent data had not confirmed his earlier conclusions in this respect, and that gossip about local witches was commoner than his first impressions had led him to believe. I get the impression that his earlier finding may have been based on informants* general statements; and this later one, on the examination of specific instances. It may be -46-that ... people put the blame for misfortunes in gene ral on to distant witches and for a specific misfortune on to someone within the community ...". Marwick, M., "Witchcraft as a Social Strain-Gauge", Australian Jour  nal of Science, Vol. XXVI, (1964), Pp. 263-8, reprinted in Marwick, M. (ed.), Witchcraft and Sorcery, op. cit., Ch. XXIV. Cf. P. 287. 19. Kluckhohn, C, Navaho Witchcraft, op. cit.. P. 111. Again, the reliability of Kluckhohn's information on this point is not clear, since, with the exception of some information on ties of kinship, we are given no concrete data on the relations between witch, victim and accuser. From the information we are given, it would seem that a disproportionate number of those accused of witchcraft, and even more of those who are bewitched, are wealthy. This might suggest that most accusations are made between persons of roughly equal socio-economic status, and that witchcraft is mainly of concern to the wealthier members of Navaho society. Ibid., Pp. 59-60. 20. Evans-Pritchard, E.E., Witchcraft..., op. cit., Pp. 104-5, 343-4. 21. Kluckhohn, C, Navaho Witchcraft, op. cit., P. 85. 22. I have drawn here on The Shorter Oxford English Dic  tionary, (Oxford University Press, London, 1959). 23. Kluckhohn, C, Navaho Witchcraft, op. cit.. P. 45. Em phasis in original. 24. Ibid., P. 133, note 3. 25. In the following section, I have drawn heavily on the exposition and critique of Pareto by Winch, P., The  Idea of a Social Science, (Routledge & Kegan Paul, Lon don, 1958), espc. Pp. 95-103. Cf. also, Pareto, V., The Mind and Society (Harcourt Brace, New York, 1935). 26. Winch, P., The Idea..., op. cit., Pp. 99-100, Emphasis in original. -47-27. This point will be developed more fully in later sections of this thesis. 28. Nadel, S.F., "Malinowski on Magic and Religion", in Firth, R. (ed.), Man and Culture. (Routledge & Kegan Paul, London, 1957), P. 196. Cf. Malinowski, B., Argonauts of the Western Pacific, (E.P. Dutton & Co., Inc., New York, 1961), P. 424. 29. Malinowski, B., Argonauts..., op. cit., P. 299. 30. Ibid.. 31. This despite the fact that, at the time of writing Argonauts of the Western Pacific, Malinowski continued to accept the validity of Frazer's theories of magic and religion. Ibid., P. 73, footnote. 32. Despite, however, the fact that the powers of the yoyova are believed to derive from magic, they seem also to involve an innate component. Thus the yoyova is believed to develop a small egg-shaped kapuwana in her body which at night assumes the various forms in which the mulukwausi appears. Whether this kapuwana is a physical substance or not seems unclear, and in Malinowski's view to ask such a question is to attempt to smuggle in our own categories where they do not exist. It is also believed that young yoyova are easily picked out from other girls by their crude tastes, and especially their liking for raw pork and fish. Ibid., Pp. 238-40. 33. Ibid., Pp. 73-5. 34. Malinowski, B., Crime and Custom in Savage Society, (Routledge & Kegan Paul, London, 1926), Pp. 85-94. 35. Malinowski, B., Argonauts..., op. cit., Pp. 76, 236-41. 36. Fortune, R., Sorcerers of Dobu, (Routledge & Kegan Paul, London, 1963). 37. Cf. Firth, R., Human Types, (Mentor, New York, 1958), -48-Ch. VIj "The Sociology of »Magic»", Sooiologus, n.s., Vol. IV, (1954), Pp. 97-116, reprinted in Firth, R., Tikopia Ritual and Belief,- (Allen & Unwin, London, 1967), Ch. IX; Hogbin, H.I., "Sorcery and Administra tion", Oceania, Vol. VI, (1935), Pp. 1-32. 38. Marwick, M., "The Study of Witchcraft", in Epstein, A.L. (ed.), The Craft of Social Anthropology, (Tavistock Publications Ltd., London, 1967), P. 233. 39. Evans-Pritchard, E.E., Witchcraft..., op. cit., P. 8. 40. Ibid., Pp. 21-3. 41. Ibid., P. 23. 42. Ibid., P. 112. 43. Evans-Pritchard, E.E., "Witchcraft (Mangu) Amongst the Azande", Sudan Notes and Records, Vol. XII, (1929), Pp. 163-249, reprinted in Marwick, M. (ed.), Witch  craft..., op. cit., Ch. II. Cf. Pp. 29-30. 44. Evans-Pritchard, E.E., Witchcraft..., op. cit., P. 109. 45. Ibid.. P. 119. 46. Ibid., P. 114. 47. Evans-Pritchard, E.E., "Witchcraft (Mangu)...", op.  cit., P. 30. 48. Evans-Pritchard, E.E., Witchcraft..., op. cit., P. 392. 49. Ibid., Pp. 394-8. 50. Kluckhohn, C, Navaho Witchcraft, op. cit., Pp. 22-4. 51. Ibid., Pp. 25-7. 52. Ibid., Pp. 31-3. 53. Ibid., Pp. 34-5. -49-54. Ibid., Pp. 40-1. In this brief outline of the cate gories of Navaho witchcraft, I have not given the Navaho terms for Frenzy Witchcraft, Sorcery, etc., because of the difficulty of reproducing the phonetic transcription used by Kluckhohn. 55. Middleton, J. and Winter, E.H., "Introduction" to Middleton, J. and Winter, E.H. (eds.), Witchcraft and  Sorcery in East Africa, (Routledge & Kegan Paul, London, 1963), P. 3. 56. An exception is Crawford, J.R., Witchcraft and Sorcery  in Rhodesia, (Oxford University Press, London, 1967). 57. Cf. Mair, L., Witchcraft, op. cit., P. 23; Marwick, M., "Introduction" to Marwick, M. (ed.), Witchcraft..., op. cit., Pp. 12-13. 58. Quoted in Turner, V., "Witchcraft and Sorcery: Taxonomy versus Dynamics", Africa, Vol. XXXLT, No. 4, (1964), Pp. 314-25. Cf. Pp. 322-3. Turner's emphasis. 59. Goode, W.J., Religion Among the Primitives, (Free Press, Glencoe, 111., 1951), P. 52. 60. Winch, P., The Idea..op. cit., P. 107. See also Pp. 105-6. 61. Turner, V., "Witchcraft...", op. cit., P. 323. 62. Wittgenstein, L., Philosophical..., op. cit., Pp. 31-6. 63. Popper, K.R., The Open Society and its Enemies, (Routledge & Kegan Paul, London, 1961), Vol. II, P. 18. 64. Ibid.. Vol. II, Pp. 10-19. 65. As far as my own use of terminology is concerned, I propose, for the purposes of this thesis, to use such terms as "witchcraft", "sorcery" and "magic" in the same sense as they are used by the particular theore ticians being discussed. CHAPTER TWO FURTHER METHODOLOGICAL CONSIDERATIONS A. Introduction In the preceding chapter, we have suggested that the principal task of the anthropologist lies in trying to ren der intelligible, to a particular audience, the way of life of another cultural group.1 In this chapter, I propose to develop further some of the implications of this view and to elaborate on some of the methodological considerations already invoked in our discussion of problems of termino logy and definition. Fundamental to the understanding of cultural life is the recognition of its nature as meaningful behaviour. An other way of expressing the same idea is to say that cul tural behaviour takes place within the context provided by a set of concepts and beliefs which are interrelated in such a way as to form a system. These systems are struc tured according to certain rules, and it follows from this 2 that cultural behaviour is rule-governed. In order to un derstand cultural behaviour, therefore, it is first neces sary to succeed in grasping the rules which order it and the notions which enter into it. It is for this reason that, in the preceding chapter, we have several times cri ticized suggested terminological systems for violating the integrity of native systems of thought. This is not to imply that a satisfactory understanding of the way of life of a particular social group is neces sarily contained in the way in which its own members con ceptualize it. Many of the rules governing cultural life may not be actually present in the consciousness of those who follow them: many linguistic conventions, for example, are unconsciously followed in this manner. Moreover, even when cultural rules are consciously applied, those apply ing them may not be aware of the full consequences of their doing so (just as a chess player may not forsee the full consequences of a move that he makes). Por cultural rules are not applied in a vacuum, but in the context of particular action-situations, and the result of their ap plication in the context of these particular situations is to generate new situations which may or may not be those which were originally intended. It follows from this that the nature of a particular way of life, considered as an on-going process, may not be fully understood by those living it, and therefore that current conceptualizations of social reality may embody falsehoods.^ Among the tasks of the anthropologist, is that of determining the extent of these misconceptions, and of discovering what role they play, as misconceptions, in the life of the social group concerned. The conceptualization of social reality is, of course, only a special case of the understanding of reality as such, and it is not only the former, but also the latter, which enters into a given way of life. As Winch expresses -52-A man's social relations with his fellows are permeated with his ideas about reality. Indeed, "permeated" is hardly a strong enough word: social relations are expressions of ideas about reality A In support of this argument, Winch gives the example of the characteristic social relations which a monk has with his fellow monks and with people outside the monastery: it would be impossible to achieve more than a superficial understanding of these relations without taking into ac count the religious ideas on which the monk's life is 5 centred. Other examples of the way in which notions of social, psychic, spiritual, biological or physico-chemical reality govern people's characteristic activities, includ ing their relations with their fellows, spring readily to mind. One might cite the way in which the characteristic behaviour of the vegetarian, the racist, the alchemist or the devotee of astrology or of the I Ching, manifests his or her particular notions of reality; but the example of the meat-eater, the liberal, the nuclear physicist, or the rationalist, would be equally valid. Of course, just as it is possible that a social consciousness may be a false consciousness, so too is it possible that an actor may not fully comprehend the way in which his social interaction network, and his way of life in general, is founded on im plicit or explicit notions of other dimensions of reality. B. Cultural Knowledge as a Cultural Phenomenon A particular image of society, and a particular view of reality, is acquired by the individual from his life in -53-society, just as a language is so acquired. It is in terms of this social consciousness, and this world-view, that the individual makes sense of his experiential world, including his universe of social interaction. This statement holds equally true for the anthropo logist as for the members of the cultures which he studies. The anthropologist does not approach the phenomena of other cultures without expectations, but does so with de finite assumptions in mind which are a product of his bio graphy in a specific socio-cultural milieu. No less than the shaman or the witch-doctor, the anthropologist has a particular social identity and a particular conceptualiza tion of this identity, of his society, of society and hu manity in general, and even of reality, be these views elaborated into a consistent philosophy.or present only as an unintegrated series of presuppositions. Most commonly, this world-view takes the form of some variant or other of a secularist liberal humanism compatible with a techno cratic social order.6 Such views deeply influence the anthropologist's analysis of other societies. As Pocock points out, even in his first piece of field-work the anthropologist neces sarily compares the categories of his own society with those of the society he is studying, and also has in mind the works of his predecessors dealing with phenomena com-7 parable to those which he finds. The anthropologist's experience of primitive society, in other words, is a function of his prior experience of his own society. In this way, the position of the anthropologist may be re--54-garded as being in essence identical to that of the his torian interpreting another historical epoch. As the his torian's work is itself the product of an historical evolu tion, so is that of the anthropologist a product of a par ticular cultural configuration. Thus, as Paul Cardan writes, ... each civilization or epoch, from the very fact that it is particular and dominated by its own obsessions, is led to suggest or uncover new meanings in the societies which pre ceded it or surround it.... These meanings can never fix or exhaust their object, not the least reason for which is that they themselves sooner or later become objects of in terpre tation.$ The following provides a clear example of the way in which social and historical thought is itself socio-historically determined: In the thinking of the ancient Greeks the dominant categories de fining social relations and history were essentially political (the power of the city, relations between cities, relations between "might" and "right", etc.). The economy only received marginal attention. This was not because the intelli gence or insight of the Greeks were less "developed" than those of mo dern man. Nor was it because there were no economic facts, or because economic facts were totally ignored. It was because in the social reality of that particular epoch the economy had not yet become a separate, auto--55-nomous factor (a factor "for itself" as Marx would say) in human develop ment. A significant analysis of the economy and of its importance for society could only take place in the 17th century and more particularly in the 18th century. It could only take place in parallel with the real development of capitalism which made of the economy the dominant element in social life. The central import ance attributed by Marx and the marxists to economic factors is but an aspect of the unfolding of this historical reality.9 C. Levels of Cultural Understanding There are two separate stages involved in the under standing of another culture. The first involves assimi lating the categories of native thought and feeling, to gether with a knowledge of the contexts in which such cate gories are employed: a process which might be compared to that of learning a foreign language. The second stage in volves going beyond these categories, in some sense, in order to indicate their significance to anthropological theory: a process which might be compared to the scienti fic description of a language. Another possible analogy to anthropological understanding is that of psychoanalysis. As Pocock argues, The analyst enters the private world of his subject in order to learn the grammar of his private language. If the analysis goes no further it is no different in kind from the under standing which may exist between any two people who know each other well. -56-It becomes scientific to the extent that the private language of inti mate understanding is translated in to a public language ... in this case ... the language of psycholo gists. But the particular act of translation does not distort the private experience of the subject and ideally it is, at least poten tially, acceptable to him as a scientific presentation of it."*^ Obviously the task of understanding the beliefs and practices of the members of another society is not equally difficult on all levels of socio-cultural organization. The relative difficulty of understanding is primarily de termined by the degree of similarity or difference of the beliefs and practices being studied to those operative in the anthropologist's own society. Technological behaviour, for example, poses no particular problems of understanding since, qua technological behaviour, it is readily assimil able to the categories of our own world-view and satisfies our own criteria of rationality. Nobody is puzzled by the fact that, in order to grow crops, seeds are planted in the ground and watered, and the garden carefully tended and weeded. Once the goal of the gardener is known - the growing of crops - and the limitations of his resources and gardening knowledge comprehended, his behaviour is im mediately intelligible to us. More than this, simply by observing technological behaviour - in its more elementary forms at least - it is often possible to infer not only the goals of the technologist, but also the limitations of the resources and technical knowledge at his disposal. We -57-might recognize that there are other ways of accomplishing the same end, but his way is also effective. Thus, even if the native observer explains the efficacy of his tech nological activities using different terms from those which we would use, these activities nevertheless in no way challenge our conceptualization of the nature of reality. The problem begins when we are forced to consider the magical rituals so often associated with what we describe as technology: the spell that is whispered over the digging-stick, the "medicines" that are used to promote growth or ward-off thieves. Are such practices to be ex plained simply as misguided technological procedures, the result of errors in the understanding of natural causation, or is there some other explanation? A similar difficulty confronts us when we consider Navaho tales of witches who enclose themselves in gourds and travel by rapidly rolling along the ground, or of other witches who can transform themselves into wolves, coyotes and bears and move at great speeds, or of dogs which may pray sickness on 11 people. Por such beliefs seem to assert the existence of powers, substances and relationships that, for us, have no reality. How, then, are we to make sense of them? Such beliefs - and the activities associated with them - constitute a special problem in that we belong to a culture whose conception of rationality is deeply affected by the concepts and methods of the natural sciences. As a result, we tend to treat such things as a belief in magic, witchcraft, or the power of oracles as a paradigm of the -58-irrational. Peter Winch points out the implications of this: The strains inherent in this situa tion are very likely to lead the an thropologist to adopt the following posture: We know that Zande beliefs in the influence of witchcraft, the efficacy of magic medicines, the role of oracles in revealing what is going on and what is going to happen, are mistaken, illusory. Scientific methods of investigation have shown conclusively that there are no such relations of cause and effect such as are implied by these beliefs and practices. All we can do then is to show how such a system of mistaken beliefs and inefficacious practices can maintain itself in the face of objections that seem to us so ob vious .12 One does not have to search far to find evidence which confirms Winch's statement. From the time of Tylor and Frazer until the present, most anthropologists have seen one of the key questions to which they must address themselves as being that of explaining why such erroneous beliefs should continue to persist. In Marwick's words, "Anthropologists have long been preoccupied with the prob lem why beliefs in magic, witchcraft and sorcery, though 1 palpably false, nevertheless continue to have influence." For Marwick himself, witchcraft and sorcery beliefs are 14 "standardized delusions", while for Monica Hunter Wilson 15 they are "the standardized nightmare of a group". In the same vein, Victor Turner describes witchcraft and sor--59-16 eery as "ugly and irrational beliefs", while Kluckhohn similarly opposes witchcraft beliefs to "more rational 17 modes of explanation". Por his part, Evans-Pritchard asks why the Azande "do not perceive the futility of their 18 magic", and why "common sense does not triumph over 19 superstition". In thus viewing magico-religious systems, anthropolo gists have implicitly accepted a hierarchy of types of thought, with western science placed in a position of superiority to all other modes, and serving as a touch stone against which to judge them. Hence the long standing controversy over the question of the extent to which the magico-religious beliefs of primitive peoples can be accomodated to models of scientific thought and be haviour. This, in itself, is a perfectly justified in terest. The trouble is, as Barnes remarks, that "Often, in such material, one finds a detailed and sophisticated treatment of anthropological material related to sketchy and often implicit notions of what is rational or scienti-fio".20 D. Their Ways of Thinking and Ours Perhaps one of the clearest examples of this tendency to operate with largely implicit notions of the nature of western thought and of scientific discourse is that of Le*vy-Bruhl. Thus, despite his explicitly declared inten tion to make a comparative study of primitive and western thought, L^vy-Bruhl did not attempt any serious examina tion of the latter, contenting himself with the statement -60-that it was already sufficiently well defined in the works of philosophers, logicians and psychologists, and therefore 21 not in need of further elaboration. Other anthropolo gists have assumed as much, if they have not been quite as explicit in saying so. Such an attitude is insupportable. No anthropologist would accept the conscious model of another society as an accurate description of reality, and there would seem to be no more justification for doing this with the conscious models of our own society. The important point is that, on the basis of these sketchy and implicit notions, a series of oppositions have been postulated between Western science and other cosmological systems. We may indicate some of these in tabular form: THEORIST MAGIC SCIENCE Tylor, Frazer subjective objective Ldvy-Bruhl prelogical & mystical logical & empirical Malinowski emotive intellectual Evans-Pritchard mystical empirical Kluckhohn fantasy-oriented reality-oriented Diagram 3: Oppositions Between Science and Magic In discussing another aspect of human thought - the manner in which the interrelation between nature and cul ture is conceptualized - Levi-Strauss has noted that the -61-mind of the anthropologist has played as large a part as have the ideas of the people being studied, in determining what the anthropologist has written: "... it is as though he were seeking consciously or unconsciously, and under the guise of scientific objectivity, to make the latter 22 ... more different than they really are". He suggests in explanation of this that: In order to place the modes of thought of the normal, white adult man on a firm foundation and simul taneously to maintain them in their integrity, nothing could be more convenient than for him to separate from himself those customs and be liefs, actually extremely hetero geneous and difficult to isolate, around which had crystallized an inert mass of ideas which would have been less offensive if it had been necessary to recognize their presence and their action in all cultures, including our own. Totemism is firstly the projection outside our own universe, as though by a kind of exorcism, of mental attitudes incom patible with the exigency of a dis continuity between man and nature which Christian thought has held to be essential. It was thus thought possible to validate this belief by making the universal exigency an at tribute of this "second nature"j which civilized man, in the vain hope of escaping from himself as well as from nature, concocts from the "primitive" or "archaic" stages of his own development.^3 In this way, "totemism" has been invented, and a similar -62-process has operated in relation to the concepts of "sacri fice" and of "ignorance of physiological paternity", both of which concepts have, significantly, been assimilated to 24 that of "totemism". The notion of "ignorance of physiological paternity", widely ascribed to the Trobriand Islanders and the Aborigines of Australia, has recently been examined in de tail by Edmund Leach. Reviewing the evidence on this point, Leach concludes that the great bulk of ethnographic data indicates that these peoples are not in fact ignorant of physiological paternity and, moreover, has always indicated  this. This, of course, raises the rather interesting question of why it is that anthropologists have clung so tenaciously to the opposite conclusion. Leach notes that the emphasis on this point for primitive peoples has been coupled with a complete failure to attempt any form of comparative analysis which would embrace the theology of Judaism or Christianity. Instead of seeking for relation ships between primitive and civilized modes of thought, anthropologists have been content to establish a simple  dichotomy between primitive ignorance and civilized theo logy, implicitly denying that there is any affinity be tween accounts of conception among primitive peoples and, 25 for example, the Christian myth of the Virgin Birth. Again we encounter "... the fact that the quest for the ultimate primitive who is quite different from civilized 26 man appeals very strongly to certain anthropologists," thus corroborating Levi-Strauss* analysis of the "totemic illusion". -63-Uhdoubtedly, this process is also part of the ten dency on the part of every civilization "... to over-27 estimate the objective orientation of its own thought." Yet how is such a tendency to be countered? It is useful at this point to turn to Lucien Goldmann's discussion of the Marxian thesis of the class determina tion of consciousness. In discussing the problem posed for objectivity in the social sciences by this thesis, Goldmann has suggested that it is necessary to recognize that certain aspects of reality may be visible from a reactionary standpoint, and yet incomprehensible within the context of the potential consciousness of a more pro gressive class. Recognition of this is the first require ment of the individual researcher, in order for him to be able to transcend the actual consciousness of any existing class. He can do this by: (a) effecting a synthesis of the elements of truth provided by the perspectives of the differing social classes; and (b) by preserving the elements of under standing already expressed earlier by this or that thinker but later abandoned under the influence of social, economic and political changes.28 Of course, no such synthesis could claim finality in any sense. All syntheses must necessarily proceed from a given perspective and, as constructions, necessarily rest on certain preconceptions and underlying assumptions. Yet a generalization of Goldmann's programme contains a very important suggestion for the present study. Primitive -64-modes of thought and notions of causality and reality are no,t only interesting in the light of our own world-views. These conceptions may themselves be used in order to un derstand better our own ways of thought, including the most advanced and sophisticated of scientific theories. A 29 few writers have seen this, but they are exceptional in this respect. The majority of anthropologists, if they have sought insight into their own cultures from the data of primitive magic and witchcraft, have limited their generalizations to a few scattered remarks on such un savoury phenomena as racism, Fascism, Stalinism or McCarthyism. In this respect, they have not really ad vanced beyond the level of insight achieved by literary figures like Arthur Miller.^ By restricting their com parisons to phenomena from their own culture which they themselves evaluate negatively, perhaps with the conscious or unconscious intention of discrediting these phenomena, contemporary anthropologists are engaged in the same kind of pursuit as Tylor and Frazer. They continue to empha size the gulf between primitive and western modes of thought, and restrict an analysis of similarities between the two to those elements of their own culture which they regard as foreign to their own outlook. E. The Importance of History G-oldmann's suggested method of transcending socio-centrism implies the necessity of a constant examination and re-examination of received ideas. An historical ap proach is therefore required for the social sciences. -65-This is more than a matter of antiquarianism or literary adornment. As Pocock explains for anthropology, The discipline as it is today con  tains its history to a remarkable degree. To put it in another way, the subject is still young, is still in the process of working out a con sensus of ideas, and divergences of assumption are perhaps more marked than they are in the longer estab lished sciences.31 History therefore holds a central place in discus sions of anthropological theory. Perhaps many anthropolo gists would dispute this. Even as historically aware a writer as E.E. Evans-Pritchard, for example, is prepared to dismiss many of the theoreticians discussed in the fol lowing pages - Tylor, Frazer, Rivers, Malinowski and Le"vy-Bruhl - as being intrinsically unimportant in contemporary theoretical discussions and mainly of interest as speci-32 mens of the thought of their time. In the face of such a rejection, one might wonder at Evans-Pritchard1s own propensity to adopt an approach to theoretical matters so typically of the "history of anthropology" type. Moreover, even if most contemporary anthropologists reject (or think they reject) the theories of their predecessors, it is not difficult to demonstrate the continuing influence of such theories in related disciplines and among laymen. The abridged edition of The Golden Bough continues to enjoy wide sales in paperback form, and numerous of the ideas of Frazer and Robertson-Smith continue to find acceptance on the part of students of the Old Testament and of Classics, -66-not to mention auch influential modern writers as Mircea Eliade and Ernst Cassirer.^ In any case, how profound is this rejection? Probably, to take a simple example, no contemporary anthropologist would accept Prazer*s model of an evolutionary progression from magic, through religion, to science, without serious reservations. Yet almost as many might accept Prazer's view that the interesting ques tion concerning primitive magico-religious systems centres on the persistence of false beliefs in magical efficacy. Moreover, even if contemporary anthropologists reject a theory, does this mean that we may dismiss it as dead wood? Not even the progress of the natural sciences, still less of the social sciences, can be adequately de picted in terms of a logical progression in the course of which truth unfailingly comes to replace falsehood. Thus, when two philosophers of the social sciences declare them selves to be "Prazerians" and set out to defend Prazer's general methodology, while at the same time criticizing a prominent contemporary anthropologist (John Beattie) for implicitly accepting "one of the weakest points of 34 Prazer's theory", it becomes plain that the old issues are far from having been settled once and for all. In deed, one might wonder if they have even been understood. An historical examination of these issues may therefore be of considerable value, and it is in fact only by means of such an examination that we can hope to render explicit our preconceptions and attempt to relativize our cate gories . F. The Development of Anthropological Theory Anthropology had its roots in the humanistic philo sophical speculation of the Enlightenment, and crystal lized as an academic discipline in the latter part of the nineteenth century. Prom the philosophes of the Enlightenment it derived a strong anti-clericalism, and a mission to free men's minds from the bonds of ignorance and superstition. This anti-religious bias was further strengthened by the bitter struggles occasioned by the theory of natural evolution, and also by the strong heri tage derived from Comtian positivism. The latter part of the nineteenth century was also, of course, the last period of frenzied empire-building on the part of the European powers: the age of the "scramble for Africa" and the "white man's burden". If the philosophes spun idealized images of Chinese civilization and of the good life in Polynesia, the generation of Tylor and Prazer encountered the reality of a weak and decadent China and the undisputed technological and military supre macy of the nations of Western Europe and North America. Most anthropologists of this period did not question the superiority of Western European civilization to any other social type existing on earth up to and including that point in time. Many also accepted the notion of the in nate superiority of the white race to all other ethnic groups. This conviction of superiority, whether racial or cultural, was reflected in the grand evolutionary schemes then so fashionable. Always these schemes culminated in -68-some institution selected from nineteenth century indus trial society: the monogamous family, private property, monotheism, or scientific thought. In relation to such institutions, all the customs, institutions and beliefs of "savages" and "barbarians" were to be regarded as so many fossilized antecedents, retarded in their development and preserved by some freak of history. Nineteenth century anthropologists therefore shared two important characteristics: firstly, the assumption of the superiority of their society to all others and, second ly, a profound antipathy towards religion. This antipathy was partly expressed in the tendency to elevate scientific methodology to a position of supreme judgement in relation to all other epistemological systems. Often these ten dencies merged, as in the writings of Sir James Prazer. In Prazer*s eyes, religion came to be seen as an histori cal survival in nineteenth century European society, a product of fallacious reasoning, and a phenomenon to be replaced in the course of social evolution by the gradual expansion of scientific knowledge. Comte had asserted much the same thing when he outlined his "Law of the Three Stages", according to which human thought had undergone an evolution from religion, through metaphysics, to its final "positive" form. Such a perspective naturally implied that the anthro pologist was looking upon primitive society from the out side. There was for him nothing to be learnt from the way of life of primitive peoples which might be relevant to his own way of life, nor any implication of the validity -69-of attempting to understand these societies in their own terms: least of all with respect to their pernicious and delusory magico-religious beliefs. These constituted only bizarre departures from rational thought, departures standing in need of some form of explanation. Such an explanation was provided, in the first instance, by the theoretical schema of associational psychology, L£vy-Bruhl did not basically modify this opposition between primitive and civilized, nor this denigration of native modes of thought. Nor did he challenge the ten dency to elevate Western science to a position of absolute arbiter or touchstone against which to measure the accept ability of other ways of conceptualizing reality. Indeed, his definition of mystic thought, as thought presupposing the existence of entities and relations not existing in reality, assumes precisely this conception of science. Furthermore, if L£vy-Bruhl, in comparison to Tylor and Frazer, marks a forward step in beginning to appreciate the variability of meaning in differing cultural contexts, in other respects he may be considered to mark a retro grade step. Both Tylor and Frazer, although tending to view the primitive as a credulous fool and slave to cus tom, nevertheless strongly emphasized that the thought processes of primitives were to be understood in basically the same terms as our own. L£vy-Bruhl denied this com pletely. However coherent native thought is, it cannot be understood by Western civilized man. Thus the anthropolo gist may describe what primitives say, but he cannot hope to grasp their concepts. Por primitives do not possess concepts proper in the sense of recognizing that some uses conform to, and others break, rules for the use of expressions. Hence the indifference of primitive thought to logical contradiction. Of course the Westerner might imagine himself a primitive and obtain some sort of under standing of primitive ideas by means of a process of em pathy. Equally well, however, he might imagine what it is to be a bear or a squirrel by means of some process of em pathy. With Malinowski, a dramatic shift occurred in anthro pological emphasis. Writers now became less interested in trying to understand primitive thought as such, and more interested in the thought of particular primitives. Pield-work suddenly assumed a crucial importance for all anthro pologists: a marked difference from the attitude towards this of, for example, Prazer. Yet, obviously, this field-work did not proceed in a theoretical vacuum, but was car ried out under the constraining influence of pre-existing ideas. The vast majority of anthropologists, for instance, continued to accept the imperial situation: if for no other reason than that there did.not seem to be any prac tical way of dismantling the empires at this time. This is not to suggest that they acted as cynical advisers to rapacious imperialist administrators. Rather, their atti tudes are more accurately described by G-ough: ... anthropologists in those days seem to have commonly played roles characteristic of white liberals, sometimes of white liberal reformers -71-... Anthropologists were of higher social status than their informants; they were usually of the dominant race, and they were protected by im perial law; yet, living closely with native peoples, they tended to take their part and protect them against the worst forms of imperialist ex ploitation. Customary relations de veloped between anthropologists and the people whose institutions they studied. Applied anthropology came into being as a kind of social work and community development effort for non-white peoples, whose future was seen in terms of gradual education and amelioration of conditions many of which had actually been imposed by their Western conquerors in the first place. 36 Thus Malinowski was to describe anthropology as having not only a scientific and cultural value, but as having a direct practical application in helping the white man to "... govern, exploit, and 'improve1 the native with less 37 pernicious results to the latter." This new anthropological role induced a change to wards native conceptual systems. Earlier anthropologists had tended to regard primitive beliefs in magic and witch craft as unqualifiedly pernicious, and as better destroyed than perpetuated. Malinowski, on the other hand, argued that such beliefs, even if fallacious, should be tampered with as little as possible by administrators, missionaries and traders, since they fulfilled a pragmatically useful function in primitive society. Hence Malinowski's empha sis on the role played by sorcery in upholding the politi--72-cal authority of the Trobriand chief. In a similar vein, Rbgbin and Fortune sought to demonstrate that, while the eradication of the belief in sorcery was a worthwhile goal to pursue, the welfare of native society was not best served by the direct interference on the part of govern-ment officials with these beliefs. This view was, of course, associated with the func tionalist analysis of social life, according to which every item belonging to a cultural complex necessarily made a contribution to the continuation of the social group considered as a perduring unit. This approach was dominant in British anthropology from the 1920's until re cent times, and also had some impact on American cultural anthropological studies. Thus it was that Kluckhohn un dertook an analysis of the witchcraft beliefs of the colo-nialized Navaho from a perspective in many respects iden tical to that of the functionalism of Malinowski and Radcliffe-Brown. Indeed, his chief difference from the latter writers lies in his far greater emphasis, especial ly when compared with Radcliffe-Brown, on the importance of the individual as the fundamental unit of functionalist analysis: an emphasis perhaps relateable to the greater disintegration of the native cultures of the United States compared to the relatively more intact cultures studied by British anthropologists. Coupled with this emphasis, was Kluckhohn's far greater sophistication of psychological theory compared with Malinowski. A further change of emphasis is marked by the publi cation of E.E. Evans-Pritchard's Witchcraft, Oracles and -73-Magic among the Azande: probably the most important single study of primitive belief in magic and witchcraft to date. In this work, there is little talk of the functions of social institutions, and the phenomena under examination are not explained by the extent to which they make for sentiments reinforcing the solidarity of Zande society. The concern is rather to show how Zande beliefs and ac tions are interrelated and how, given the premises on which they are based, they form a logical system. Nowhere in the book is there anything resembling a law of human society, or even anything approaching a statement about witchcraft in general. The analysis is concerned rather with presenting the meaning of Zande beliefs and actions, and some understanding of how these beliefs accord with the formal social organization of the Azande. There emerges an implicit comparison between their witchcraft and our no tions of belief, causality, and moral system, and also a heightened consciousness of what we ourselves mean by these terms.... It is impor tant to note that by this stage the individual institution - in this case, witchcraft - is only the point of entry to the perception of sets of relations. In short, one can be gin to speak of the structural ana lysis of social life as opposed to the functional analysis of social structures. 39 G. Evans-Pritchard and Kluckhohn Kluckhohn*s work therefore represents the culmination of a theoretical tendency initiated - first and foremost --74-by Malinowski, while the approach of Evans-Pritchard re presents the initiation of a new direction of anthropolo gical analysis, and one which has become increasingly sig nificant in recent years. Since these two studies were published within seven years of each other, and - somewhat curiously - written independently of each other, it would seem a fruitful project to undertake a comparative analysis of these as models for the understanding of primitive ideas of witchcraft. In Kluckhohn's work, we see the functionalism of Malinowski, greatly refined, attempting to make sense of that field of data most difficult to un derstand in terms of Malinowski*s general theory of magic, and most problematic for his functionalist approach as a whole: i.e., witch beliefs. In Evans-Pritchard*s work, we see Malinowski*s perspective already implicitly rejected (hence the hostility of the latter towards it*^), and the initiation of a new approach to the understanding of primitive magic and witchcraft beliefs. It is possible to contrast the work of Evans-Pritchard and of Kluckhohn on a number of dimensions. Thus the for mer is concerned with meaning, the latter with function; Evans-Pritchard is attempting to understand the Zande world-view, while Kluckhohn is trying to explain that of 41 the Navaho; Evans-Pritchard places his main emphasis on the rationality and logical coherence of Zande beliefs, even although he considers these to be objectively false, while Kluckhohn treats Navaho witch beliefs as fundamen tally irrational and essentially to be understood as the -75-resultants of the unconscious processes of the human psyche. To all these differences, one must add the overriding difference between KLuckhohn's psychologism and Evans-Pritchard' s sociologism. This, to be sure, is nothing new in terms of an antithesis in the social sciences: nor is it an antithesis that has by any means been resolved. Pocock notes that the question of the priority of psyche or society was already an issue of debate between Macaulay 43 and James Mill in 1829; it continues to be a point of contention in the most recent of discussions in the social 44 sciences. It seems unlikely that this question should ever be resolved, since it cannot be decided by arguments on an abstract methodological level. The question of the priority of psyche or society is closely bound to one's prior conception of the type of question the social sciences should attempt to answer, which in turn hinges on one's actual ideas concerning the nature of human society and personality. Prom the standpoint of this thesis, what is interest ing in the psychological model is the substantive theory and the implicit assumptions with which it is associated. Por the psychological model has not been invoked in order to explain rational thought processes, but only those which are considered to depart from our criteria of ra tionality. The sociological approach, on the other hand, has tended to look for criteria of rationality within each culture, and not to assume that what would be irrational -76-in our society need necessarily be so in another cultural context (and vice versa). It follows that the sociologi cal approach - at least in its structuralist variant - is much more concerned with seeing a culture in its own terms. The psychological approach, in contrast, definitely involves the analysis of other cultures in terms extra neous to them. Notes and References 1. Ch. I, P. 3. 2. The meaning of the statement that all specifically human behaviour is rule-governed is explicated in Winch, P., The Idea of a Social Science. (Routledge & Kegan Paul, London, 1958), espc. Pp. 24-33, 40-65. Winch suggests that an action may be spoken of as in volving the application of a rule if it is possible to distinguish between a right and a wrong way of per forming the action. In his critique of Winch, D.R. Bell argues that there are certain obscurities in herent in this formulation of Winch's, and contends that it is necessary to distinguish at least two senses in which an activity may be rule-governed: (i) in the sense that an activity is conceptualized, this particular conceptualization fitting in with other concepts of the language; and (ii) in the sense in which a rule applies forbidding or enjoining an activity for some reason. Bell, D.R., "The Idea of a Social Science", Aristotelian Society Supplement. Vol. XLI, (1967), P. 119. See also the critique by Maclntyre of Winch's position: Maclntyre, A., "The Idea of a Social Science", Aristotelian Society  Supplement. Vol. XLI, (1967), Pp. 95-114. 3. Recognition of this point is, of course, widespread in the social sciences, implicitly or explicitly forming in part the theoretical base of such divergent metho dological approaches as those of Preudianism, Marxism, functionalism and structuralism. It is expressed in such notions as those of false consciousness, manifest function, rationalization, conscious model and ideo logy. 4. Winch, P., The Idea.... op. cit.. P. 23. 5. Ibid.. 6. The concept of technocracy is developed in Roszak, T., -78-The Making of a Counter Culture, (Doubleday & Co., Inc., Garden City, New York, 1969). 7. Pocock, D.F., Social Anthropology, (Sheed & Ward, London, 1961), P. 90. 8. Cardan, P., History and Revolution, (Solidarity, London, 1971), P. 28. 9. Cardan, P., The Fate of Marxism, (Solidarity, Clydeside, n.d.), P. 4. 10. Pocock, D.F., Social Anthropology, op. cit., Pp. 88-9. 11. Kluckhohn, C, Navaho Witchcraft, (Beacon Press, Boston, 1967), Pp. 26-7, 61, 138-48, 150 note 7. 12. Winch, P., "Understanding a Primitive Society", American Philosophical Quarterly, Vol. I, No. 4, (Oct. 1964), P. 307. Emphasis in original. 13. Marwick, M. (ed.), Witchcraft and Sorcery, (Penguin Books Ltd., Harmondsworth, 1970), P. 319. 14. Marwick, M., "The Study of Witchcraft", in Epstein, A.L. (ed.), The Craft of Social Anthropology, (Tavistock Publications Ltd., London, 1967), P. 238. 15. Wilson, M.H., "Witch Beliefs and Social Structure", American Journal of Sociology, Vol. LVT, No. 4, (Jan. 1951), P. 313-16. Turner, V., "Witchcraft and Sorcery: Taxonomy versus Dynamics", Africa, Vol. XXXIV, No. 4, (1964), P. 316. 17. Kluckhohn, C, Navaho Witchcraft, op. cit., P. 83. 18. Evans-Pritchard, E.E., Witchcraft, Oracles and Magic  among the Azande, (Oxford University Press, London, 1937), P. 475. 19. Ibid., P. 193. -79-20. Barnes, S.B., "Paradigms - Scientific and Social", Man (n.s.), Vol. IV, No. 1, (1968), P. 94. 21. LeVy-Bruhl, L., How Natives Think, (Washington Square Press, New York, 1966), P. 19. 22. Levi-Strauss, C, Totemiam, (Merlin Press, London, 1962), P. 1. Emphasis in original. 23. Ibid.. P. 3. 24. Ibid., Pp. 2-3. 25. Leach, E.R., "Virgin Birth% in Leach, E.R., Genesis  as Myth and Other Essays, (Jonathan Cape, London, 1969), Pp. 109-10. 26. Ibid.. P. 97. Emphasis in original. 27. Levi-Strauss, C, The Savage Mind, (Weidenfeld & Nicholson, London, 1966), P. 10. 28. Goldmann, L., The Human Sciences and Philosophy, (Jonathan Cape, London, 1969), P* 58. 29. See Polanyi, M., Personal Knowledge, (Routledge & Kegan Paul, London, 1958), P. 58; Barnes, S.B., "Paradigms...", op. cit., espc. Pp. 99-101. 30. Miller, A., The Crucible. (Bantam Books, New York, 1959). Of course I am not here intending to denigrate Miller's remarkable play, nor his purposes. The point I wish to emphasize is that we don't need anthropolo gists to tell us what a "witch-hunt" is. 31. Pocock, D.P., Social Anthropology, op. cit.. P. 3. Emphasis in original. 32. Evans-Pritchard, E.E., Theories of Primitive Religion, (Oxford University Press, London, 1965), P. 100. -80-33. See Douglas, M., Purity and Danger, (Penguin Books Ltd., Harmondsworth, 1970), Pp. 16-40. 34. Jarvie, I.C. and Agassi, J., "The Problem of the Rationality of Magic", British Journal of Sociology, Vol. XVIII, (1967), P. 68. 35. It is important to note that Levy-Bruhl's contention that primitive thought is ultimately incomprehensible was not simply a logical deduction on the part of an armchair anthropologist, but was based upon the testi mony of pioneer field-workers, like Cushing and Best, who had actually lived among primitive peoples and learnt their languages. See Levy-Bruhl, L., How  Natives Think, op. cit., Pp. 55-6. 36. Gough, K., "New Proposals for Anthropologists", Current Anthropology, Vol. IX, (Dec. 1968), P. 403. 37. Malinowski, B., Crime and Custom in Savage Society, (Routledge & Kegan Paul, London, 1926), P. xi. 38. Hogbin, H.I., "Sorcery and Administration", Oceania, Vol. VI, No. 1, (Sept. 1935), Pp. 1-32; Fortune, R., Sorcerers of Dobu, (Routledge and Kegan Paul, London, 1963), Appendix III, "Administration and Sorcery". 39. Pocock, D.F., Social Anthropology, op. cit., P. 73. 40. Malinowski in fact even went so far as to attempt to prevent Evans-Pritchard from publishing Witchcraft,  Oracles and Magic among the Azande. Dr. K. Burridge, pers. comm.. 41. Understanding involves rendering the actions of an in dividual intelligible by reference to the thoughts, goals or intentions of the actor. Explanation, on the other hand, involves rendering a series of actions in telligible by reference to a theory extraneous to the behaviour observed. The difference between these two might be likened to the distinction Winch draws be tween applying one's knowledge of a language in order -81-to make sense of a conversation and applying one's knowledge of the laws of mechanics in order to compre hend the workings of a watch. Winch, P., The Idea..., op. cit., P. 133. 42. Kluckhohn does acknowledge that Navaho witch beliefs possess a certain coherence and logical structure. He does not, however, elucidate this logic which he treats somewhat in the manner of the logic of the in sane . 43. Pocock, D.P., Social Anthropology, op. cit., Pp. 34-5. 44. See the discussion of the theories of Herbert Marcuse and Norman Brown in Roszak, T., The Making..., op.  cit., Ch. III. CHAPTER THREE THE PROGRESSIONISTS A. Introduotion To understand the views of Tylor and Frazer on magic and witchcraft, it is necessary to locate these views in two contexts. The first is that of the intense Victorian concern with religion, and in particular the hitter con flict between some of the nascent natural and human sciences and a literalist theology. The second is that of the general suppositions of their*progressionist methodo logy. Between these, of course, was a very close relation. The Victorians experienced a tremendous religious crisis, which was rooted in a number of different develop ments. The growth of geology and archaeology, and the formulation of the theory of natural evolution, undermined the validity of a literal interpretation of Scripture from a scientific point of view. In addition to this, there was the impact of the Left Hegelians and of Renan, who subjected the Bible to an exacting textual scrutiny that was often devastating and was too scholarly to be lightly 1 dismissed. Finally, the development of comparative mythology and comparative religion began to treat all religious beliefs and rituals as phenomena of the same 2 order and, by implication, of the same validity. These developments occasioned deep and intense con flicts. The denunciation of Lyell because his geology -83-seemed to conflict with scriptural accounts of the Creation and Deluge, the bitter struggles between the Darwinians and the anti-evolutionists, the censure of Jowett for suggesting that the Old Testament must be in terpreted like any other book: in many ways, events such as these were typical of the age. In consequence of this, anthropologists felt that they were living in the midst of an immense crisis in the history of thought. This is reflected in Muller's remark of 1878: Every day, every week, every month, every quarter, the most widely read journals seem just now to vie with each other in telling us that the time for religion is past, that faith is a hallucination or an in fantile disease, that the gods have , at last been found out and exploded. In this crisis, and the conflicts it generated, an thropologists played an important part. Some, like Muller (a devout Lutheran) and Robertson-Smith (who never wavered in his faith in the Bible as a divinely-inspired revela tion), were relatively cautious in their comments on 4 Christianity. Not so the intellectual tendency which was to culminate in Prazer*s work. Most of the important nineteenth century anthropolo gists derived from Dissenting backgrounds, and in addition to this were strongly influenced by such anti-religious intellectual traditions as Enlightenment scepticism and rationalism, Utilitarianism and Comtian positivism. Spencer, Morgan, McLennan, Lubbock, Tylor and Prazer, were all atheists or agnostics and hostile to religion. Thus one of Frazer's purposes in writing The Golden  Bough seems to have been to discredit revealed religion by-showing how one or other of its essential features (the resurrection of the man-god, for example) is analogous to what may be found in pagan religions.^ As J.W. Burrow 7 points out, in doing this he was in essence following in Tylor*s footsteps, for Tylor explicitly stated that he saw anthropology as having two main functions: "to impress men*s minds with the idea of development", and "to expose the remains of the crude old culture which have passed in-g to superstition, and to mark these out for destruction." Tylor*s method in attempting this task consisted in outlining a series of stages of progress, culminating in Victorian society. It has sometimes been asserted that this construction of stages of progress derived from the application of Darwin's ideas from the biological to the q cultural realm. Such a view ignores the very real dif ferences between the biological and cultural models. Por Darwin, in the course of evolution, natural forms dif ferentiate. Por the progressionists - Spencer, Morgan, Tylor and Frazer - all social forms tend to evolve in the same direction, although at different rates. Despite the enormous impact of Darwin on nineteenth and twentieth century thought, the methodology of Tylor and Prazer probably owed more to Herbert Spencer. As early as 1850, nine years before the publication of On the  Origin of the Species, Spencer drew an analogy between society and organism, an analogy he was to elaborate in -85-The Principles of Sociology. As organisms grow, so do societies, although the factors controlling growth in each case are different. Existing "savage11 or "barbarous" societies have been arrested in their growth, and repre sent early stages in the growth of abstract "Society". Contemporary primitive societies could therefore be used to illustrate stages in the temporal process of social evolution. This argument, that existing primitive societies re present stages through which our own civilization has passed, was taken up by Prazer and Tylor, who coupled it with the doctrine of the "psychic unity" of mankind. The main axioms of the resultant methodology may be summarized as follows: i. Human institutions, shaped by a similar human nature, succeed each other in a series substantially uniform over the earth. ii. By utilizing the comparative method, and ignoring dif ferences resulting from environmental and historical con tingencies, it is possible to abstract this common human nature and determine the evolutionary sequence of human institutions. iii. The operations of the human mind, which are mani fested in this evolution, are governed by laws "as de finite as those which govern the motion of the waves, the combination of acids and bases, and the growth of plants and animals."11 -86-B. Magic and the Association of Ideas In accordance with this methodology, both Tylor and Frazer considered magic to constitute an historical sur vival of a once universal mode of conceptualization: a mode of conceptualization that remained, however, amenable to analysis in basically the same terms as our own. The model for this analysis was derived from the associational psychology then dominant. Both Tylor and Frazer considered magic to be founded on an intellectual confusion, representing a mistaken ap plication of the association of ideas. In Tylor's formu lation, primitive man was incapable of seeing that the symbolization of an object was a purely subjective pheno menon and that manipulation of symbolic representations does not entail any modification of the object symbolized. Because of this, association of an object1in thought was believed to entail association in the objective world. Man, as yet in a low intellectual condition, having come to associate in thought those things that he found by experience to be connected in fact, proceeded erroneously to invert this action, and to conclude that association in thought must in volve similar connexion in reality. He thus attempted to discover, to foretell, and to cause events by means of processes which we can now see to have only an ideal signific ance. By a vast mass of evidence from savage, barbaric, and civilized life, magic arts which must have thus resulted from mistaking an ideal for a real connexion, may be -87-clearly traced from the lower cul ture which they are of, to the ^ higher culture which they are in. In support of this contention, Tylor cited a number of ethnographic examples, which Frazer was later to clas sify as examples of the operation of a supposed Law of Sympathy: ... among the Dayaks, young men sometimes abstain from the flesh of the deer, lest it should make them timid, and before a pig-hunt they avoid oil, lest the game should slip through their fingers, and in the same way the flesh of slow-going and cowardly animals is not to be eaten by the warriors of South America; but they love the meat of tigers, stags, and boars, for courage and speed. *«* Prazer considered these examples to typify the opera tion of an imagined Law of Sympathy which, if never analysed as such by the "unreflective magician", could nevertheless be abstracted by the "philosophic student"• The implicit assumption underlying the imagined law was: ... that things act on each other at a distance through a secret sympathy, the impulse being transmitted from one to the other by means of what we may conceive as a kind of invisible ether, not unlike that which is pos tulated by modern science for a pre cisely similar purpose, namely, to explain how things can physically affect each other through a space which appears to be empty.1 * This assumed Law of Sympathy was considered by Frazer -88-to have two applications: i. Homeopathic (imitative) Magic, based on the association of ideas by similarity and implicitly assuming the opera tion in nature of a Law of Similarity. ii. Contagious Magic, based on the association of ideas by contiguity and implicitly assuming the operation in nature 15 of a Law of Contagion. The relationship between these two types of magic is shown in the following diagram: SYMPATHETIC MAGIC (Law of Sympathy) HOMEOPATHIC MAGIC (Law of Similarity) CONTAGIOUS MAGIC (Law of Contagion) Diagram 4: Homeopathic and Contagious Magic Homeopathic magic is based on the mistaken assumption that things which resemble each other may exercise an in fluence on each other. An example would be the evil-wishing sand-paintings made by Navaho Witches and Sorcerers, or the practice of the same Sorcerers of making an image in clay or wood of an intended victim which is then tor tured by being stuck with a pointed object, or by having -89-projectiles shot into it. Navahos believe that by thus treating the likeness of a person, Sorcerers can cause 16 illness or death to befall the person himself. Contagious magic is based on the error of assuming that things which have been in contact with each other may continue to exercise an influence on each other after they have been physically separated. Por instance, a relation ship is sometimes thought to exist between a man and some severed part of his person: hair, nails, spittle, blood, teeth, excreta or soiled clothing. Thus a sorcerer who succeeds in securing these objects is believed to hold a power over their original possessor. Hence the Prussian belief that if you beat the garment of a thief, he will 17 fall ill. Hence also, Navaho secrecy about urination and defecation and care in the disposal of hair-clippings, 18 the placenta and menstrual blood. Of course, Prazer did not consider these principles to be operative only in the case of destructive magic but considered them to be equally characteristic of, for example, love, hunting and rain-making magic. Indeed, Prazer used the terms "sorcerer" and "magician" inter changeably and nowhere attempted to define a theory speci fic to destructive magic, as distinct from a theory of magic in general. It is also important to note that these universal principles of magic were not thought of as nec-cessarily existing in "pure" form: ... in practice the two branches are often combined: or, to be more exact, while homeopathic or imitative magic -90-may be practised by itself, con tagious magic will generally be found to involve an application of the homeopathic or imitative prin ciple.19 Por Prazer, these principles of homeopathic and con tagious magic were operative not only positively, but also in the case of ritual prohibitions. Cross-cutting the broad division between homeopathic and contagious magic, therefore, Prazer established a second dichotomy along another axis: that between sorcery and taboo, or the posi tive and negative precepts of magic. The resultant theo retical scheme is shown below: MAGIC THEORETICAL PRACTICAL (Pseudo-Science) (Pseudo-Art) SORCERY TABOO (+ Magic) (- Magic) Diagram 5: Magic, Sorcery and Taboo -91-C, Magic, Science and Religion Tylor did not attempt to draw a firm distinction be tween magic and religion, being content to offer as a "minimum definition" of the latter, "the belief in Spiri tual Beings", assigning the remainder of the supernatural 20 to the domain of magic. Similarly, he did not attempt to discuss the question of the relationship of magic to science, although implicit in his position is a differentiation between a subjective, fallacious association of phenomena (magic) and an as sociation of phenomena between which there exists a real or objective link (science). Both these ideas were developed by Prazer, who may, perhaps, be credited with having initiated systematic dis cussion of the relation of magic to science and religion. According to Prazer, the essential difference between magic and religion was that the former implicitly assumed the operation of mechanical laws of causality in nature, whereas religion postulated that nature is subject to the direction of some "superhuman being(s)", possibly capri cious, who must be propitiated. Magic and religion, therefore, represent two cosmological systems, two at tempts to reduce the universe to order, although (as with homeopathic and contagious magic) these two systems rarely exist in pure form, being usually fused in ethnographic 21 reality. The distinction between magic and science in Prazer's scheme is far less clear, though he argued the close kin--92-ship of these two modes of thought on the grounds that both rest on the assumption that nature is governed by im mutable mechanical laws. The only clear difference be tween the two that he seems to have drawn is that in the case of magic this belief in the operation of immutable laws throughout nature remains implicit, whereas with 22 science this assumption is rendered explicit. In addi tion to this, however, Prazer seems to have thought of science, as did Tylor, as formulating laws which correspond to objective reality, in contrast to the illusory presup-23 positions of the magician.. The differentiation of three such modes of thought immediately poses the question of how they are interrelated. The nineteenth century was, of course, the period in which most social science was cast in the mould of progres sionist models of the succession of historical stages. To his credit, however, Tylor did not seek to interrelate magic, science and religion in this manner, despite his general commitment to the progressionist methodology. In stead, he argued that magic, science and religion are to be found in all societies, although in the more advanced cultures animistic and magical ideas come to play a lesser role in human thought than in earlier stages of develop ment and linger on mainly as survivals. His scheme may therefore be represented as in Diagram 6 on the following page. Prazer went a step further than this. Magic, science and religion are not only of varying importance from one -93-Higher Stages of Culture MAGIC Lower Stages of Culture Diagram 6: Tylor*s View of Magic, Science and Religion society to another, but could actually be arranged in a hierarchy of stages. According to this hierarchy, in the course of human social evolution magic is gradually re placed by religion, which in turn is replaced by science. This scheme may be illustrated as follows: Higher Stages of Culture SCIENCE RELIGION MAGIC Lower Stages of Culture Diagram 7: Frazer's View of Magic. Science and Religion -94-In Prazer's view, since it was inevitable that the first attempts to understand reality should be largely mistaken, it was a reasonable assumption that magical (false) notions would have preceded scientific (true) ones in the course of intellectual evolution. Religion he placed between magic and science, on the grounds that it represents a less primitive mode of thought than magic, despite the latter's close kinship to science. This argu ment he justified on three grounds: i. The association of ideas underlying magic is an ele mentary mental process, characteristic of even animal in telligence. The notion that nature is controlled by in visible beings, on the other hand, is indicative of a much more complex way of thinking and must be reserved to human 24 intelligence. ii. Religious belief may be differentiated into an enor mous number of varieties, largely affecting the more thoughtful members of the communi"fcy. Magic, in contrast to this, constitutes a "universal substratum of uniformity", 25 characteristic of the dull-witted and superstitious. iii. The Australian Aborigines, according to Prazer the most primitive of existing peoples, practise magic uni versally. On the other hand religion, in the sense of the propitiation of higher powers, is almost unknown among them: "... nobody dreams of propitiating gods by prayer and sacrifice."^ Magic therefore represented the primordial mode of thought, transcended only when some of the more intelli gent members of humanity became conscious of its inef--95-ficacy. Confronted by the failure of their magic, these individuals proceeded to postulate that nature was not governed by immutable mechanical laws, but was subject to the direction of some mightier power(s) whose favour must be secured. In this way magical thought gave way to re ligious, although the transition was not pure and traces of magic continue to persist in most religions: in the beliefs of the French Catholic peasantry concerning the Mass of the Holy Spirit and the Mass of St. Secaire, for 27 example. Yet in the course of time this religious mode of thought also proves unsatisfactory, for it assumes that the succession of natural events is not invariable but is subject to alteration. The keener-minded perceive that order actually does exist, and as their comprehension of this order gradually extends they come to reject the re ligious mode of thought and revert to the postulate of an inflexible regularity in the order of nature. In this 28 way, scientific thought comes to replace religious. D. The Stability of Magical Belief The fundamental characteristic of primitive magic being defined by Tylor and Frazer as objective falsehood, the question at once arises as to why this falsehood should not be immediately perceived, in view of its dis crepancy with empirical reality. Why, in other words, should primitive peoples continue to cling to magical no tions when these are so obviously fallacious? In answer to this question, Tylor suggested six main reasons: -96-i. Some of the results aimed at by magicians are actually achieved, although for different reasons than the practi tioners of magic believe. The power of suggestion, for example, might slay a victim, or a naturally curative agent be incorporated into a ritual healing process. Alternatively, a magical rite might coincide with the result it was designed to achieve by sheer chance. ii. "Conjurers' tricks" are sometimes used to maintain the prestige of the magician, although magic is not simply a matter of fraud but also "a sincere but fallacious system of philosophy". iii. Successes always outweigh failures in the eyes of those who believe in magic. iv. The plasticity of primitive notions of success and failure makes it difficult to specify when a magical ritual has definitely failed. v. Magical performances are typically associated with so many ritual prohibitions and requirements that the failure of a rite can always be explained by arguing that one of the difficult preconditions necessary for success was not met. vi. It is always possible to claim that the effect of magic has been nullified or reduced by the operation of 29 counter-magic. . E. A Critique of Tylor and Frazer We cannot deny the possible operation of all these factors, although we might wonder what sort of explanation it is of a supposedly once-universal conceptual system, to -97-say that it originated in an error in logic and is main tained by a combination of secondary elaborations, in cidental successes, credulity and fraud.^ Moreover, re lated to this is a fundamental problem inevitably encoun tered by the methodology of Tylor and Prazer. If a common mental process is postulated to underlie a variety of cul tural forms, then one is faced with the problem of ex plaining this variety. Or, to put this in the terms in which Tylor and Prazer would have seen it, the hypothesis of the psychic unity of mankind must be reconciled with the fact of progress. This problem is not in fact insoluble. One might, for example, resolve it in terms of the differential ex perience organized by the common mental functions, and from here be led into an examination of the social struc turing of ideas. Tylor, however, did not choose this path but resorted instead to the theories of racial determina tion so common in his day. Thus, although in Primitive Culture Tylor resolved to "... eliminate considerations of hereditary varieties or races of man, and to treat mankind as homogenous in nature, 31 though placed in different grades of civilization", he nevertheless went on in the same work to compare the men tal capacities and morality of "savages" to those of chil dren, while in Anthropology he specifically attributed this moral and intellectual "inferiority" to the supposedly lesser degree of development of the brain of non-white 32 peoples. -98-Such a view accords perfectly with Tylor's explana tion of magical practises as resulting from mere intellec tual confusion, as also with the nineteenth century stereotype of the "savage": as credulous, incapable of generalization, speaking rudimentary languages and com-33 municating by grimaces, etc.. Tylor's view of primitive peoples therefore reveals the smug certainty of intellec-34 tual and moral superiority typical of his era, and also indicates his antipathy to his object of study: indeed, he sought its obliteration. He therefore appears in a dual, though fused, guise. On the one hand, he clearly repre sents an intellectual justification for nineteenth century imperialism. On the other, he seeks to eradicate the last vestiges of those "most pernicious illusions that have 35 ever vexed mankind" which have continued to persist in the "highest" civilization, by showing their origins. Such a task was made all the more important for Tylor in that he considered that the progress of civilization did not constitute a single, unbroken advance. He noted the "revival" of witch belief in Europe in the period be tween the thirteenth and seventeenth centuries, and con sidered the development of Spiritualism in his own time to be a similarly dangerous example of "revival in culture". A great philosophic-religious doc trine, flourishing in the lower cul ture but dwindling in the higher, has re-established itself in full vigour. The world is again swarming with powerful disembodied spiritual beings, whose direct action on -99-th ought and matter is again con fidently asserted, as in those times and countries where physical science had not as yet so far succeeded in extruding those spirits and their in fluences from the system of nature. 36 Yet such "revivals" in culture, Tylor was merely to describe. They are not systematically integrated into his progressionist model, and therefore remain unaccounted for. A very similar criticism may he levelled at Prazer, who was not in any satisfactory manner able to account for the evolutionary progression which he argued led from magic, through religion, to science. Prazer's suggestion, of course, was that the whole process could be explained in terms of the intelligence of some of the more perspi cacious members of each society. These, confronted by the failure of their magical or religious rites, would even tually see through the confusion of ideas underlying their beliefs and come to the conclusion that the universe is governed by principles different from those they assumed operative. Yet why this should have happened in different societies at different times; why the Arunta, for example, unlike those in certain other societies, should not have been able to perceive the fallacy of their magical beliefs: questions such as these are met with a blank within the context of Prazer*s system. In addition to this, it is necessary to add that none of Prazer's grounds for regarding magic as a more primi tive mode of thought than religion can be accepted. As -100-Robert Lowie argued: i. It may be true that animals engage in the association of ideas: but it is as surely untrue that they carry out magical rites as it is that they postulate the existence of religious essences. As soon as one passes from the association of ideas underlying magic to the magical pro cesses themselves, "the extreme simplicity alleged by 37 Prazer vanishes in thin air." ii. It is equally untrue that, considered cross-culturally, magic constitutes a substratum of intellectual uniformity, whereas religious ideas may be differentiated into a large number of varieties. Magic shows far fewer uniformities than Prazer alleged. The Arunta, for example, lacked con tagious magic; divination was far more pronounced in the Old World than the New; the use of the spell was highly elaborated in Oceania, but unimportant in most of North America. In other words, Prazer did not base his gene ralization on an objective appraisal of available data: If religions are compared in their specific characteristics and magical faiths only as regards their abstract common traits, the former will of course appear diverse and the latter uniform. A fair survey, on the other hand, will bring out frequent recurrences of religious no less than of magical practice.38 iii. Prazer's contention that the Australian Aborigines lacked any form of religion is untenable. In fact, al though Prazer retained this assertion in his 1922 edition of The Golden Bough, it had already been refuted by the -101-work of Mrs. K. Langloh Parker on the Euahlayi tribe as 30, early as 1905. It is necessary, however, to balance the force of all these criticisms of Tylor and Prazer against the fact that both sought to explain magico-religious phenomena as essentially rational, if mistaken, phenomena. Magic, for them, constituted a coherent mode of thought, fully com prehensible once the basic principles on which it is founded are understood. Thus, despite Prazer*s tendency to view the believer in magic as a credulous fool, and Tylor1s racism, neither postulated a gap between primitive and civilized mentalities comparable to that postulated by L€vy-Bruhl. -102-Notes and References 1. Evans-Pritchard, E.E., "Religion and the Anthropo logist", in Evans-Pritchard, E.E.j Essays in Social  Anthropology, (Faber & Faber, London, 1962), P. 34. 2. Thus inducing the Bishop of Gloucester to condemn at tempts "to put into competition the sacred books of India and the Holy Scriptures". Ibid., P. 35. 3. Quoted in Evans-Pritchard, E.E., Theories of Primitive  Religion, (Oxford University Press, London,1965), P. 100. 4. This circumspection did not save them from running afoul of ecclesiastical authority. Muller was denied the Chair of Sanskrit at Oxford because his teaching was considered subversive of the Christian faith, while Robertson-Smith was removed from the Chair of Hebrew at the University of Aberdeen for publishing certain critical remarks on the dating, order and com position of the Books of the Old Testament. 5. Evans-Pritchard, E.E., "Religion...", op. cit., P. 35. As Evans-Pritchard points out, this tradition of agnosticism and atheism has continued to prevail, both in British and in American anthropology, until the present day. 6. Ibid., Pp. 35-6. I follow here the interpretation of Evans-Pritchard. Leach has recently contested this, mainly on the grounds of Frazer*s caution in his direct comments on Christianity. See Leach, E.R., "Frazer and Malinowski", Encounter, Vol. XXV, (Nov. 1965), Pp. 24-36. See also Jarvie's critique of Leach:-Jarvie, I.C., "Academic Fashions and Grandfather Killing - In Defence of Frazer", Encounter, Vol. XXVI, (April 1966), Pp. 53-5. 7. Burrow, J.W., Evolution and Society, (Cambridge University Press, London, 1966), P. 250. -103-8. Tylor, E.B., Primitive Culture. (John Murray, London, 1920), Vol. II, P. 410. Quoted in Burrow, J.W., Evolution..., op. cit., P. 258. 9. Cf. Jarvie, I.e., The Revolution in Anthropology. (Routledge & Kegan Paul, London, 1967), P. 9. 10. Childe, V.G., Social Evolution. (Watts & Co., London, 1951), Pp. 2-5. 11. Tylor, E.B., Primitive Culture, op. cit., Vol. I, P. 2. 12« Ibid.. Vol. I, P. 133. 13. Tylor E.B., Researches Into the Early History of  Mankind. (John Murray, London, 1870), P. 133. 14. Prazer, J.G., The Golden Bough, abridged ed., (Macmillan & Co. Ltd., London, 1960), Vol. I, P. 16. 15. Jakobson has suggested that all intersubjective sym bolic acts may be placed in relation to one of two axes: metaphoric and metonymic. The first of these involves the selection and substitution of symbols and is dependent on the perception of similarity; the second combines symbols in a particular context ac cording to the perception of contiguity. The first axis is therefore paradigmatic, and the second syntag-matic. Jakobson further suggests that the differen tiation of these two axes corresponds to Prazer*s dis tinction between homeopathic magic (metaphoric axis) and contagious magic (metonymic axis). In light of this correspondence Prazer*s distinction, although un developed, becomes all the more significant, and the possibility of an analysis of magical beliefs and practices modelled on the methods of structural linguistics may be envisaged. Cf. Jakobson, R., "Deux aspects du langage et deux types d'aphasies", in Jakobson, R., Essais de Unguistique generals, (Les Editions de Minuit, Paris, 1963), espc. Pp. 61-7. 16. Kluckhohn, C, Navaho Witchcraft, (Beacon Press, Boston, 1967), P. 32. -104-17. Frazer, J.G., The Golden Bough, op. cit., Vol. I, * P. 57. 18. Kluckhohn, C, Navaho Witchcraft, op. cit., P. 54. 19. Frazer, J.G., The Golden Bough, op. cit., Vol. I, 20. Tylor, E.B., Primitive Culture, op. cit.. Vol. II, P. 424. 21. Frazer, J.G., The Golden Bough, op. cit., Vol. I, - P. 67. 22. Ibid.. 23. Ibid., Vol. I, P. 65. See Ch. IX of this thesis for a fuller discussion of Frazer*s view of science. 24. Frazer, J.G., The Golden Bough, op. cit., Vol. I, P. 71. 25. Ibid., Vol. I, Pp. 72-3. 26. Ibid.. 27. Ibid., Vol. I, Pp. 68-77. 28. It is necessary to remark at this point, however, that there is a certain ambiguity to Frazer*s thought on this matter. Thus, towards the close of The Golden  Bough, he actually suggests the co-existence of magic and science from the earliest times. Adoption of such a point-of-view, however, poses special problems for Frazer*s progressional sequence. For if science and magic are co-extensive, how can one establish an evolutionary succession between them? Cf. Ibid., Vol. II, P. 933. 29. Tylor, E.B., Primitive Culture, op. cit., Vol. I, Pp. 133-5. Tylor*s analysis of the secondary protec tive mechanisms of magical belief has become classic, and Prazer was unable to improve on it in any way. -105-Malinowski was to add the further reason, for the per petuation of magical beliefs, that a sorcerer will often frankly admit his responsibility for an illness, whether true or not, in order to enhance his reputa tion. Apart from this new suggestion, however, Malinowski was to repeat Tylor's arguments. Cf. Malinowski, B., Argonauts of the Western Pacific, (E.P. Dutton & Co., Inc., New York, 1961), P. 76. Tylor's arguments are also repeated in Firth, R., Human Types, (Mentor, New York, 1958), P. 128. Probably the ful lest analysis of the secondary protective mechanisms of magical belief is that of Evans-Pritchard, E.E., Witchcraft, Oracles and Magic among the Azande, (Oxford University Press, London, 1937), Pp. 314-44, 475-8. Even here the debt to Tylor is obvious* 30. Hence Durkheim's criticism of the approach of Tylor and Prazer: "Quand les philosophes du XvTIIe siecle faisaient de la religion une vaste erreur imaginee par les prStres, ils pouvaient du moins en expliquer la persistance par 1'inters*t que la caste sacerdotale avait a tromper les foules. Mais si les peuples eux-mSmes ont 6t6 les inventeurs de ces systemes d*id£es erronndes en meme temps qu'ils en Staient les dupes, comment cette duperie extraordinaire a-t-elle pu se perpStuer tout le long de l'histoire?" Durkheim, E., "Examen critique des systemes classiques sur les origines de la pensie religieuse", Revue philosophique  de la Prance et de 1'Stranger, Vol. LXVII, (Jan. 1909), P. 28. 31. Tylor, E.B., Primitive Culture, op. cit., Vol. I, - P. 7. Quoted in Harris, M., The Rise of Anthropo  logical Theory, (Thomas Y. Crowell, New York, 1968), P. 140. 32. Tylor, E.B., Anthropology, (D. Appleton & Co., New York, 1898), Pp. 74-5. 33. See the examples given by Evans-Pritchard, E.E., Theories..., op. cit., Pp. 105-6. -106-34. Frazer is often castigated for his "complacency and undisguised contempt of primitive society". Douglas, M., Purity and Danger, (Penguin Books ltd., Harmondsworth, 1970), P. 36. But it was Tylor, who is usually accorded much more sympathetic treatment, who wrote that: "... there is this plain difference be tween low and high races of men, that the dull-minded barbarian has not power of thought enough to come up to the civilized man's best moral standard. The wild man of the forest, forgetful of yesterday and careless of tomorrow, lolling in his hammock when his wants are satisfied, has little of the play of memory and fore sight which is ever unrolling before our minds the panorama of our own past and future life...". Tylor, E.B., Anthropology, op. cit., P. 407. Further, "... a Londoner who should attempt to lead the atrocious life which the real savage may lead with im punity and even respect, would be a criminal and only allowed to follow his savage models during his short intervals out of gaol. Savage morals are real enough, but they are far looser and weaker than ours. We may, I think, apply the oft-repeated comparison of savages to children as fairly to their moral as to their in tellectual condition." Tylor, E.B., Primitive Culture, op. cit., Vol. I, P. 31. 35. Ibid., Vol. I, P. 112. 36. Ibid., Vol. I, Pp. 142-3. 37. Lowie, R.H., Primitive Religion, (Boni & Liveright, New York, 1924), Pp. 143-4. 38. Ibid., P. 146. 39. Ibid., P. 145. That the Australian Aborigines lacked any form of religion was, of course, a common assump tion in Frazer's time. Cf. Evans-Pritchard, E.E., Theories..., op. cit., P. 197. CHAPTER FOUR LUCIEN LEVY-BRUHL A. Introduction The work of Lucien Levy-Bruhl has rarely been accorded a sympathetic treatment by British or American anthropo logists. As Mary Douglas points out, most textbooks on comparative religion are emphatic about the mistakes that he made, but say nothing of the value of the questions he raised.1 Some have not even provided an accurate presen-2 tation of his ideas. It is undeniable that a forceful critique may be made of Levy-Bruhl's theories, but for a fair evaluation it is necessary to view his work within its historical context. Only then may we form an accurate idea of his contributions. Levy-Bruhl's theory of primitive mentality (for he wrote on this in general, rather than about religion, magic or witchcraft specifically) has two sides. On its critical side, it is an attack on the "intellectualist" interpreta tions of magic and religion offered by "the English school", especially Tylor and Frazer. In its essentials, this attack accords with the views of Durkheim and the other writers of the Ann£e sociologique. On its construc tive side, however, it has a markedly distinctive and ori ginal character. The critique by Levy-Bruhl of Tylor and Frazer was on two levels. Firstly, he attacked the associational psycho--1 Oo logy which was the basis of their theories. This; he criti cized as being inadequate as psychology, in that it did not take proper account of the importance of the emotional and motor elements in mental life and of their influence on intellectual life properly so-called.^ This criticism was similar to that made by a number of post-Prazerian anthropologists - Marett and Malinowski, for example - who made of it the starting point of an analysis of magic and religion as "affective" phenomena, thus retaining a psy chological frame of reference. Levy-Bruhl, however - with Durkheim - went a step further than this, in repudiating the very idea of attempting an explanation of what are essentially social phenomena in psychological terms. The methodology of Tylor and Prazer rested on two im portant and closely related presuppositions. Firstly, it assumed the validity of introspection as a mode of attain ing cultural knowledge. Both Tylor and Frazer assumed that their modes of thought were essentially the same as those of primitive peoples, and both considered that in order to deduce how primitive man must think and feel in particular circumstances it was sufficient to imagine what their own thoughts and feelings would be in such circumstances. Secondly, it assumed that formally similar customs from differing cultural contexts could be ascribed the same meaning. The idea was not yet current that such customs were connected with others in a particular institutional context and formed a complex of meaning. Explanation in terms of individual psychology- had led away from an -109-appreciation of this insight. As a result, the approach of Tylor and Frazer was at tended by serious dangers, often leading the anthropologist to read erroneous, or factually unjustifiable, meanings into practices. This is well illustrated by an example taken from Frazer and cited by Levy-Bruhl in order to show the shortcomings of the English school of anthropology. Frazer had interpreted the burial custom of placing a gold coin in the mouth of the deceased as an attempt to provide the dead with a means of purchasing food in their after-life. As such, he interpreted this custom as an historical substitution for an earlier practice of placing food in the mouth of the dead. On the face of it, such an explanation seems plausible, as do so many of those of Tylor and Frazer. But, as Levy-Bruhl pointed out, in the; one case where this theory can actually be checked, is is incorrect. For among the Ancient Chinese, the placing of gold, jade and pearls in the mouth of the deceased was; not done with the aim of providing the departed with the means of purchasing food in the next world. In Chinese thought, these substances were conceived: of as substances of the celestial sphere, composed of Yang matter, and as powerful counter-agents of corruption and decay. Among the Chinese, therefore, this custom represented an attempt to preserve the corpse from decomposition and thus render possible the 4 continued use of the body after death. The trouble with the Tylor-Frazer method of intro spection, in assuming that what would be reasonable infer--lic ences and logical conclusions for us would also be so for the members of other societies, is that it leaves out of account the differing social experience of the members of each society. Ultimately, therefore, it rests on the pos-tulate of an unsocialized mental life: We might just as well hope to make: acientific use of the idea of a hu man individual mind imagined to be devoid of all experience whatever. Would it be worth while to try/ and reconstruct the method in which such a mind would represent the natural phenomena which occurred within and! around, him? As a matter of fact we have no means of knowing what such a mind would be like. As far back as we can go, however primitive; the races we may study, we shall never find any minds which are not social ized, if we may put it thus, not al ready concerned with an infinite number of collective representations: which have been transmitted by tradition, the origin of which is lost in obscurity.5 Primitive beliefs should not, therefore, be viewed as individual responses to a presumed universal need for ex planatory devices, nor as individual deductions rendering intelligible personal experiences. Rather, these beliefs must be taken as primary, as patterns of thought which de termine the thinking of the particular individual. Thought processes are social facts, determined by collective rep  resentations . These Levy-Bruhl defined as: ... common to the members: of a given social group; they are transmitted -111-from one generation to another; they impress themselves upon its indi vidual members, and awaken in them sentiments of respect, fear, adora tion, and so on, according to the circumstances of the case. Their existence does not depend upon the individual; not that they imply a collective unity distinct from the individuals composing the social group, but because they present them selves in aspects which cannot be accounted for by considering indi viduals merely as such. Thus it is that a language, although, properly speaking, it exists only in the minds of the individuals who speak it, is none the less an incontest able social reality, founded upon an ensemble of collective representa tions, for it imposes its claims on each one of these individuals; it is in existence before his day, and it survives him.6 Patterns of thought, being in this way social pheno mena, must necessarily vary from one society to another, the various aspects of society being interdependent. This was not to deny a certain basis of homogeneity to the thought patterns of all societies. Obviously, L^vy-Bruhl considered, this must exist, insofar as in all societies/ languages are spoken, traditions transmitted and institu tions maintained. But since societies also vary profoundly in their organization, we must expect patterns of thought to vary concomitantly. Accordingly, the idea of reducing all mental operations to a single type, explicable in terms of the mental functioning of the "adult civilized white -112-7 man", must be rejected. What Levy-Bruhl should have attempted at this point, as Evans-Pritchard was to remark, is a study of how par ticular thought patterns are related to particular modes of g social organization. But he made no attempt to demonstrate in detail the social determination of thought, and argued in fact that the state of knowledge of his time made such a detailed comparative study impossible. He therefore pro posed instead, by way of a preliminary investigation, to analyse only "the most general laws" governing the collec tive representations of primitive peoples. I shall endeavour to construct, if not a type, at any rate an ensemble of characteristics which are common to a group of neighbouring types, andi in this way to define the essential features of the mentality peculiar to undeveloped peoples.9 In order to differentiate these features; as clearly as possible, he proposed to compare this mentality with "our own": ... i.e. with that of the races which are the product of "Mediterranean" civilization, in which a rationalis tic philosophy and positive science have been developed.10 By undertaking this comparison of these two types of men tality - the two types between which, he believed, dif ferences were most strongly marked - Levy-Bruhl believed he would highlight the distinctive features of both, making it easier to analyse later transitional and intermediate types. -113-The civilized mental type, however, he sought to: use merely as a base for comparative purposes and did not examine in depth in any way. He such an examination un necessary, since the civilized mode of thought was already "sufficiently well defined in the works of philosophers:, logicians and psychologists", and not in need of further elaboration. B. The Nature of Primitive Mentality Por the purposes of describing primitive mentality, Levy-Bruhl felt compelled to devise a new terminology. It was obvious, he thought, that if this mentality was govern ed by principles different from our own, then it could not be understood by means of the terminology devised, by logicians and psychologists, for the purpose of analysing the latter. The distinction drawn by psychologists between emotional and intellectual phenomena, for example, could not be applied to the analysis of the collective represen tations of primitives. Por the collective representations of primitives are suffused with feelings of hope and fear, and include emotional and motor elements as integral 12 parts. L^vy-Bruhl therefore suggested three special terms in order to designate the attributes of primitive thought: prelogicality, mysticism, and the law of partici pation. The use of these terms has been undeniably respon sible for many of the misunderstandings of, and much of the hostility towards, Levy-Bruhl*s work.1^ What, then, did he use them to refer to? When he spoke of primitive thought as being prelogical, -114-Levy-Bruhl did not mean to imply that primitives are un intelligent, that they are incapable of coherent thought, or even that their modes of reasoning necessarily violate the rules of logic. Nor did he intend to imply an evolu tionary sequence by his terminology/:^ By Prelogical we do not mean to assert that such a mentality constitutes a kind of antecedent stage, in point of time, to the birth of logical thought. Have there ever existed groups of hu man or prehuman beings whose collec tive representations have not yet been subject to the laws of logic? We do not know, and in any case1, it seems to be very improbable. At any rate, the mentality of these unde veloped peoples which, for want of a better term, I call prelogical, does not partake of that nature. It is not antilogical; it is; not alogical either. By designating it "pre logical" I merely wish to state that it does not bind itself down;, as our thought does, to avoiding contradic tion. It obeys the law of participa tion first and foremost. Thus oriented, it does not expressly de light in what is contradictory (which would, make it merely absurd in our eyes), but neither does it take pains to avoid, it. It is often wholly in different to it, and that makes it so hard to follow/. 15 Perhaps, as Evans-Pritchard says, Levy-Bruhl was being a little too subtle here, for he meant only that primitive thought does not always present the same logical require ments, and that it is therefore uncritical."^ -115-Passing from this characterization, of the way in which the elements comprising the content of primitive thought are interrelated, to an analysis of that content itself, Levy-Bruhl suggested that the distinguishing characteristic of this could be described as mystic. In adopting this term, he did not intend to imply any re ference to any form of transcendental religious experience. Rather, he employed it, ... in the strictly defined sense in which "mystic" implies belief in forces and: influences and actions which, though imperceptible to senses, are nevertheless real.17 Evans-Pritchard states that when Levy-Bruhl spoke of the mystic character of primitive mentality, he meant no more than what other anthropologists have meant when they have spoken of the primitive's "belief in the super-18 natural". It is important, however, to understand Levy-Bruhl's:; preference for one term rather than the other. In his view, there is no "natural" for the savage, and hence no "supernatural". This is a distinction that we make, not one intrinsic to primitive thought, which does not operate in terms of such a dualistic conceptualization of reality. The superstitious- man, and frequently also the religious man, among us, be lieves in a twofold order of reality, the one visible, intangible, "spiri tual", forming a mystic sphere which encompasses the first. But the pri mitive's mentality does not recognize two distinct worlds in contact with each other, and more or less inter--116-penetrating. To him there is but one. Every reality, like every influence, is mystic, and consequently every perception is also mystic.19 What exactly Levy-Bruhl meant by describing the per ceptions of primitives as mystic is not entirely clear, but need not detain us here as it is incidental to his main argument. There is* however, a valuable suggestion contain ed in his writings on this point. This is that the atten tion paid to phenomena is in large measure determined by the collective representations of society, and must there fore be expected to vary concomitantly with variations in social organization. Educated Europeans, for example, normally pay little attention to their shadows, because for them the shadow signifies: merely a negation of light. The Bakwiri of West Africa, on the other hand, attribute great significance to their shadows and are careful to avoid "losing" them under the mid-day sun. Their collective rep resentations direct their attention more strongly towards this aspect of physical reality.20 According to Levy-Bruhl, these mystic collective rep resentations are connected in a network of participations, so that one exercises an influence on the others. This: law of participation he considered to be the principle peculiar to primitive mentality par excellence. The opposition be tween some/other, or one/many, does not impose on this thought the necessity of affirming one term if the other is denied, or vice versa. In this way, the Bororo may declare themselves parakeets, and in doing so claim more than a -117-mere relationship is claimed: an actual identity is affirm ed. Such an identity is affirmed by all communities of the totemic type, and is an example of the law of participation. Participation may also be manifested through such in fluences as homeopathic and contagious magic, or tele kinesis. Thus, what a wife does or does not do in camp may, for example, affect her husband's hunting activities. All the facts that Tylor and Prazer grouped under the rubric of sympathetic magic, were ascribed by LeVy-Bruhl to the prin-21 ciple of participation. It is important, however, to appreciate his advance on their position: His analysis is not like that of the just-so stories we have earlier con sidered, for he does not try to ex plain primitive magic and religion by a theory purporting to show how they might have come about. He takes them as given, and seeks only to show their structure and the way in which they are evidence of a distinctive men tality common to all societies of a op certain type, C. A Critique of Levy-Bruhl How successful may we adjudge Levy-Bruhl'S enter-• «23 prise? We may begin with what is perhaps the most obvious criticism to be levelled against him: that his dichotomy between primitive and civilized is too crude to be of much use. We cannot treat Azande, Bororo, Chinese, Iroquois, Maori and Zuni cultures as all constituting a single type, for they have little in common, even when placed in opposi--118-tion to European culture. Secondly, this same "European culture" was treated by Levy-Bruhl in vague terms. When he spoke of the mentality of this culture being the product of Mediterranean civili zation, of a positive science and a rationalistic philo sophy, who was he including in his designation? Por there are obvious differences between the dominant ideas of, say, Russian peasants, Welsh miners, Breton fishermen, Italian priests, Swiss bankers or functionaries of the Communist Party of the Soviet Union: to suggest only a few of the countless divisions and subdivisions one could make on the basis of national, educational, political, religious or oc cupational categories. Moreover, who in the following example is reasoning mystically, indicating a belief in the operation of in visible forces: the South African missionaries, or the natives who "only believe what they see", and of whom it is recorded that, ... in the midst of the laughter and applause of the populace, the heathen enquirer is heard saying "Can the God of the white men he seen with our eyes?.....and if Morimo (God) is ab solutely invisible how. can a reason able being worship a hidden thing?24 Whatever their other failings, Tylor and Prazer at least did not allow a simple dichotomy between an undif ferentiated primitive and an undifferentiated civilized mentality to obscure for them some of the variations within the latter, or blind them to some of the similarities -119-between certain European beliefs and customs and cor responding beliefs and customs in various primitive so cieties.2"* This failure of Levy-Bruhl*s was compounded by his further failure to analyse the characteristics of the civi lized mentality which he considered already adequately described in the writings of philosophers, psychologists and logicians. Por, on the basis of precisely these writ ings, ViIfredo Pareto was to make an analysis of European mentality in which logicality played about as important a role as it did in the analysis of primitive mentality made 26 by Levy-Bruhl. Many of Levy-Bruhl*s inadequacies stem from the fact that he was: unaware that what he termed "mystic thought" is; a function of particular situations. Different ideas are evoked by different objects in different situations, and mystic qualities are therefore not necessarily attributed to objects outside their ritual context. The Azande, for example, place stones in the forks of trees in order to delay the setting of the sun. But the stones used for this purpose are casually picked up and have a mystical signi ficance only during the rite in which they are used. The sight of such a stone in any other situation does not necessarily evoke the idea of the setting sun. Similarly, the Azande often use their ancestral shrines as; convenient posts on which to lean their spears or hang their baskests;, and at such times have no interest in their shrines except as convenient posts. At religious ceremonies, however, -120-their attitudes are very different. Again, no necessary contradiction is involved in such affirmations as that a man is a parakeet, that the sun is a white cockatoo, or that t»inS are birds.28 Perhaps many of Levy-Bruhl*s errors in conceptualizing primitive mentality were inevitable, considering the nature of many of the sources he was compelled to draw on for his basic data. As Evans-Pritchard explains, His; authorities had collected all the information they could get about the mystical beliefs held by a community of savages about some phenomenon and pieced them: together into a coordi nated ideological structure. These beliefs, like the myths which Europeans also record, may have been collected over a long period of time and from dozens of: informants. The resulting pattern of belief may be a fiction since it may never be actual ly present in a man's consciousness and may not even be known to him in its entirety. This fact would have emergedl if records of everything a savage does and says throughout a single day were recorded for then we would be able to compare our own thoughts more adequately with the real thoughts of savages instead of with an abstraction: pieced together from persistent enquiries conducted in an atmosphere quite unlike that of the savage*s ordinary milieu and in which it is the European who evokes the beliefs by his questions rather than the objects with which they are associated. It would also have emerged had Levy-Bruhl attempted to -121-contrast the formalised beliefs of Europe with those of savages, had he, for instance, attempted to contrast the formal doctrine of Christianity with the formal doctrines of savage religion. What he has done, in fact, is to take the formalised doctrines of savage religions as though they were identical with the actual mental experience of individuals. It is easy to see that it would never do to regard as; identical the thoughts of a Christian with Christian thought. Moreover, primitive thought as pieced together in this manner by European observers is full of contradictions which do not arise in real life be cause the bits of belief are evoked in different situations.29 L^vy-Bruhl's posthumously published Notebooks shows that he was himself keenly aware> of many of the inade quacies of his earlier writings."*0 Thus he abandoned the term "prelogical" as inappropriately suggesting a sequen tial development from prelogical to logical thought, where as, in reality, these two types of thought have co-existed. He therefore extended the concept of the mystical to cover types of thought not governed by the rules of Aristotelian logic, and qualified the simple dichotomy he had establish ed between civilized and primitive mentalities. Thus he wrote: Let me expressly correct what I be lieved to be true in 1910: there is no primitive mentality which is dis tinct from the other.... There is mystical mentality more marked and more easily observed among primitives -122-than in our own society, but present everywhere in the human mind."31 We must not, however, allow these criticisms - and Levy-Bruhl*s own acknowledgement of their validity - to blind us to the importance of his contributions. Por LeVy-Bruhl was among the first to emphasize that primitive be liefs are integrated systematically, that particular be liefs must be understood in terms of the total conceptual structure of which they form a part, and that these con ceptual structures must be related back to the level of social organization. Yet he himself, in his later works, was not to extend these insights. This was to be left to those anthropologists who had the advantage of carrying out the field-work he was unable to. This was to safeguard them from the distorted accounts of primitive beliefs and customs on which LeVy-Bruhl, like Tylor and Prazer before him, had had to rely. -123-Notes and References 1. Douglas, M., Purity and Danger. (Penguin Books Ltd., Harmondsworth, 1970), P. 82. 2. Malinowski provides an example of the misrepresentation of Levy-Bruhl*s ideas: "Professor L^vy-Bruhl tells us, to put it in a nutshell, that primitive man has no sober moods at all, that he is hopelessly and com pletely immersed in a mystical frame of mind. In capable of dispassionate and consistent observation, devoid of the power of abstraction, hampered by *a decided aversion towards reasoning,' he is unable to draw any benefit from experience, to construct or com prehend even the most elementary laws of nature. 'For minds thus orientated there is no fact purely physical.* Nor can there exist for them any clear idea of sub stance and attribute, cause and effect, identity and contradiction. Their outlook is that of confused superstition, •prelogical,• made of mystic •participa tions* and •exclusions.»" Malinowski, B., Magic.  Science and Religion and Other Essays. (Doubleday & Co. Inc., Garden City, New York, 1954), P. 25. On the next page of the same work, Malinowski asks, "First, has the savage any rational outlook, any rational mastery of his surroundings, or is he, as M. L^vy-Bruhl and his school maintain, entirely •mystical •? The answer will be that every primitive community is in possession of a considerable store of knowledge, based on experience and fashioned by reason." Ibid.. P. 26. Malinowski is here making no allowance for the fact that L^vy-Bruhl was constructing an ideal type of primitive thought, not attempting a description of the actual content of primitives* minds. He is, moreover, postulating an opposition between mystic thought and rationality which was quite contrary to Levy-Bruhl*s intentions. It is worth noting Levy-Bruhl*s own words: "... these charac teristics (of prelogicality) apply only to the collec tive representations and their connections. Considered as an individual, the primitive, in so far as he thinks and acts independently of these collective representa--124-tions where possible, will usually feel, argue and act as we should expect him to. The inferences he draws will be just those which would seem reasonable to us in like circumstances." Levy-Bruhl, L., How Natives Think, (Washington Square Press Inc., New York, 1966), P. 63. 3. Ibid., Pp. 4-5. 4. Ibid.. P. 12. 5. Ibid., Pp. 13—14- This parallels Durkheim's criticism of the method employed by Muller, Tylor and Prazer: "elle suppose une veritable creation ex nihilo." Durkheim, E., "Examen critique des systemes classiques sur les origines de la pensSe religieuse", Revue philo- sophique de la France et de 1'Stranger, Vol. LXVII, (Feb. 1909), P. 162. 6. LeVy-Bruhl, L., How Natives Think, op. cit., P. 3« 7. Ibid.. P. 18. 8. Evans-Pritchard, E.E., "LeVy-Bruhl's Theory of Primitive Mentality", Bulletin of the Faculty of Arts, University of Egypt, Vol. II, Part I, (May 1934), Pp. 3-4, 11-12. 9. Levy-Bruhl, L., How Natives Think, op. cit., Pp. 18-19. 10. Ibid., P. 19. 11. Ibid.. 12. Ibid., Pp. 22-6. See also P. 324, where the "most primitive" collective representations are equated with "collective mental states of extreme emotional tension." 13. See Bunzel, R.L., "Introduction" to Levy-Bruhl, L., How  Natives Think, op. cit., P. vi; also Evans-Pritchard, E.E., "Levy-Bruhl^ Theory ...", op. cit., P. 2. 14. An historical priority of primitive to civilized men tality is, however, implied. See LeVy-Bruhl, L., How  Natives Think, op. cit., Ch. IX. -125-15. Ibid., P. 63. Emphasis in original. 16. Evans-Pritchard, E.E., Theories of Primitive Religion. (Oxford University Press, London, 1965), P. 82. 17. Levy-Bruhl, L., How Natives Think, op. cit.. P. 25. 18. Evans-Pritchard, E.E., Theories .... op. cit., P. 83. 19. Levy-Bruhl, L., How Natives Think, op. cit., P. 54. See also P. 234. 20. Ibid., Pp. 40-1. The psycho-philosophical theory un derlying Levy-Bruhl *.s assertion that the perceptions of primitives differ from our own, is well expressed by Cantoni: "L*esperienza non h una passiva registrazione, in chiave umana, di una realta in se" gia data e pre-costituita. Essa presuppone e implica una soggettivita ricca di iniziative culturali che integrano l*uomo nei suo mondo ambientale. L'esperienza si viene costruendo in base a strutture categoriali, a finalita culturali che le conferiscono un senso, un ordine, un valore..•• Ricostruire il mondo dell'esperienza primitiva equivale a descriverer il sistema dei significati, dei valori, delle finalita, delle emozioni, delle immagini, dei simboli che dominano la mens e il comportamento degli uomini primitivi." ("Experience is not a passive registration, in human key, of a reality already given and preconstituted. It presupposes and implies a human subjectivity, rich in cultural initiatives which in tegrate man into his world. Experience comes to one building itself on a base of category structures and cultural purposes that confer on it a sense, order and value.... Reconstructing the world of primitive ex perience is equivalent to describing the system of meanings, values, purposes, emotions, images and sym bols which dominate the mind and behaviour of primitive man.") Cantoni, R., II Pensiero dei primitivi, (II Saggiatore, Milan, 1963), P. 54. The following un doubtedly provides a clear example of what Le>y-Bruhl meant when he wrote of the mystic perception of reality, and illustrates the "cultural initiatives which in tegrate man into his world" referred to by Cantoni: "An -126-informant told me that many years before he was sitting in a tent one afternoon during a storm, together with an old man and his wife. There was one clap of thunder after another. Suddenly the old man turned to his wife and asked, 'Did you hear what was said?' 'No,' she re plied, 'I didn't catch it.' My informant, an accul-turated Indian, told me that he did not at first know what the old man and his wife referred to. It was, of course, the thunder. The old man thought that one of the Thunder Birds had said something to him. He was reacting to this sound in the same way he would respond to a human being, whose words he did not understand. The casualness of the remark and even the trivial character of the anecdote demonstrate the 'psychological depth' of the 'social relations' with other-than-human beings that become explicit in the behaviour of the Ojibwa as a consequence of the cognitive 'set* induced by their culture." Hallowell, I.A., quoted by Hymes, D., "Toward Ethnographies of Communication: The Analysis of Communicative Events", in Giglioli, P.P. (ed.), Language and Social Context. (Penguin Books Ltd., Harmondsworth, 1972), Pp. 26-7. LeVy-Bruhl and Cantoni were undoubtedly correct to affirm that the perceptions of primitives are governed by cultural imperatives, and that the primitive therefore lives in a different ex periential world from our own. They erred, however, in failing to recognize that our own perceptions are also governed by cultural imperatives. Thus Cantoni argues that our experience of natural reality is governed by logico-experimental structures which are ordered accord ing to the ideals of "pure reason". This he contrasts with primitive experience, in which emotional factors predominate. Cantoni, R., H pensiero..., op. cit., Pp. 54-5. It would, however, surely be difficult to sustain that, in the Ojibwa case quoted above, emo tional factors predominate over rational. Moreover, the logico-experimental structures referred to by Cantoni do not exist in vacuo, but are the product of a definite evolution hardly explicable by reference to logico-experimental criteria alone. Prom Kuhn's dis cussion of scientific method, it is plain that scien tists do not merely perceive, but to an important ex--127-tent perceive according to their prior expectations. Thus a change in theory may lead to the reporting of a wide range of "new" phenomena. Kuhn cites the example of the new astronomical discoveries - many of them made with traditional instruments - which followed on the triumph of the Copernican paradigm: "Using traditional instruments, some as simple as a piece of thread, late sixteenth-century astronomers repeatedly discovered that comets wandered at will through the space pre viously reserved for the immutable planets and stars. The very ease and rapidity with which astronomers saw new things when looking at old objects with old instru ments may make us wish to say that, after Copernicus, astronomers lived in a different world. In any case, their research responded as though that were the case." Kuhn, T.S., The Structure of Scientific Revolutions, (University of Chicago Press, Chicago, 1962), Pp. 115-6. 21. Levy-Bruhl, L., How Natives Think, op. cit., Pp. 21-2. 22. Evans-Pritchard, E.E., Theories .... op. cit., P. 86. 23. In the following evaluation, I have drawn heavily on Evans-Pritchard, E.E., "Le>y-Bruhl»s Theory ..."; on. cit.; Theories ..., op. cit.. 24. Evans-Pritchard, E.E., "Levy-Bruhl's Theory ...", op.  cit., P. 8. 25. Bunzel states, on the basis of private discussions, that Levy-Bruhl was prepared to consider the Mass as an example of prelogical thought. Bunzel, R., "Introduc tion", op. cit., P. vi. Evans-Pritchard, also on the basis of private conversations, declares that Levy-Bruhl considered Christianity and Judaism to be ex amples of prelogical and mystic mentality. He suggests that Levy-Bruhl made no allusions to these religions in his writings, in order to avoid giving offence. Evans-Pritchard, E.E., Theories...., op. cit.. P. 91-26. Pareto, V., The Mind and Society, (Harcourt Brace, New York, 1935). See also, Evans-Pritchard, E.E., "Science -128-and Sentiment: An Exposition and Criticism of the Writings of Pareto", Bulletin of the Faculty of Arts, University of Egypt, Vol. Ill, Part II, (Dee. 1935), Pp. 163-92; Theories..., op. cit.. Pp. 92-9; Winch, P., The Idea of a Social Science, (Routledge & Kegan Paul, London, 1958), Pp. 95-111. 27. Evans-Pritchard, E.E., "Levy-Bruhl's Theory...", op.  cit., P. 28. 28. See, for example, Levi-Strauss• analysis of the Nuer affirmation that twins are birds. Levi-Strauss, C., Totemism, (Merlin Press, London, 1962), P. 80. 29. Evans-Pritchard, E.E., "Levy-Bruhl1s Theory...", op.  cit., Pp. 28-9. 30. LeVy-Bruhl, L., Les carnets de Lucien Levy-Bruhl, (Presses Universitaires de France, Paris, 1949)* 31. Ibid., P. 131. Quoted in Bunzel, R.L., "Introduction", op. cit., P.-xvii. CHAPTER FIVE W.H.R. RIVERS A. Introduction The work of W.H.R. Rivers is frequently neglected to day, although Claude Levi-Strauss has seen fit to honour him 1 as the "Galileo" of anthropology. Probably few other an thropologists would accept such a generous assessment of 2 Rivers* importance in the development of the discipline, but it is nevertheless undeniable that in the opening chap ters of his Medicine, Magic and Religion. Rivers brilliantly anticipated later interpretations of witchcraft and sorcery as constituting rational and logically coherent systems of thought. Unlike Prazer, Rivers did not set out to analyse magic as an a priori category which he himself constructed and for which he then had to find some sort of explanation. Rather, he undertook to examine the way in which primitive peoples (and in particular, those of Melanesia and New Guinea) think about disease and other misfortunes, and to study how their theories concerning these are applied in concrete situa tions. In doing so, he made an important advance on Tylor, Prazer and LSvy-Bruhl, and his work may even,be considered to have been potentially more fruitful than that of Malinowski. B. Medicine, Magic and Religion Rivers began by noting that the distinctions drawn in 130-Western society between magic, medicine and religion, have kittle applicability outside this context, the roles of sor cerer (a term that he, like Frazer, used to refer to the practitioner of any form of magic), priest and leech (medi cal practitioner) frequently being fused in other cultural contexts.^ This observation, however, did not lead Rivers into attempting a polarization of primitive and modern men talities, nor even to attempt a polarization of respective curative techniques. On the contrary, Rivers stressed that primitive methods of curing could be understood in basically the same terms as European medicine: that is to say, as con stituting a logical, coherent system, which attempts to cope rationally with diseases and illnesses, and comprehensible once its basic presuppositions are understood. One element of the concept of disease, and perhaps the most impor tant, is that it includes within its scope the factor of causation. There are usually clear-cut ideas concern ing the immediate conditions which lead to the appearence of disease. One happy result of this fact is that we are able to approach our subject by way of etiology, and are thus led to deal with the medicine of savage peoples from the same standpoint as that of modern medi cine, which rests, or should rest, entirely upon the foundation of etio logy. By starting from etiology we shall find ourselves led on as natu rally to diagnosis and treatment, as is the case in our own system of medicine.* Thus, even where the roles of leech, priest and sorcerer are fused, there are still theories of the causation of ill--131-ness and other misfortunes, procedures corresponding to diag nosis and prognosis, and modes of treatment which may he re-5 garded as equivalent to a system of therapeutics. But, granted this, where lies the difference between European and primitive medicine, or, to pose this question in only a slightly different form, what did Rivers really mean when he stated that the primitive leech is not differentiated from the priest and the sorcerer? Rivers1 answer to this question, perhaps the obvious an swer, was that the basic difference lies in the ascription of the causes of disease. In accordance with this, he suggested three broad classes of etiology: i. where disease is attributed to what we would call "natural causes"; ii. where disease is attributed to the actions of a human agent; iii. where disease is attributed to the actions of some non-human, spiritual or supernatural being, who tends to be per sonified.^ In Europe and North America, it is no longer common to ascribe human causes to the origins of disease, except in cases like that of murder. Moreover, even with the latter case, the human agent is seen as being merely a means per mitting the operation of natural causes. In contrast to this, primitive peoples ascribe the causes of disease mainly 7 to the actions of human agents or non-human beings. Primitive etiologies, in other words, tend to be magico-religious. Diseases and injuries are ascribed to human and non-human agencies, and not only in those cases where there -132-is no obvious antecedent which would explain the disease or misfortune in terms of natural causes, but also in those cases where the natural cause is obvious: Thus, if a man is killed or injured by falling from a tree in the Island of Ambrim in the New Hebrides, the fall is not ascribed to a loose branch, or to some failure of co ordination of the movements of the climber, but the accident, as we loosely call it, is put to the ac count of the sorcerer. It is prob able that the sequence of ideas in the Melanesian mind is that, in a business so familiar as that of climbing trees, accidents would not happen unless someone has interfered with the normal course of events. If a sorcerer had not loosened a branch, or produced an illusion whereby the victim had seen a branch where there was none, he would not have fallen to the ground." From this etiological base, are derived processes of 9 treatment and acts of revenge. If a disease, for example, is believed to have resulted from some morbific substance or essence having been projected into the victim's body, the treatment which will follow from the etiology will be to re move the morbific objects or essences from the body of the victim. In some cases, it may be necessary to discover the agent, who alone can remove what he has implanted in the body. In such cases, some method of divination may have to be employed in order to discover by whom the disease has been 10 inflicted. In other words, once native theories of the causation of disease are understood, it will be possible to -133-understand modes of treatment of it. The main point, then, is the essential rationality of the medical procedures of such peoples as the Melanesians and Papuans. The practices of these peoples in relation to disease are not a medley of disconnected and meaningless cus toms, but are inspired by definite ideas concerning the causation of disease. Their modes of treatment follow directly from their ideas concerning etiology and pathology. Prom our modern standpoint we are able to see that these ideas are wrong. But the important point is that, however wrong may be the be liefs of the Papuans and Melanesians concerning the causation of disease, their practices are the logical con sequence of those beliefs. We may even say that these people practise an art of medicine which is in some respects more rational than our own, in that its modes of diagnosis and treatment follow more directly from their ideas concerning the causation of disease. According to the opinion of the civilized world, these ideas of causation are wrong, or contain but grains of truth here and there; but once grant these ideas, and the body of medical practice follows therefrom with a logical consistency which it may take us long to emulate in our pursuit of a medicine founded upon the sciences of physiology and psychology.^1 C. A Critique of Rivers Essentially a transitional figure, however, it would be wrong to over-emphasize the modernity of Rivers* views. In 134-many respects his trichotomy between leech, sorcerer and priest is little more than a restatement of Frazer's dif ferentiation between magic, science and religion. The dif ference between leech and sorcerer, moreover, is not made any clearer than Frazer*s rather obscure distinction between magic and science, except that the leech is concerned with a narrowly specific range of phenomena (i.e., disease), the 12 sorcerer with a broader range. But how is such a dif ferentiation to be sustained when, as Rivers himself makes clear, leech and sorcerer may be merged in a single role? Yet, despite the similarity of his trichotomy to Frazer's, and despite his explicit restatement of Frazer's differentiation between magic and religion,1^ Rivers never theless attacked Frazer for implying that the basis of ma gical action rested on an abstract or even mystical body of ideas, "... opposed to the concrete nature of the mental 14 processes of peoples of rude culture." Thus the contagious magic of the Kai rests, not on some mystical belief in ac tion at a distance, but on the belief that the sorcerer has in his possession a part of the soul or vital essence of the person whom he wishes to destroy. Such positive knowledge as we pos sess concerning the psychological processes underlying the blend of medicine and magic leads us into no mystical dawn of the human mind, but introduces us to concepts and beliefs of the same order as those which direct our own social activi ties.15 This view, of course, was totally opposed to that of Levy-Bruhl. Hence, while recognizing that a given corpus -135-of beliefs and practices might not always form a strictly logical system (our own beliefs and practices included), Rivers strongly attacked LSvy-Bruhl's notion of prelogical mentality: ... in the department of his acti vity in which he endeavours to cope with disease, savage man is no il logical or prelogical creature ... his actions are guided by reasoning as definite as that which we can claim for our own medical prac tices.^ Stated briefly, Rivers' strength in relation to pre ceding and contemporaneous theorists may be said to hinge on four main facts. Firstly, as against Tylor and Frazer, he placed a strong emphasis on the relative unimportance of charlatanism in the magical treatment of disease: this was an important point since, so long as the magical specialist was regarded as little more than a wiley trickster cashing-in on the naive superstitions of his credulous fellow-tribesmen, the understanding of the intellectual structure of magical beliefs, together with their role in a given way of life, was almost by definition ruled out. Secondly, again in contrast to Tylor and Frazer, Rivers did not regard magic as based upon a simple mistake in logic, but rather insisted that magic was governed by a logic quite as rigour-ous as that governing our own practical activities. A con comitant of this point is that, in contrast to Malinowski, Rivers did not assign magic to the affective antechamber of logic. Finally, unlike LSvy-Bruhl, Rivers did not oppose primitive and civilized mentalities. -136-Yet Rivers nowhere developed his ideas on this subject in detail. Medicine. Magic and Religion was published pos thumously, and is today largely unknown. Moreover, the later chapters of this work are given over to studies of the evolution and diffusion patterns of medicine, magic and re ligion, and not to an elaboration of the ideas contained in the earlier sections. We are left, therefore, with a bril liant and suggestive short sketch - nothing more. -137-Notes and References 1. Lgvi-Strauss, C, Structural Anthropology, (Basic Books, - New York, 1963), P. 162. 2. A more typical assessment of Rivers' importance in the history of anthropology would seem to he that of Pocock: "To Rivers is owed the encouragement of field-work and the stress on its capital importance for the anthropolo gist. His theoretical position represents little ad vance in the subject." Rivers is thus grouped together with such figures as A.C. Haddon and C.G. Seligman. Cf. Pocock, D.P., Social Anthropology. (Sheed & Ward, London, 1961), P* 48. 3. Rivers, W.H.R., Medicine, Magic and Religion. (Kegan Paul, Trench, Trubner & Co. Ltd., London, 1927), P. 1. 4. Ibid., P. 7. 5. Ibid., P. 6. 6. Ibid., P. 7. 7. Ibid., PP . 7-8. 8. Ibid., Pp . 9-10. 9. Ibid., P. 10. 10. Ibid., P. 14. 11. Ibid.. P. 52. 12. Ibid., P. 4. 13. Ibid.. 14. Ibid., P. 27. 15. Ibid., P. 28. 16. Ibid., P. 53. CHAPTER SIX AN AFFECTIVE THEORY OF MAGIO A. Introduction If any one figure was to be singled out as the "foun der" or "father" of modern (British) social anthropology, that figure would undoubtedly be Bronislaw Malinowski. It was Malinowski more than any other single person, who was responsible for dissociating anthropology from its nine teenth and early twentieth century evolutionary and dif-fusionist concerns, who insisted on the crucial importance of seeing every culture as a unified, integral whole, and who emphasized that each cultural pattern could only be un derstood by locating it within the context of this whole. It was also Malinowski who initiated intensive field-work based upon the idea of the "participant observer", thus freeing anthropology from its former dependence on the second-hand reports furnished by missionaries, traders, and administrators, and the superficial observations which were the outcome of only a few weeks in the field, like those of the Torres Straits expedition. Pre-Malinowskian anthropology was explicitly compara tive, and was concerned with what we might call the "big" questions: what are the origins of totemism, what is the nature of religion, how is magic distinguished from science, for example. Under Malinowski*s influence, anthropology abandoned such concerns. As Evans-Pritchard explains, -139-Whereas the nineteenth-century anthro pologist sought to answer such ques tions as "What is the sociological significance of religion?", no anthro pologist, or at any rate no sensible anthropologist, would ask such a ques tion today. Rather he seeks- ta deter mine, for instance, the part played by the ancestor cult in the social system of the type we call segmentary lineage system among certain African peoples.... The viewpoint in social anthropology today may be summed up by saying that we now think we can learn more about the nature of human society by really detailed, intensive and observational studies, conducted in a series of a few selected societies: with the aim of solving limited problems, than by at tempting generalizations on a wider scale from, literature.-*-In retrospect, however, most "revolutionaries" usually appear less radical than they did in their own time. . More over, as a number of authors have pointed out, Malinowski*s meticulous field-work was not matched: by any comparable capacity for systematic theoretical analysis. His thought may therefore be said to move on two levels: that of his field-work reports (which still stand a® models; of their kindl), and that of culturological generalization. The quality of these is very unequal. As Pocock remarks, "today nocbody reads Malinowski for his theories; his field, monographs, on the other hand ... remain as masterpieces of fieldi-work and provide an account of one people which for 2 its detail has not been equalled". As a theoretician, Malinowski was among the last to -140-ask the "big" questions. As a field-worker, he showed anthropologists some of the more fruitful and restricted tasks of empirical research and concern with the particular rather than the universal. Obviously there must be a con nection between these levels - someone must ask the "big" questions if empirical work is to proceed systematically and fruitfully. But the relation between the two is com plex, and the problems of transition from one level to the other never presented themselves to Malinowski. He was, in this respect, an essentially transitional figure. As Nadel writes, Putting it somewhat crudely, Malinowski's thought moved on two levels only - on the level of the particular society, the Trobriands, where he did his fundamental and exemplary field research; and on the level of primitive man and society at large, and indeed of Man and Society at large. In his more general writ ings Malinowski did refer also to other primitive societies; but he did so in the main only for the sake of supporting evidence, of secondary im portance. He never thought strictly in comparative terms. His generali zations jump straight from the Trobrianders to Humanity...3 B. Malinowski's Psychologism As Robert Redfield observes, however, the criticism that Malinowski made a theoretical leap from the Trobriand Islanders to humanity itself loses much of its force once it is granted that a common human nature underlies a uni--141-versal culture pattern. It was, of course, precisely this assumption that was central to Malinowski1s theore tical endeavour. He never accepted the methodology elabo rated by writers like Durkheim, Mauss, Hubert, Levy-Bruhl and Radcliffe-Brown, which asserted the autonomy of social facts. In his own words, The tendency represented largely by the sociological school of Durkheim, and clearly expressed in Professor Radcliffe-Brown*s approach ... the tendency to ignore completely the individual and to eliminate the bio logical element from the functional analysis of culture, must in my opin ion be overcome.... the Durkheimian conception of society has to be sup plemented in order to be really ser viceable in fieldwork, in theoreti cal studies, and in the practical application of sociology. Malinowski did not, therefore, reject the psycho logical reductionism of Tylor and Prazer, and - perhaps in consequence - did not develop any systematic sociological -theory, other than the rather vague "functionalism" with which his name will always be associated. Hence his gene ralizations about magic were formulated in terms of assum ed psychological universals, rather than sociologically. It is important to note, however, how different was Malinowski*s psychologism from that of Tylor and Prazer. The post-Victorian era, and especially the period follow ing the First World War, saw a deprecation in the role reason was thought to play in human affairs. Thus, in place of the intellectualistic interpretations of magic -142-and religion offered by Tylor and Frazer, we find writers like Freud, Marett, Malinowski and Lowie interpreting ma gic and religion as essentially emotive phenomena. C. The Uniformity of Primitive Magic We have already noted that Frazer considered magic -in comparison to religion - to constitute a universal sub-stratum of intellectual uniformity. Much the same opin ion was held by Malinowski, who considered magic: ... an entirely sober, prosaic, even clumsy art; enacted for purely prac tical reasons, governed by crude and shallow beliefs, carried out in a simple and monotonous technique.... Primitive magic - every field an thropologist knows it to his cost -is extremely monotonous and unexcit ing, strictly limited in its means of action, circumscribed in its be liefs, stunted in its fundamental assumptions. Follow one rite, study one spell, grasp the principles of magical belief, art and sociology in one case, and you will know not only all the acts of the tribe, but, add ing a variant here and there, you will be able to settle as a magical practitioner in any part of the world yet fortunate enough to have faith in that desirable art.7 Hence the justification for generalizing from a single case and for postulating a common psychological base underlying all magic. For Malinowski was not re ferring here to a set of shared basic features which might be used to establish a "minimum definition" of ma gic and which must be uniform precisely because of this. -143-Por Malinowski, magic was uniform and repetitive in an en tire array of features. Thus the "typical act" of magic always involves "the dramatic expression of emotion" as g "the essence of the act", and most magic also bears a firm 9 link with mythology. In addition to this, there are al ways two typical elements associated with the belief in ma gical efficiency: i. phonetic effects - imitations of natural sounds like the whistling of the wind, the roar of the sea and the noises of animals. These sounds symbolize certain pheno mena and are thus believed to produce them magically. Or else they express certain emotional states associated with the desire which is to be rea lized by the means of magic.^ ii. the use of words to state, evoke or command the desired aim. Thus the sorcerer will mention all the symptoms of the disease which he is inflicting, or in the lethal for mula he will describe the end of his victim.... Or again, the magician uses words and sentences which express the emotion under stress of which he works his magic, and the action which gives expression to his emotion. The sorcerer in tones of fury will have to repeat such verbs as "I break - I twist - I burn - I destroy," enume rating with each of them the various parts of the body and internal organs of his victim.11 Malinowski in fact saw the use of words - embodied in the traditional corpus of the spell - as the most important -144-single characteristic of primitive magic: the most important element in ma gic is the spell. The spell is that part of magic which is occult, handed over in magical filiation, known only to the practitioner. To the natives knowledge of magic means knowledge of spell, and in an analysis of any act of witchcraft it will always he found that the ritual centers round the ut terance of the spell. The formula is always the core of the magical per formance .12 Unfortunately the facts of ethnography utterly fail to hear out these generalizations. In 1924, one year prior to the publication of Malinowski*s essay on "Magic, Science and Religion", Robert Lowie had already pointed out - in criticizing Frazer - how extensive the diversity in magical 13 beliefs and practices appears when considered globally. Subsequent research has entirely confirmed Lowie*s view point and not Malinowski's. Thus Evans-Pritchard, in an early comparative essay, drew attention to the marked dif ferences between Trobriand and Zande magic. In Zande ma gic, for example, the spell is relatively unimportant and lacks the fixedness which is so heavily stressed in the Trobriand Islands, while on the other hand, the Azande place a much greater emphasis on the use of "medicines". Zande magic also lacks any developed mythological tradi-14 tion. Indeed, even some of Malinowski's own data from the Trobriand Islands is hard to square with his broad generalizations about the nature of magic - the heavy em phasis he placed on emotional involvement during magical performances, for example, does not seem to be consistent -145-with some of the data he himself reported. However, as Nadel says, it is less important that we concern ourselves with these generalizations than that we attempt to understand the assumptions which underlie them. These, he suggests, are two: i. Malinowski's emphasis on the strong link between magic and mythology derives from his belief that magic is essen tially a traditionalistic and conservative force protecting the established social order, and from his conclusion that magical formulae must be thought to be timeless and never to have been tampered with (even if actually changing and adapting all the time), for credence in their effectiveness to be maintained. Thus it was that Malinowski wrote that: ... the essence of all magic is its traditional integrity. Magic can on ly be efficient if it has been trans mitted without loss and without flaw from one generation to the other, till it has come down from primeval times to the present performer. Ma gic, therefore, requires a pedigree, a sort of traditional passport in its travel across time. This is supplied by the myth of magic.^ ii. the emphasis placed on the affective nature of magic de rives from Malinowski's assumption that magic manifests an inevitable human effort to achieve the fulfillment of 17 "strong, unrealizable desires". P. Anxiety and Magic According to Malinowski, in every primitive community two clearly discernible domains may be distinguished - that of the sacred, and that of the profane. The former involves -146-all beliefs and actions of a magical or religious charac ter: ... traditional acts and obser vances, carried out in reverence and awe, hedged around with prohibitions and special behaviour. Such acts and observances are always associat ed with beliefs in supernatural forces, especially those of magic, or with ideas about beings, spirits, ghosts, dead ancestors, or gods..."'" The domain of the profane, on the other hand, corres ponds to the rudiments of science. No art, eraft, or me thod of subsistence could be successfully carried out un less checked by careful observation, a firm belief in the regularity of natural processes, and confidence in the power of reason. The profane is therefore based on empiri cal knowledge and an acceptance of logic. Now, according to Malinowski, primitive man lives in conditions in which the technical skills and practical knowledge constituting the realm of the profane, provide* him with only a limited measure of control and mastery of his environment. Beyond this realm, rational knowledge is of no help, and it is here that magic operates as an essential complement to empirical knowledge and skills. Thus the native: ... knows as well as you do that there are natural forces by mental and physical effort. His knowledge is limited, no doubt, but as far as it goes it is sound and proof against mysticism. If the fences are broken down, if the seed is destroyed or has been dried or washed away, he will have recourse not to magic, but 1 -147-to work, guided by knowledge and reason. His experience has taught him also, on the other hand, that in spite of all his forethought and be yond all his efforts there are agen cies and forces which one year be stow unwonted and unearned benefits of fertility, making everything run smooth and well, rain and sun appear at the right moment, noxious insects remain in abeyance, the harvest yields a superabundant crop; and an other year again the same agencies bring ill-luck and bad chance, pur sue him from beginning to end and thwart all his most strenuous ef forts and his best-founded knowl edge. To control these influences and these only he employs magic.^ Por Malinowski there was between these two sets of conditions - and therefore in the means for controlling them - a clear-cut division. Hence, empirical knowledge must always be sharply distinguished in the native mind from magic. As Leach remarks, in asserting this dichotomy between the objective-rational and the subjective-meta physical, Malinowski was following in the tradition of 20 Tylor and Prazer strictly. Yet Malinowski also departs radically from the tradition of Tylor and Prazer. Por these writers, primitive man was incapable of this cate gory distinction, which was given only in more highly evolved societies. Indeed, it was crucial for Prazer's theory of magic as "bastard science" to assume that primi tives did not differentiate between the empirical and the magical. Malinowski, on the other hand, argued that the primitive was just as capable as civilized man of making -148-such distinctions, though the force of his argument must surely he diminished by his own admission that he himself was not always able to judge "where rational procedure ended and which were the supergatory activities, whether magical or aesthetic".21 Leaving this to one side, however, let us consider why it was that Malinowski contended that primitive man invoked magic in order to control those activities he re cognized he was incapable of controlling by means of empi rical knowledge and rational thought. In answering this question, Malinowski argued that magic constitutes a sub stitute activity, originating in a reaction of fear, hope and anxiety induced by the inadequacy of empirical knowl edge and rational means: Let us realize once more the type of situation in which we find ma gic. Man engaged in a series of practical activities comes to a gap; the hunter is disappointed by his quarry, the sailor misses pro pitious, winds, the canoe builder has to deal with some material of which he is never certain that it will stand the strain, or the healthy person suddenly feels his strength failing. What does man do naturally under such conditions, setting aside all magic, belief and ritual? Forsaken by his knowledge, baffled by his past experience and by his technical skill, he realizes his impotence. Yet his desire grips him only the more strongly; his anxiety, his fears and hopes, induce a tension in his organism which drives him to some sort of -149-aetivity. Whether he be savage or civilized, whether in possession of magic or entirely ignorant of its existence, passive inaction, the only thing dictated by reason, is the last thing in which he can ac quiesce . His nervous system and his whole organism drive him to some substitute activity. Obsessed by the idea of the desired end, he sees it and feels it. His organism reproduces the acts suggested by the anticipations of hope, dictated by the emotion of passion so strongly felt.22 Thus a man will spontaneously clench his fist at the thought of an enemy whom he is powerless to harm and hurl words of hatred against him, a lover will address en treaties to visions of his unattainable or non-responsive beauty, an anxious fisherman or hunter will imagine and describe in words, visions of a magnificent catch. According to Malinowski, these reactions to over powering emotion or obsessive fear or desire "are natural responses based on a universal psycho-physiological me-23 onanism". All the spontaneous actions of a man in a situation of frustration involve extended expression of emotions by forecasting images of wished-for results or the expression of passion in gestures or words. More over, governing the entire action, is the image of the end: It supplies the motive-force of the reaction, it apparently organizes and directs words and acts towards a definite purpose. The substitute action in which the passion finds -150-its vent, and which is due to impo- . tence, has subjectively all the value of a real action, to which emotion would, if not impeded, naturally have led. As the tension spends itself in these words and gestures the obsess ing visions fade away, the desired end seems nearer satisfaction, we re gain our balance, once more at har mony with life,24 These spontaneous enactments are therefore cathartic, in providing a release for frustrated emotions. Moreover, in thus acting out intense emotional states, a strong feeling is obtained that the words and gestures have actu ally done something towards achieving the desired goal, thus leaving a conviction of the reality of these acts, ... as if of something done by a i power revealed to man. This power, born of mental and physiological ob session, seems to get hold of us from the outside, and to primitive man, or to the credulous and untutor ed mind of all ages, the spontaneous spell, the spontaneous rite, and the spontaneous belief in their efficien cy must appear as a direct revelation from some external and no doubt im personal sources.25 As Malinowski himself emphasized, however, magic is not a matter of "spontaneous spells" or of "spontaneous rites", but is an activity in which the words uttered and the rites enacted are codified by tradition. One cannot simply invent magic - one must have a knowledge of the ap propriate spells or medicines, and of the correct method of bringing spell or medicine into contact with its 26 object. ' It would seem then, that there is a gulf be-151-tween the spontaneous activities described by Malinowski, and magical lore. Yet Malinowski argued a close kinship between these spontaneous reactions and the traditions of magic. "Kin ship", indeed, seems hardly a strong enough word to ex press the strength of the relationship as Malinowski saw it: emotional reaction is "not only one of the sources but 27 the very fountainhead of magical belief". Magical ritual, most of the princi ples of magic, most of its spells and substances, have been revealed to man in those passionate experi ences which assail him in the impas ses of his instinctive life and of his practical pursuits, in those gaps and breaches left in the ever-imperfect wall of culture which he erects between himself and the be setting temptations and dangers of his destiny.... Thus the founda tions of magical belief and practise are not taken from the air, but are due to a number of experiences actu ally lived through, in which man re ceives the revelation of his power to attain the desired end.28 Magic therefore fixes upon these spontaneous rites and spells, and standardizes them into traditional forms. By doing this, it supplies primitive man with a definite means of handling those situations he must deal with in every important pursuit or activity where his empirical knowledge is insufficient. In this way, it exercises an important pragmatic function, in that it enables man to carry out his important tasks with confidence, and to maintain his poise and mental integrity when subject to 152-strong emotional states. Magic thus ritualizes man's op timism, and by doing so encourages confidence in place of doubt, determination in place of hesitation and optimism in place of pessimism. Thus, despite "all the crudity and 29 irrelevance of magic", it is of fundamental importance within primitive society: ... without its power and guidance early man could not have mastered his practical difficulties as he has done, nor could man have advanced to the higher stages of culture. Hence the universal occurrence of magic in primitive societies and its enormous sway. Hence do we find magic an in variable adjunct of all important activities.30 Magic therefore ... fulfills an indispensable func tion within culture. It satisfies a definite need which cannot be satis fied by any other factors of primi tive civilization.31 E. Black Magic Malinowski's general theory of magic is perhaps weak est of all when called upon to explain the facts of witch craft and sorcery. Por, if the function of magic is to master the risks of the environment, why add to them by 32 inventing sorcery and witchcraft? Malinowski himself offered two reasons for the exist ence of black magic. Firstly, it is a natural human re action to frustrated hate and impotent anger, being in this way comparable to a similarly impassioned love magic: ... such is human nature that a -153-man's desire is as much satisfied by the thwarting of others as by the advancement of himself. To the sociological play of desire and counter-desire, of ambition and spite, of success and envy, there corresponds the play of magic and counter-magic or of magic white and black.33 Secondly, Malinowski contended that the belief in magic must safeguard itself in certain ways against being weakened by the absence of success. One of the ways in which it does this is, as Tylor had argued long before, by invoking the belief in black magic. Every magic has its counter-magic, of which the destructive magic prompted by 34 evil desires is an example. But, as Nadel says, these explanations are incidental rather than systematic and fail to treat the evil use of magic as a problem in its own right. In borrowing from a common sense psychology, Malinowski ignored the fact that sociological and ethical questions, no less than psycho logical, are posed by the use of black magic: Por if a society acknowledges the presence of occult destructive powers in its midst, that is, of agencies threatening its norms and stability yet available to its mem bers, this must indicate that the structure of the society itself in vites or even requires the presence of these agencies ...35 Malinowski's failure to deal with this problem in a satisfactory manner, not only sociologically but even psy chologically, is borne out by his treatment of witchcraft -154-in particular* In this respect, he stands in striking contrast to Evans-Pritchard and Kluckhohn. Thus, despite the fact that Malinowski based his general theory of magic directly on his Trobriand material, he wrote little about the Trobriand yoyova except to emphasize that the beliefs concerning her do not form a consistent body of knowledge and that, being in this way confused in the minds of the Trobrianders, these beliefs provide another example of how, "The native feels and fears his belief rather than 36 formulates it clearly to himself." This failure of treatment is also reflected in Malinowski's final formulation of the utilitarian value of magic to the primitive community and in his conclusion that magic constituted a necessary step in the course of human evolution. In this final formulation, the negative side of magic no longer finds any mention. Likewise, as Leach points out, "In Coral Gardens, where magical beliefs and practises are dealt with at length and treated as functionally positive practical working tools, the exis tence of witchcraft beliefs is completely ignored, the word 'witchcraft* being used merely as a synonym for sor-37 eery, in the sense of negative magic." P. The Politico-Legal Functions of Sorcery Malinowski's remarks on the Trobriand bwaga'u were somewhat more extensive than those he devoted to the yoyova. Doubtless, this was because he was able to recon cile the corpus of beliefs concerning the bwaga'u more easily with his theoretical schema. In discussing sorcery, Malinowski emphasized its cha--155-racter as a source of power, wealth and influence and, as such, how it tends to function in such a way as to perpe tuate the traditional social order. It is chiefs and men of rank who have first claim on the services of the Trobriand sorcerer, and he would not lend himself to the unjust requests of lesser men. Too rich and powerful to do anything outside the law, it is mainly when it is a question of punishing the wrong-doing of another that he will accept a fee and champion a cause. In many such cases, on learning that the sorcerer is work ing against him, the wrong-doer will take fright and hasten to make amends. "Thus ordinarily, black magic acts as a genuine legal force, for it is used in carrying out the rules of tribal law, it prevents the use of violence and 38 restores equilibrium." Sorcery is especially important in maintaining the authority of the chief. Before the coming of the whites, the chief was able to use direct physical violence against those guilty of a direct breach of etiquette or ceremony, or guilty of flagrant offenses like adultery with one of his wives. Today, sorcery is the main method of enforcing the exclusive rights and privileges of the chief. Sorcery is therefore ranged on the side of the powerful and the wealthy as a support of vested interest and, in the long run, of law and order. "It is always a conservative force, and it furnishes really the main source of punishment and 39 retribution indispensable in any orderly society." More over, while sometimes used to commit wrongs against a weak er man on behalf of one more powerful, sorcery is never em--156-ployed in direct opposition to the law. Rather, sorcery is a way of emphasizing the status quo, of expressing old in equalities and counter-acting the formation of new ones. Moreover, "... since conservatism is the most important trend of primitive society, sorcery on the whole is a bene-40 ficient agency, of enormous value for early culture." Malinowski's account of the functional utility of sor cery to Trobriand society does not therefore conflict with his general theory of magic. Sorcery is "useful" to Trobriand society, and the attempts of white missionaries and administrators to stamp it out are misguided. Yet the very brevity of these remarks is an indication of the poverty of Malinowski's ideas on this subject. C. A Critique of Malinowski Malinowski's failure to deal satisfactorily with the phenomena of witchcraft and sorcery does not, of course, represent a simple theoretical aberration, but derives from the imperfections of his general theory of magic. The foundation-stone of Malinowski's theory is the view that magic represents a cathartic response to the psychological tensions which are generated in situations of danger and uncertainty. It is a logical deduction from this theory that, since the risks and dangers associated with different undertakings are not equally great in all spheres of life, little or no magic should be associated with those activities which are attended by few risks or dangers while, conversely, especially risky activities ought to be permeated with magical beliefs and practices. 157-This deduction is in fact consistent with the data we are presented on Trobriand fishing magic: While in the villages on the inner lagoon fishing is done in an easy and absolutely reliable manner by the method of poisoning, yielding abundant results without danger and uncertainty, there are on the shores of the open sea dangerous modes of fishing and also certain types in which the yield greatly varies according to whether shoals of fish appear beforehand or not. It is most significant that in the lagoon fishing, where man can rely completely upon his knowledge and skill, magic does not exist, while in the open-sea fishing, full of danger and uncertainty, there is extensive magical ritual to secure safety and good results.*1 This example seems to clinch Malinowski's argument, un til we remember that the adequacy of a theory must be judged on the basis of far more than a single case. And when we turn to consider other examples of magical belief and acti vity, it becomes plain that Malinowski*s theory is far less satisfactory than it might seem at first sight. To begin with, not all magic can be assigned with equal facility to the function of dealing with the risks, dangers and otherwise uncontrollable events attending an undertak ing. So well-grounded a technology as gardening on a tropi cal island, for example, faces few risks and should there fore stand in little need of magical support. There is al ways, of course, the possibility of drought and crop fail ure, but these risks are remote when compared to those at--158-tending such activities as voyaging on an open sea to dis tant islands. Yet there is a rich magical lore associated with Trobriand gardening. One might, of course, try to salvage Malinowski's argu ment by pointing out that, in the Trobriand Islands, the production of food enters deeply into the context of inter personal relations, and is important in the renewal of per sonal bonds and the winning and maintenance of prestige. In consequence of this, gardening is an activity which engages strong emotional involvement, and thus gives rise to pas sions which greatly magnify the gravity of risks and chance effects. Advancing such an argument, however, requires the making of a rather important adjustment to Malinowski's theory. As Nadel comments, ... the risks involved are in a sense artificial ones. This is no longer a question of the tragic shortcomings of human inventiveness in the face of nature. Rather, ma gic serves to protect a particular people from failures which are fail ures only because their own social system has decreed them to be such and, in a sense, invented them. Differently put, magic is a tool used in the realization of the for tuitously given social values, not only a weapon in man's eternal fight against fate.*2 The same argument applies to another form of Trobriand magic which seems frivolous, or even unwarranted, in terms of Malinowski's theory. This is the magic which is used in order to achieve excellence in dancing, and to protect the dancer from the envy and black magic of his rivals. As -159-Nadel again comments, It is clearly the society itself which decrees that this excellence is desirable and enviable; so that the society invents both the desire and the tragedy of its frustration, and then, in addition, has to invent also the supernatural device to save people from a risk of their own mak ing.43 This criticism stands close to that levelled at Malinowski's theory of magic by A.H. Radcliffe-Brown. Repud-44 iating Malinowski*s distinction between magic and religion? Radcliffe-Brown argued that one could equally validly con tend the opposite opinion to Malinowski with regard to some examples of ritual enactment. That is to say, rituals are not necessarily the product of individual feelings of inse curity and danger but, on the contrary, may even create such feelings in the individual. Radcliffe-Brown supports this argument by refering to the food taboos which are associated with childbirth in the Andaman Islands. Here, when a woman is expecting a child, and for some weeks following the birth of the baby, both she and her husband must abstain from the eating of certain foods which are permitted in normal cir cumstances. These prohibited foods are the flesh of turtle, dugong and pig. Should the taboo be violated, it is be lieved that the person responsible, and perhaps also the child, will fall ill. Hence Radcliffe-Brown•s conclusion that, .•• while one anthropological theory is that magic and religion give men confidence, comfort and a sense of security, it could equally well be -160-argued that they give men fears and anxieties from which they would otherwise he free ..."45 Eadcliffe-Brown would seem to have pin-pointed a fatal flaw in Malinowski's theory but, as Homans has shown, by dint of further modification the theory can be salvaged and incorporate the objection to it. Homans points out that im plicit in Malinowski's theory (that magical ritual provides an outlet for tensions generated in situations of uncertain ty) is the assumption that, provided a magical ritual is properly performed, anxiety will remain latent. It follows from this that, should a suitable ritual not be properly performed, anxiety will necessarily be felt. Thus, with the Andamanese example, The anxiety has, so to speak, been displaced from the original situa tion. But even granted that it has been displaced, Malinowski's general theory is confirmed by the existence of a secondary ritual which has the function of dispelling the secondary anxiety which arises from a breach of ritual and tradition. We call this the ritual of purification, of expiation.46 This leads Homans to an elaboration of the anxiety theory of ritual, based on some conceptual distinctions not made, or not explicitly made, by Malinowski. Thus Homans distinguishes between: i. Primary anxiety, which arises when an individual is seek ing to realize a given end and does not possess a technique certain of guaranteeing him or her the desired result; ii. Primary ritual, which is performed in the above circum--161-stances in accordance with definite social norms. Where the ritual tradition is weak, as in our own society, individuals will invent rituals in situations of this type; iii• Secondary (displaced) anxiety, which is experienced when primary rituals are not properly performed. This atti tude becomes generalized and is experienced when any of the traditions of society is not observed; iv. Secondary rituals, which are the rituals of purification and expiation which have the function of dispelling secon dary anxiety* These will be invented where not already 47 existing as a social tradition. Por Homans, therefore, as for Malinowski, magic and other ritual forms are primarily responses to situations of anxiety. These responses secure no direct benefits, but are indirectly beneficial both to the individual and to society, since they instill confidence in the individual, dispelling anxieties and disciplining the social order. Yet despite his sophistication of Malinowski*s theory, Homans, by his own admission, does not provide us with a theory of ritual, but with a theory applicable to certain ritual occasions only. His argument, in other words, concerns rituals only to the extent to which they do arise from anxiety, and "... there is no implication that other sentiments besides 48 anxiety do not give rise to ritual.n But which sentiments and on what occasions? Can different sentiments give rise to the same ritual on different occasions? Can the same sentiments give rise to different rituals, in the same cir cumstances, on the part of different actors? Is ritual be haviour subject to a unifying interpretation on the level of 162-sentiment at all? Questions such as these are given no con sideration. How, moreover, are we to recognize "ritual" be haviour? According to Homans, any action which does not produce a direct practical result on the external world is to be classified as ritual, whatever the actor enacting it 49 might happen to think. Native systems of classification are therefore to be ignored, and the actor's own explanation of his or her behaviour is to be treated as merely a kind of 50 rationalization which is not to be taken at its face value. While Homans is at least consistent, his approach allows us no way of differentiating genuinely ritual from mistaken em pirical notions. An ineffective mode of medical treatment, for example, must logically be equated with such other very different phenomena as Hopi rain dances, the Catholic Mass and the American Pledge of Allegiance. Homans' consistency, then, is achieved only at the cost of violating the integ rity of other systems of thought, and he is therefore open to much the same kind of criticism as that which we have al-51 ready made of Pareto. In comparison to Malinowski, there fore, who always emphasized the importance of native cate gories of thought, Homans' work represents a retrograde step. It is also worth-while noting another criticism which may be levelled at the affective theory of magic: that it is unable to explain cultural variations in magical beliefs and practices, and therefore cannot account for those situations where human knowledge cannot predict the issue yet where ma gic is absent. For instance, the Tikopians, unlike the Trobrianders, have no love magic, while the Manus Islanders, -163-again unlike the Trobrianders, have no sailing magic. One way of meeting this objection is by invoking the notion of "functional substitutes". One might, in other words, argue that the functions fulfilled in one society by magic are fulfilled by some other cultural element(s) in another so ciety. This is, in effect, the line of reasoning adopted by Firth when he writes that magic represents only one of a number of possible cultural responses to uncertainty, which may also take the form of "... a reliance on a beneficent Sod, a reliance upon the theory of probability - which is another name for science, or a simple fatalism which rejects 52 both science and God." So far as Malinowski himself is concerned, however, this line of reasoning would seem ruled out by his own assertion, quoted earlier in this chapter, that magic fulfills an indispensable function which cannot 53 be fulfilled by any other element of primitive culture. Moreover, Firth's mode of argument merely shifts the problem from "why do the Manus Islanders, unlike the Trobrianders, not practise magic in the face of certain forms of uncer tainty?" to "why do the Manus Islanders adopt one rather than another cultural response to certain forms of uncer tainty?". This question, Firth himself confesses he is un-54 able to answer. Invoking the notion of functional substitutes, there fore, succeeds in salvaging the affective theory of magic only by rendering it completely untestable in terms of em pirical evidence. Recognizing the limitations of Malinowski's theory as expressed in its original form, and attempting to safeguard it from criticism by adopting the -164— modifications to it suggested by Homans and Firth, can in fact lead only to very vague generalizations along the lines of: "anxiety, and/or other sentiments, may give rise to magical or other ritual procedures, or, in certain un  specified circumstances, may give rise to a different cul tural response." Expressed in this way, it becomes clear that the anxiety theory of magic, as further developed by Firth and Homans, may be reconciled with any conceivable item of empirical information. In other words, the theory can explain everything - and nothing. Yet, even although his general theory of magic ul timately leads one into a blind alley, Malinowski may never theless be credited with having provided an important im pulse to the study of witchcraft and sorcery. Both Navaho  Witchcraft and Witchcraft, Oracles and Magic among the  Azande owe an important debt to Malinowski. For example, Kluckhohn*s theory that Navaho witchcraft beliefs provide an outlet for the release of tensions arising from the general and specific conditions of Navaho life, holds many points in common with Malinowski's anxiety theory of magic. Similarly, even although Evans-Pritchard*s analysis of Zande witchcraft assumes methodological and theoretical principles quite different from those of Malinowski, it was only by means of employing the techniques of meticulous field-work, which Malinowski was first and foremost in de veloping, that the Zande study was made possible. Moreover, for his emphasis on the necessity of understanding native categories of thought, and on the importance of seeing how particular patterns of belief are translated into action in -165-specific behavioural contexts, the study of witchcraft and sorcery - indeed, the entire field of anthropology - will always remain indebted to the genius of Malinowski. -166-Notes and References 1. Evans-Pritchard, E.E., Social Anthropology, (Cohen & West ltd., London, 1951), Pp. 91-2. As a statement of anthropological attitudes, this probably holds true even today, although writings like those of Levi-Strauss, Leach, Horton and Douglas indicate a renewed interest in some of the "bigger" questions and a return to broader theoretical concerns. 2. Pocock, D.P., Social Anthropology, (Sheed & Ward, London, 1961), P. 53. To some extent, this distinction between Malinowski's "theoretical" and his "ethno graphic" writings is misleading. Por Malinowski^ theory was deeply grounded in his Trobriand material (of which it was often little more than a generaliza tion) while, conversely, his ethnography was deeply influenced by his theoretical preconceptions. 3. Nadel, S.P., "Malinowski on Magic, Science and Religion", in Firth, R. (ed.), Man and Culture, (Routledge & Kegan Paul, London, 1957), P. 190. 4. Redfield, R., "Introduction" to Malinowski, B., Magic.  Science and Religion and Other Essays, (Doubleday & Co. Inc., Garden City, New York, 1954), P. 10. 5. Malinowski, B., "Introduction" to Hogbin, H.I., Law and  Order in Polynesia, (London, 1934), P. xxxviii. Quoted in Homans, G.C., "Anxiety and-Ritual: The Theories of Malinowski and Radcliffe-Brown", American Anthropologist, Vol. XLIII, (April-June 1941), P. 169. Bobbs-Merrill reprint, S-121. 6. See P. 94 of this thesis. 7. Malinowski, B., "Magic, Science and Religion", in Malinowski, B., Magic..., op. cit., P. 70. 8. Ibid., P. 71. 9. Ibid., P. 74. -167-10. Ibid.. Pp. 73-4. 11. Ibid., P. 74. 12. Ibid.. P. 73. 13. Lowie, R.H., Primitive Religion, (Boni & Liveright, New York, 1924), Pp. 143-4. See P. 100 of this thesis. 14. Evans-Pritchard, E.E., "The Morphology and Function of Magic", American Anthropologist. Vol. XXXI, (1929), Pp. 619-41. Reprinted in Middleton, J. (ed.), Magic.  Witchcraft and Curing, (The Natural History Press, Garden City, New York, 1967), Pp. 1*22. Horton, however, argues that all African magic may be reduced to a ver bal base, and suggests that the non-verbal symbols com monly employed in African magic derive their powers from verbal designation. If correct, this suggests that Malinowski was right to emphasize the cross-cultural significance of the spell. Horton, R., "African Traditional Thought and Western Science", in Marwick, M. (ed.), Witchcraft and Sorcery, (Penguin Books Ltd., Harmondsworth, 1970), P. 353. 15. See Malinowski, B., Argonauts of the Western Pacific, (E.P. Dutton & Co. Inc., New York, 1961), Pp. 424-5. 16. Malinowski, B., "Myth in Primitive Psychology", in Malinowski, B., Magic..., op. cit., P. 141. 17. Nadel, S.F., "Malinowski on Magic...", op. cit., P. 141. 18. Malinowski, B., "Magic...", op. cit., P. 17. 19. Ibid., Pp. 28-9. 20. Leach, E.R., "The Epistemological Background to Malinowski's Empiricism", in Firth, R. (ed.), Man and  Culture. op. cit., P. 128. 21. Malinowski, B., Coral Gardens and their Magic, (Allen & Unwin, London, 1966), Vol. I, P. 460. Quoted in Leach, E.R., "The Epistemological...", op. cit.. P. 128. -168-22. Malinowski, B., "Magic...", op. cit., P. 79. 23. Ibid., P. 80. 24. Ibid., Pp. 80-1. 25. Ibid., P. 81. 26. Ibid.. Pp. 74-6; "Myth...", op. cit., Pp. 138-43. 27. Malinowski, B., "Magic...", op. cit., P. 82. 28. Ibid., Pp. 81-2. 29. Ibid., P. 90. 30. Ibid.. 31. Malinowski, B., "Anthropology", Encyclopaedia - Britannica, (London, 1926), 1st. Supplementary Volume, P. 132. Quoted in Merton, R*, Social Theory and Social  Structure, (The Free Press, Glencoe, 1949), 36. 32. It is worth noting that Homans' elaboration of the anxiety theory of magic, discussed on Pp. 160-2 of this thesis, also fails to deal with the phenomena of witch craft and sorcery. 33• Malinowski, B., "Magic...", op. cit., P. 85. 34. Ibid., Pp. 85-6. 35. Nadel, S.F., "Malinowski on Magic...", op. cit.. P. 194. 36. Malinowski, B., Argonauts..., op. cit., P. 239. 37. Leach, E.R., "The Epistemological Background...", op.  cit., P. 129. 38. Malinowski, B., Crime and Custom in Savage Society, (Routledge & Kegan Paul, London, 1926), P. 86. 39* Ibid., P. 93. 169-40. Ibid.. P. 94. 41. Malinowski, B., "Magic...", op. cit., Pp. 30-1. 42. Nadel, S.P., "Malinowski on Magic...", op. cit., P. 193. 43. Ibid.. 44. According to Malinowski, magic employs limited tech niques to achieve limited goals, whereas religion con sists of a body of self-contained acts, complex and rich in content, and is expressed in acts which have value but not utility, the end of which being realized in the acts themselves. Cf. Malinowski, B., "Magic...", op. cit.. Pp. 88-9. 45. Radcliffe-Brown, A.R., Structure and Function in  Primitive Society, (Cohen & West Ltd., London, 1952), Pp. 148-9. 46. Homans, G.C., "Anxiety and Ritual...", op. cit., P. 170. 47. Ibid., P. 171. 48. Ibid., P. 167. 49. Ibid., P. 172. 50. Ibid., Pp. 165-6, 171. 51. See Pp. 14-17, 50? of this thesis, 52. Firth, R., Human Types, (Mentor, New York, 1958), P. 130. 53. Ibid.. 54. Ibid.. CHAPTER SEVEN EVANS-PRITCHARD'S ANALYSIS OF ZANDE WITCHCRAFT A. Introduction Max Glucl—an reveals that it was early noticed, by Europeans in contact with Africans, that the belief in witchcraft involved the idea that Africans thought it "... singular that they alone should be sick while all the 1 people around them were enjoying good health." This ob servation contained an important clue to the understanding of witchcraft beliefs. Another was contained in the obser vation that people accused their personal enemies of be witching them: an observation which led many to conclude that witchcraft observations were obviously fraudulent. ... these separate clues were brought together by Professor Evans-Pritchard to explain the logic, the intellec tual coherence, of witchcraft beliefs in their relation to natural events and to society. He did this in his analysis of Witchcraft, Oracles and  Magic among the Azande of the Anglo- Egyptian Sudan. Subsequent research in other African tribes has confirmed this analysis entirely.2 Evans-Pritchard's study may be described as the first systematic analysis of the witchcraft beliefs of a primi tive people as constituting a coherent system of collective representations, as well as the first clear analysis of how these collective representations are invoked in particular 171-action-contexts. As such, it has had a tremendous impact on British social anthropology and has served to define most of the problems to which anthropologists have directed their attention, in studying witchcraft and sorcery, in the period following the Second World War. But although deepest in the field of anthropology, the influence of Evans-Pritchard1 s work has by no means been limited to the circle of his professional colleagues. Thus, his analysis has provided the starting point for an intensive village level study of witchcraft in Tudor and Stuart Essex by a professional historian,3 and has even at-4 tracted attention in the field of the philosophy of science. It is necessary to view Evans-Pritchard's work in relation to that of Levy-Bruhl. Already, in a series of papers published at the University of Egypt, Evans-Pritchard had shown himself unsympathetic to the psycho-logistic reductionism then prevalent in British anthropo logy. He revealed his methodological standpoint as being much closer to that of L^vy-Bruhl, Durkheim and the con-5 tributors to the Ane'e sociologique. He was particularly concerned to defend Levy-Bruhl's writings, which he con sidered extremely valuable theoretically, against the mis understandings and misrepresentations of them, widespread then as now.** It was L^vy-Bruhl who first stated with emphasis that primitive peoples hold distinctive systems of beliefs in herent in their conceptual frameworks and reflected in their languages. Because of this, the mental structures of other peoples cannot be understood in the same terms as our 172-own, and the methodological basis of Tylorian anthropology-was destroyed. Similarly, it was LSvy-Bruhl who insisted strongly on seeing particular modes of thought as deriving from particular types of social organization. Thus, some of the most basic questions which Evans-Pritchard sought to deal with in his study - how the Azande see the world, how their beliefs form a coherent system, how these beliefs are manifested behaviourally, what are the contexts of their mystic and non-mystic thought - could well have been posed 7 by LSvy-Bruhl himself. B. Evans-Pritchard1s Viewpoint In contrast to writers like Prazer, Malinowski and Kluckhohn, Evans-Pritchard was more interested in providing us with an understanding of how the Azande themselves think about witchcraft and magic, and of how these ideas enter into their way of life, than in attempting a causal or quasi-causal explanation of these phenomena. Hence, he took the thought patterns of the Azande as given (much as a linguist would take a language as given), and sought to re veal their structure rather than determine their historical or psychological origins. Indeed, this analogy between Evans-Pritchard's endeavour and that of a linguist is par ticularly apt, since he refers in more than one place to Azande beliefs as constituting an "idiom" and of his main aim as being that of rendering this idiom intelligible to us. He therefore sought to resolve problems which he des cribed (in a manner already suggesting where his differ ences with I»e*vy-Bruhl would be found) as being of the fol lowing order: -173-Is Zande thought so different from ours that we can only describe their speech and actions without compre hending them, or is it essentially like our own though expressed in an idiom to which we are unaccustomed? What are the motives of Zande be haviour? What are their notions of reality? How are these motives and notions expressed in custom?" He therefore eschewed "... current psychological and sociological explanations of mystical notions and ritual behaviour", and instead of seeking to answer the question of why it is that the Azande attribute particular events to the operation of magic and witchcraft, sought to elucidate 9 how they do this. In other words, he was searching for information in the reasons Azande themselves give for the occurrence of phenomena. Thus his method involved trying to "... explain a fact by citing other facts from the same culture and by noting interdependencies between facts."10 C. Witchcraft Explains Misfortune We have already, in an earlier section of this thesis, noted the essential facts of Zande witchcraft (mangu) as an organic, inherited substance, and of Zande sorcery as a technique employing bad medicines.11 The first major ques tion to which Evans-Pritchard addressed himself in relation to these beliefs was that of determining the contexts in which they are invoked. The short answer to this question is that they are invoked in situations of misfortune, how ever trivial and petty, as a "... ready and stereotyped 12 means of reactkng to such events". ... there is no niche or corner of -174-Zande culture into which it (i.e. witchcraft) does not twist itself. If blight seizes the ground-nut crop it is witchcraft; if termites do not rise when their swarming is due and a cold useless night is spent waiting for their flight it is witchcraft; if a wife is sulky and unresponsive to her husband it is witchcraft; if a prince is cold and distant with his subject it is witchcraft; if a magi cal rite fails to achieve its purpose it is witchcraft; if, in fact, any failure or misfortune falls upon any one at any time and in relation to any of the manifold activities of his life it may be due to witchcraft. This is not to suggest that the Azande believe witch craft to be the sole cause of phenomena, or even the sole cause of misfortune. They are well aware, for example, of the physical nature of elephants as huge beasts capable of throwing men into the air, piercing them with their tusks, and crushing them under their knees. They are equally well aware that granaries collapse because termites eat away the supports which are holding them up, and that if people are sitting under a granary at the moment when it collapses they will be injured by its heavy structure. The Azande, in other words, are quite as capable of observing a se quence of events within a chain of natural causation as we are. On the other hand, the Azande do not limit themselves to asking how it is that such events take place, but also seek to account for why they take place. For example, it may very well be true that a particular granary collapsed because its supports had been eaten away by termites. It -175-may also be true that those who were sitting under it at the moment of its collapse were injured because of its heavy structure. But why, ask the Azande, should those particular people have been sitting under that particular granary right at the moment of its collapse? Similarly, it is certainly true that elephants are physically capable of attacking and injuring people. But this is not a common event. Why should one man, on one particular occasion, happen to come face-to-face with an elephant which attacks him? Why this man and not some other man? Why on this occasion and not on other occasions? Why this elephant and not some other elephant? We would dismiss these questions as illegitimate, at least in our more scientific moods. Such events are to be explained as coincidences, as contingent upon the chance intersection of two chains of causality between which there is no intrinsic relationship. The Azande, in contrast, seek to go a step further than this, in providing an expla natory framework which accounts for just such random events on other than a chance basis. This explanatory framework is provided by the idiom of witch beliefs. A boy knocked his foot against a small stump of wood in the centre of a bush path, a frequent happening in Africa, and suffered pain and incon venience in consequence. Owing to its position on his toe it was impossible to keep the cut free from dirt and it began to fester. He declared that witchcraft had made him knock his foot against the stump.... I told the boy that he had knocked his foot against the stump of wood because he -176-had been careless, and that witch craft had not placed it in the path, for it had grown there naturally. He agreed that witchcraft had nothing to do with the stump of wood being in his path but added that he had kept his eyes open for stumps, as indeed every Zande does most carefully, and that if he had not been bewitched he would have seen the stump. As a con clusive argument for his view, he re marked that all cuts do not take days to heal but, on the contrary, close quickly, for that is the nature of cuts. Why then, had his sore fes tered and remained open if there were no witchcraft behind it?1* Witchcraft is therefore an explanatory mode which ac counts for the particular and variable conditions of an event, as distinct from the universal and general condi tions. As such, it in no way contradicts empirical obser vation, and may even encompass scientific theories within its ambit. Thus a Pondo teacher in South Africa remarked to Monica Wilson, arguing in a manner identical to that of the Azande, that: It may be quite true that typhus is carried by lice, but who sent the in fected louse? Why did it bite one man and not another?15 As LSvi-Strauss remarks of this type of explanation, Seen in this way, the first differ ence between magic and science is therefore that magic postulates a complete and all-embracing determin ism. Science, on the other hand, is based on a distinction between levels: only some of these admit forms of de terminism; on others the same forms -177-16 of determinism are held not to apply. The Azande themselves explain their notions of causa tion by means of an analogy to "the second spear". Thus, if a man is killed by an elephant, the elephant is the first spear and witchcraft is the second spear, and to gether they killed the man. Similarly, if a man kills an other in war, then he is the first spear and witchcraft is 17 the second spear, and together they kill the man. It also should not be assumed that witchcraft consti tutes the only agency of what Europeans might term "mystic" danger. A sorcerer may make magic with bad medicines and kill a man. A breach of taboos, like those forbidding in cest, may lead to serious illness. The death of babies from certain diseases is ascribed to the action of the Supreme Being. Thus, if a man develops leprosy and there is a history of incest in his case, then the cause of his illness is incest, not witchcraft. Again, if a potter has sexual intercourse the night before firing his pots, and the pots are broken in the firing process, then the cause of the breakage is not witchcraft but failure to observe the necessary taboos preparatory to firing. In the same way, if the parents of a child have sexual relations before it is weaned and the child falls ill, the question of witch craft does not arise. Por the child's illness is clearly to be attributed to the parents' breach of a ritual pro hibition. In some cases, however, a complex situation arises where witchcraft operates together with other mystic forces in such a way as to lead to the death of someone. Thus, in 178-the cases mentioned above, if the leper or the child were to die, there would really be three causes operating. Firstly, there is the illness itself - leprosy, or perhaps a fever in the case of the child. These diseases are not seen as being a product of witchcraft but exist in their own right, much as a buffalo or an elephant exists in its own right. Secondly, there is the breach of the taboo: that on incest in one case, that on sexual intercourse be fore weaning in the other. The child developed fever, and the man leprosy, because the taboo was broken. But death need not have ensued. Not all persons die from leprosy, not all children from fevers. Certainly, fever and leprosy were developed because of the violation of certain taboos. But unless witchcraft had been present as the "second 18 spear", they would not have led to death. D. Action Against Witchcraft Evans-Pritchard drew an analogy between the Zande con cept of witchcraft and our concept of bad luck. When, in spite of forethought and technical competence, a person suffers a misfortune, we say that he or she is having bad luck. The Azande say that the person has been bewitched. And, should the situation which is thus described have al ready taken place, we content ourselves with the thought that the failure was "just bad luck", just as the Azande content themselves with the thought that the failure was a result of witchcraft. The difference in reactions between Europeans and Azande arises not when an unfortunate event has already taken place (with the exception of death), but when a misfortune is actually in the process of falling -179-upon one, or is anticipated. We, like the Azande, make an effort to elude or terminate the misfortune, according to our ideas of the objective conditions causing it. But since the Azande conceive of these objective conditions very differently from the manner in which we do, they con centrate their attention on the factor which they see as being of greatest importance: i.e., witchcraft. Witches are detected by means of oracles and by witch doctors. The most important oracles are the poison oracle, which is considered the most reliable, the termite oracle and the rubbing-board oracle, which is considered the least reliable. The reliability of witch-doctors is thought to be about the same as that of the rubbing-board oracle. Witch-doctors hold public seances at which they divine the names of witches who happen to be inflicting, or threatening to inflict, misfortunes on a client. At these seances, the witch-doctor dances with a group of his col leagues to the sound of drums, hand-bells and other instru ments. In the course of his dance, he cross-examines his client, perhaps demanding the names of his wives, neigh bours, or those who have taken part in some activity with him. He then dances, in a trance-like state, with the names of three or four possible witches on his mind and after a period of some hours (a seance may last from mid day to sunset) advises his client of whom he thinks is res ponsible. This he usually does discreetly, perhaps by hints and inuendo, perhaps by whispering the name of the witch to his client in private after the spectators have dispersed. Generally speaking, witches are not denounced -180-publicly by witch-doctors, and no man would proceed to direct action against a witch on the basis of a witch doctor's information alone. Por the witch-doctor is a hu man being, and is therefore fallible. His revelations must 19 be checked against the verdict of the poison oracle. The least reliable of Zande oracles is the rubbing-board oracle. It consists of two pieces: a small table like surface which is supported by two legs and a tail, and an adjoining piece of wood which fits the "table" like a lid. The operator squeezes various juices, or grates wood, onto the surface of the table, and then dips the lid into a gourd of water and proceeds to slide it back and forth across the surface of the table. Usually the lid moves smoothly back and forth, but occasionally it sticks to the lower surface so firmly that no amount of jerking will move it. It is in this way that the oracle answers questions: by smooth sliding, or by sticking. The operator therefore addresses questions to the oracle according to the formula, "if such is the case, rubbing-board oracle, stick; if such is not the case, rubbing-board oracle, run smoothly". Por example, "If so-and-so is bewitching my home, rubbing-board stick." Afterwards you ask it again, "Rubbing-board, it is hot witchcraft, I place witchcraft on one side, it is wingi; another thing, sorcery, is about to spoil my home, rubbing-board stick."20 The Azande do not have complete faith in the operation of the rubbing-board oracle but consider that it may err, and that it may be manipulated by men. Its judgements must -181-therefore be checked against the poison oracle, unless serious illness is involved and speed is all-important. Otherwise, the oracle is consulted on issues of lesser im portance, or in order to clarify a case so that it may be put to the poison oracle for final judgement. If a man is ill, for example, a great many people may occur to him as being possibly responsible. It would be tedious and expen sive to place six or seven names before the poison oracle when, perhaps, the last name is the correct one. But the rubbing-board can sort out the relevant name in a matter of minutes and the poison oracle then may be consulted for confirmation. This efficiency makes the rubbing-board the most often used of all Zande oracles. The termite oracle is more highly esteemed than the rubbing-board oracle, and no one would place a verdict of the termite oracle before the rubbing-board for confirma tion. There are no expenses involved (this is its chief advantage), and a man need only find a termite mound and insert two branches, from different trees, in the runs and return next day to see which has been eaten. The oracle is addressed in such words as, n0 termites, I will die this year, eat dakpa. I will not die, eat kpoyo." Thus if dakpa is eaten and kpoyo is left untouched, it is a pro phecy of misfortune; if kpoyo is eaten and dakpa is left untouched, of good fortune. If both the branches are eaten, but one more than the other, it is regarded as a qualified answer, tending towards either a positive or a negative verdict, according to the situation. If both branches are eaten about equally, the Azande may simply say 182-that the ants were hungry and ate to satisfie their appe tites, that a taboo has been broken, or that witchcraft has interfered with the oracle. If neither branch is eaten, this means that the termites have refused an answer, and the Azande try another mound. All important verdicts are 22 submitted to the poison oracle for confirmation. The poison oracle is the most important of all Zande oracles. The Azande rely completely on its decisions, and these have the force of law when obtained on the orders of a prince. Consultation of the oracle is regarded as indis pensable in all matters strongly affecting individual wel fare, in serious legal disputes, in important collective undertakings, or on any occasion regarded as dangerous or socially important: ... how can a Zande do without his poison oracle? His life would be of little worth. Witches would make his wife and children sick and destroy his crops and render his hunting use less. Every endeavour would be frus trated, every labour and pain would be to no purpose. At any moment a witch might kill him and he could do nothing to protect himself and his family. Men would violate his wife and steal his goods, and how would he be able to identify and avenge him self on adulterer and thief? Without the aid of his poison oracle he knows that he is helpless and at the mercy of every evil person. It is his guide and his counsellor ... The oracle tells a Zande what to do at every crisis of life. It reveals his enemies, tells him where he may seek safety from danger, shows him hidden -183-mystic forces, and discloses past and future. Truly a Zande cannot live without his benge. To deprive him of that would be to deprive him of life itself.23 The oracle is administered by an operator who must be in a ritually pure condition. That is to say, he must have abstained from having had sexual relations with women, from the eating of elephant flesh, fish and certain vegetables, and from the smoking of hemp. These prohibitions cover from five to six days prior to consultation of the oracle in the case of sexual relations, and three to four days in the case of the forbidden foods. The poison oracle is consulted in secret, and no one except trusted friends are advised that there is to be a seance. A large audience is not wanted, since its members would want to ask the oracle all about their own affairs. The oracle takes the form of administering benge (apparently a poison related to the strychnine family) to fowls. As the poison is administered, the oracle is ad dressed according to the formula, "If such is the case, poison oracle kill the fowl", or, "If such is the case, poison oracle spare the fowl". The fowl is then lifted in the hands of the operator, jerked back and forth in order to stir up the poison, and replaced on the ground. The effect of benge on fowls varies. Occasionally; though not often, it kills them at once, even before they are picked up from the ground by the operator. More com monly, a fowl is not seriously affected until it is picked up from the ground when, if it is going to die, it under--184-n goes a series of spasms culminating in vomiting and death. Some fowls appear little affected by the poison until re turned to the ground, when they suddenly expire. Others are quite unaffected by the poison and when, after having been jerked backwards and forwards for a while, they are thrown to the ground, peck around unconcernedly. Very rarely, the poison seems to kill a fowl which later re covers . The oracle normally involves two tests. If a fowl dies in the first test then another fowl must survive in the second test, and vice versa, for a judgement to be accepted as valid. Usually, although not invariably, the questions are so framed that the oracle must kill a fowl in the first test and spare another fowl in the corroborating test to give an affirmative reply, and must spare a fowl in the first test and kill another fowl in the corroborative test to give a negative reply. If two fowls live, or two fowls die, then the verdict is regarded as invalid and the oracle must be consulted on another occasion.24 The Azande are selective in the doubts and questions they put before the oracles, for there is always witchcraft about and it is impossible to eliminate from one's life. It is therefore only about important matters, usually con nected with the state of their health, that they consult the oracles. When a man falls ill, he usually retires to a grass hut in the bush where he can remain hidden from witchcraft and organize counter-measures. He asks a close kinsman, a son-in-law, or some other person on whom he can rely, to -185-consult the oracle on his behalf in order to identify the witch who is injuring him. The first oracle consulted is usually the rubbing-board oracle, which will select the names of several witches who may be responsible from a large number. Then, if the man is poor, he will place the names selected before the termite oracle for confirmation; but if he is able to obtain benge and chickens, he will consult the poison oracle. Following identification of the witch by the poison oracle, two possible lines of action are open to the sick man and his kin. One possibility is to make public ora tion, in which the kinsmen declare that they know the name of the witch who is injuring their relative but that they do not wish to expose it and thus shame him, and that they expect him to return their courtesy by leaving their kins man in peace. This procedure is especially suitable if the witch is a person of some importance whom they do not wish to affront, or someone who enjoys the esteem and respect of his fellows and they do not wish to humiliate him. Their object is to humour him, for an open quarrel would only ir ritate him, perhaps leading him to kill the victim out right. This procedure is followed, however, only when authorized by the rubbing-board oracle. The more usual procedure is to cut off a wing from the fowl which has died to the name of the witch, thrust it on the end of a small, stick and forward it to the witch, usually through a messenger of one of the deputies of the prince. On his arrival, the messenger lays the wing on the ground before the witch and says that he has been sent on -186-account of the sickness of so-and-so. The witch, almost invariably, replies courteously that he is unconscious of hurting anyone, but that if such is the case he is sorry and that if it is he alone who is responsible the victim will soon recover, since he wishes him health and happi ness. To indicate his good intentions, he calls for a gourdful of water and when his wife brings it he takes a mouthful and blows it out in a thin spray over the wing lying before him on the ground. He then says, so that the messenger may hear and repeat his words, that if he is a witch, then he is unaware of his state; that he is not in tentionally causing witchcraft, and that he addresses the witchcraft in his belly, beseaching it to become "cool". The messenger then returns to the deputy, to report what he has done and seen, and the deputy then informs the kin of the sick man that the task has been fulfilled. Should, however, the illness continue, the oracles will be con sulted afresh, to discover if the repentance exhibited by the witch was genuine or fraudulent, or whether some new witch has started to trouble the victim in the meantime. In either case, the formal presentation of chickens* wings is carried out as before, through the intermediacy of a 25 prince*s deputy. Almost all deaths are attributed to witchcraft, and must be avenged. In pre-European times, vengeance was exacted in one of three ways: by the slaughter of the witch, by the payment of compensation, or by means of lethal vengeance magic.5 Normally, however, witches were seldom slain, unless they committed a second or third mur--187-der or killed an important person. Today, vengeance magic is the sole means employed, and the success of its opera tion is confirmed by the poison oracle of the prince. The use of this vengeance magic is more a requirement of pious duty than a result of anger and hatred, and the kin of the dead man are not permitted to cease mourning until the guilty witch has been struck down. Who he is, will be known only to the prince and to the kin of his victim. One can deduce the fact that their vengeance magic has struck only because the kin are no longer observing the taboos of mourning. It is useless to ask them who was its victim, for it is their private affair and a secret between them selves and their prince. Besides, as a verdict of the 26 poison oracle, such information may not be disclosed. It should be noted that magic is utilized against witches not only as vengeance magic, but also for protec tive purposes. Thus, magicians are employed to bury magic on the thresholds of homes in order to protect them against witchcraft and sorcery by destroying witches and sorcerers who intend the occupants ill. Thus one spell Evans-Pritchard records as having been uttered over medicines buried in the entrance to a homestead, runs: That man who may come to bewitch me, and bewitch my hunting, and to be witch my wives and children, may he die.27 Similarly, a spell to protect a man's gardening: The medicine which I place in my cul tivations - whatever witch comes to injure my cultivations, to harm my food-plants so that they will not -188-28 flourish, may he die. E. The Direction of Witchcraft Accusations We have already noted that European explorers, mis sionaries and administrators in Africa often observed that accusations of witchcraft tended to be made against per sonal enemies: hence the frequent conclusion that these 29 charges were fraudulent. Evans-Pritchard*s Zande data, however, shows that this is a mistaken conclusion to draw. Por although it is personal enemies whom the Azande most often accuse of bewitching them, such accusations by no means represent an underhand way of striking a blow against someone against whom a grudge is held. Bather, they are a logical outcome of Zande doctrines concerning the nature of witchcraft. The Azande are not interested in the possession of witchcraft substance in itself, but only in witch activity proper. Any person may be born a witch, but it does not follow that they will necessarily cause ill to their fel lows, for their witchcraft substance may remain "cool". Witchcraft substance is not in itself dangerous, but only when it is activated by the ill will, jealousy or spite of its possessor. A witch attacks a man when motivated by hatred, envy or greed. It follows from this that a Zande who is suffering from or is threatened by some misfortune, consults the oracles primarily to the names of those whom he thinks are likely to bear him such ill will. He is well aware that others take pleasure in his troubles and pain and are displeased at his good fortune. He knows that if he becomes rich the -189-poor will hate him, that if he rises in social position his inferiors will be jealous of his authority, that if he is handsome the less favoured will envy his looks, that if he is talent ed as a hunter, a singer, a fighter, or a rhetoritician, he will earn the malice of those less gifted, and that if he enjoys the regard of the prince and of his neighbours he will be de- ,Q tested for his prestige and popularity. There are plenty of occasions for hostility and ill-feeling in the context of Zande life, and this is often ex pressed in gossip in the privacy of the huts or in whisper ings to close friends in the safety of the bush. Sometimes such bitterness results in a complaint being made to the prince's court. More often, a man who knows that others are jealous of him will do nothing. He continues to be polite, and tries to remain on friendly terms. But should some misfortune strike him, he will at once conclude that one of these men is bewitching him, and will place their names before the oracles to determine who among them is responsible. Oracle consultations therefore ex press histories of personal relation ships, for, as a rule, a man only places before an oracle names of those who might have injured him on account of some definite events which he believes to have occasioned their enmity. It is often possible by adroit questioning to trace back the placing of a name before the oracle to its source in some past incident. Since accusations in this way reflect personal en mities, it can be seen that certain people will not be -1 go-considered when the sick man searches his mind for the names of those who might he injuring him. Thus, commoners do not accuse nobles, and only rarely do unimportant com moners accuse influential commoners. This is partly be cause it would be inadvisable for those of low social standing to insult those of higher social status by thus accusing them, but mainly it is because social contact between those of high and low social status is limited to situations where behaviour is fixed by notions of rank. A man quarrels with and is jealous of his social equals. A noble is socially so separated from commoners that were a commoner to quarrel with him it would be treason. Commoners bear ill-will against commoners and princes hate princes. Likewise a wealthy commoner will be patron to a poorer commoner and there will sel dom be malice between them because the incentive to malice and the op portunity for creating it do not easily arise. A rich commoner will envy another rich commoner and a poor man will be jealous of another poor man. Offence is more easily taken at the words or actions of an equal than of a superior or an in ferior .32 Thus women tend to accuse other women of witchcraft, since there is no social contact between men and unrelated women, while men accuse other men. The main exception to this rule is that a woman may bewitch her husband. No man, however, would bewitch his wife, since by doing so he would only be hurting himself. Similarly, children do not normally have relations with adults of a type which might -191-be expected to generate hatred. In consequence, children are not thought of as bewitching adults. In those cases where an adult bewitches a child, this is usually thought of as being a way of striking against the father. It is among householders of roughly equal status who come into close relations with one another that there is the greatest opportunity of squabbles, and it is these people who most frequently place one an other's names before the oracles when they or members of their fami lies are sick.33 Evans-Pritchard's work therefore contains two impor tant and valuable suggestions, both of which have been re searched and confirmed by later specialists in the study of witchcraft and sorcery: firstly, that accusations of witchcraft and sorcery, as functions of personal relations, are indicative of the main lines of stress and tension in a society; secondly, that accusations tend to be made be tween people who stand in a relationship of competition in 34 regard to scarce resources. It is important to note, however, that Evans-Pritchard did not fall into the trap of postulating a com plete sociological determinism, but emphasized that among the Azande accusations of witchcraft derive primarily from the experience of misfortune and cannot therefore be re garded as a function of personal enmities only. Thus a man who suffers a misfortune knows that he has been bewitched, and only then does he search his mind to find out who is likely to bear him a grudge and therefore have bewitched him. If he is unable to recall any incidents that might 192-have caused a man to hate him, and if he has no particular enemies, he must still consult the oracles in order to discover the witch so that his misfortune will cease. Thus it is that even a prince will sometimes accuse a com moner of witchcraft (although the reverse never happens), since his misfortunes must be accounted for and countered, even when those whom he accuses are not his personal 35 enemies. P. Scepticism and the Variability of Zande Beliefs There is an implication, in anthropological references to "the" Azande, "the" Navaho or "the" Trobrianders, that all Azande, Navahos or Trobrianders share the same ideas, or, putting the matter a little more crudely, think the same. That such an assumption cannot be supported for the Navaho was clearly shown by Kluckhohn, who was careful to indicate divergences of opinion among his informants. Thus it is clear from Kluckhohn*s data that Navahos are by no means agreed on the exact nature of witchcraft, the cha racteristics of different forms of witchcraft, or the dif ferences between witches and ghosts.^ Evans-Pritchard did not attempt a detailed breakdown of variations in Zande belief comparable to that provided by Kluckhohn for the Navaho - perhaps the different con ditions of their field-work are a relevant factor here -but he did provide us with some information regarding such variations, and particularly on the limitations of Zande scepticism. Such information is, of course, essential, for if we are to make generalizations about witchcraft be liefs constituting "closed systems of thought", then we -193-must possess empirical data on just how closed such sys tems are. Similarly, if we are to comprehend social change and the processes whereby people become critical of their beliefs and abandon or transform them, we must know, first of all, how great a degree of scepticism pre-exists, and what are the pre-existing variations in patterns of belief within the culture concerned. Evans-Pritchard discovered a considerable amount of scepticism existing in many sectors of Zande culture. He suggested that differences between individuals in regard to the extent of their scepticism is dependent on four factors: upbringing, range of social contacts, variation of individual experience and personality. The Azande are especially sceptical about witch doctors although, of course, the degree of this scepticism varies markedly from one person to another. Mekana, whose remarks were often tinged with refreshing cynicism, was accustomed to speak lightly of witch doctors. He contrasts in a remark able manner with Kamanga, who was a fervent believer in all kinds of ma gic, and especially in the powers of witch-doctors, a belief which months of mild effort on my part failed to break down.37 Zande scepticism of witch-doctors is manifested not only in personal remarks, but also in the tests put to them and the jokes sometimes played on them by young nobles and commoners of influence. Sometimes a young noble puts a piece of iron, or indeed, any object, in a pot and tells the witch-doctors to -194-divine what is inside it. A com moner friend of mine, Mbira, once placed a knife in a covered pot and summoned witch-doctors to tell him what the pot contained. After three witch-doctors had danced for the better part of the day and had made a number of wildly incorrect guesses about what was in the pot, the fourth still maintained silence and continued to dance. Mbira called to him and told him in the presence of the audience that the sun was sink ing and that people were desirous of returning to their homes and sug gested that he should say what was in the pot as soon as possible, so the seance might break up. A little while afterwards Mbira left the audience and went,.on private busi ness into one of his huts. The witch-doctor followed him and sec retly asked him there to tell him what was in the pot so that he could save his reputation. Mbira, how ever, refused his request, called him a knave, and told him that he would get nothing for his day's ex ertion. 38 It would be a mistake to attribute such scepticism to the result of European influence, for it is sometimes found among the oldest men who are quite free from any such in fluence. Thus one old man remarked to Evans-Pritchard that what the witch-doctors told their audience was merely sup position, and that what they offer in the guise of inspired 39 oracles are only likely guesses. Similarly, it would be equally mistaken to exaggerate the limits of this scepti cism. Although many Azande believe that witch-doctors are 195-often incorrect in their diagnoses of witches, and practise deception and fraud in their capacity as leeches, they do not thereby dismiss the idea of witch-doctorhood as such. Most Azande believe that while the majority of witch doctors are quacks, there are still a few entirely reliable practitioners. This faith in the powers of witch-doctors is especially strong in relation to tales recounted of the powers of witch-doctors among foreign peoples and of past 40 times: those of four or five generations ago, for example. It should also be noted that scepticism about witch-doctors is a function of particular situations: I have noticed that men who fre quently spoke with a measure of con tempt about witch-doctors have made speed to visit them when in pain. Also a man who readily accuses witch doctors of deception when they ex pose him as a witch as readily ap plauds their skill when they expose someone else, particularly if it is enemy.* Zande scepticism is not restricted to witch-doctors, but also extends to other spheres of belief. The Azande are well aware, for example, that the rubbing-board oracle may be improperly manipulated, and it is for this reason that they consider the rubbing-board to be inferior to 42 other oracles. Many Azande are also sceptical about pro tective and punitive magic. Hence, while many men are par ticularly anxious to acquire medicines, and use them fre quently, there are others who: ... do not care whether they possess medicines or not, and who only use them when it is customary and then -196-without enthusiasm. They do not even have deep faith in the majority of medicines. Mekana and Kamanga were excellent examples of these two types. Mekana was not interested in magic, nor did he concern himself with it. Kamanga was really super stitious and believed in every kind of magic ...43 G-. The Persistence and Stability of Zande Belief Evans-Pritchard records that Europeans to whom he ex plained Zande ideas of magic and the power of oracles, re acted with incredulity and even contempt: In their questioning to me they have sought to explain away Zande beha viour by rationalizing it, that is to say, by interpreting it in terms of our own culture. They assume that Azande must understand the qualities of poison as we understand them, or that they attribute a per sonality to the oracle, a mind that judges as men judge, but with higher prescience, or that the oracle is manipulated by the operator whose cunning conserves the faith of lay men. They ask what happens when the result of one test contradicts the other which it ought to confirm if the verdict be valid; what happens when the findings of the oracles are belied by experience; and what hap pens when the oracles give contrary answers to the same question.44 Such questions were not confined to Evans-Pritchard*s audiences, but also formed an important part of his own line of enquiry. Thus, during his discussion of Zande scepticism towards witch-doctors he poses the question as 197-to "... why common sense does not triumph over supersti-45 tion." In this aspect of his work, Evans-Pritchard stands very clearly in an old tradition of anthropological research: that of trying to explain why beliefs in the ef ficacy of magic, although manifestly false, continue to persist. But, in seeking to answer this time-honoured question, Evans-Pritchard raised the level of discussion far beyond that of his predecessors. He thus provided us with an account of the stability of structures of belief which transcends not only the narrow range of the facts of Zande ethnography, but even his own philosophical frame work . Confronted by the facts of Zande belief in witchcraft and magic, the European is likely to adduce a number of ob jections which to him render this belief invalid. Let us briefly outline some of these objections, in order to see why they do not have the same force for the Azande as they might at first sight be expected to: i. the poison oracle contradicts itself by sometimes an swering first yes, and then no, to the same question, and vice versa; ii. an internally consistent judgement given by the oracle may be contradicted by future experience; iii. Zande doctrines concerning the inheritance of witch craft lead to a logical contradiction. If one man is proven to be a witch, then the whole of his clan must also be witches too, since the Zande clan is a group of people related to each other through the male line. On the other hand, if a post-mortem were to disclose that a man was not -198-a witch, it would follow that none of his fellow clansmen were witches either. Thus a few positive results of post mortems scattered among the clans would soon prove that everyone was a witch, while a few negative results would prove the opposite; iv. the existence of vengeance magic reduces the explana tion of death by witchcraft to an absurdity, and vice versa. Thus the death of a man X is avenged upon a witch Y, whose death is avenged in turn upon another witch Z, and so on ad infinitum. Why is it that the Azande do not see the force of these objections to their beliefs? Let us consider each in turn. If the poison oracle contradicts itself in response to a question put to it, the Zande who is consulting it is not bewildered but has at hand a number of ready-made explana tions to account for this contradiction. The wrong variety of poison may have been gathered; a taboo may have been violated by some person handling the poison; the people from whose land the poison was gathered may have been in sufficiently paid and thus angered against the collectors, thus affecting the quality of the poison; the poison may have been kept too long and have lost its strength; ghosts may have attacked the oracle if some of the poison gathered has not been offered by the collector to his father as first-fruits; someone may be practising sorcery against the oracle; or the poison may have become exhausted with use: 46 for any of these reasons the oracle may contradict itself. Zande belief is therefore elaborated in such a way -199-that it is a relatively simple matter to explain why the oracle may sometimes contradict itself. These secondary elaborations, or supportive beliefs, are differently in voked in different contexts. Por example, if at the first seance the poison oracle kills all the fowls without excep tion, the operators generally conclude that this is because the poison was collected from the wrong type of creeper. The poison is therefore stored for some time before it is used again, in order to see if it will become "cool" (i.e., discriminating). After a period of time the poison is then tested again, and either proves that it has become cool, or is thrown away. This explanation is adopted only when fresh poison has been gathered and is being tested in order to determine its worth. If a sample of poison is first passed as being discriminating, only to kill all the fowls at a later seance some other explanation, such as witch-47 craft, must be sought. These secondary elaborations of belief therefore themselves form a system, being interre lated in a coherent pattern where each particular belief acts to support the others. Consequently, it is impossible for Europeans to discredit the power of the oracles by pointing out that they often contradict themselves, or by arguing that benge is a natural poison. Indeed, such argu ments would strike the Azande as being naive and stupid: ... to ask Azande, as I have often asked them, what would happen if they were to administer an extra portion of poison to a fowl which has recovered from the usual doses, or, if they were to place some of the poison in a man's food, is to -200-ask silly questions.... no one has ever been fool enough to waste good oracle poison in making such point less experiments, experiments which only a European could imagine. Proper benge is endowed with potency by man's abstinence and his knowl edge of tradition and will function only in the conditions of a seance. ... I am sure that no Zande would ever be convinced that you could kill a fowl or person with benge un less it had been gathered, adminis tered, and addressed in the tradi tional manner. Were a European to make a test which proved Zande opin ion wrong they would stand amazed at the credulity of the European who attempted such an experiment. If the fowl died, they would simply say that it was not good benge• The very fact of the fowl dying proves to them its badness. Normally there is very little opportunity for the oracle to be proved wrong by subsequent events. Usually put to it provide answers that cannot easily be challenged by subsequent experience, since the enquirer accepts the verdict and does not seek to treat it as a scientific hypo thesis to be tested by experience. Thus no man will build his home in a certain place if he has been warned by the oracle that he will die there. As a result, he never knows what would have happened if he had defied the warning of the oracle. Moreover, the oracle is not asked to answer questions which are formulated in a precise manner. Ques tions are framed in broad terms, and deal with wide issues. Por instance, instead of asking, "Will I get a bushbuck if -201-I go hunting tomorrow?" the consultant demands, "If I make my hunting in such-and-such an area this year, will I have a successful season, or will it he spoilt by witchcraft?" Because of this, many of the questions put before the oracle are difficult to test by experience in any con clusive manner. Moreover, should there be any direct con tradiction between the prophecy of the oracle and empirical experience, this can be explained on the same basis as when the oracle contradicts itself: i.e., as a result of witch craft, of the violation of a taboo, or of the operation of 49 some other mystic agency. In addition to this, the oracle is usually consulted on such matters as witchcraft, sorcery and ghosts, and the information which it yields refers to phenomena of which it itself is the sole evidence. Its results therefore not only do not contradict experience, but deal with phenomena 50 which transcend direct sensory perception. It must also be noted that the authority of the oracle is sustained by the hierarchical structure of Zande society. All verdicts of the poison oracle are backed by the full authority of the king, and the decisions of his oracles are beyond appeal. If an appeal was permitted from the king's oracle to that of a commoner's, or if there was no final deciding oracle, general confusion would ensue. Everybody would be able to claim verdicts in support of his own point of view, and there would be no way of deciding between them. Thus in all legal disputes, the authority of the king safeguards the authority of the poison oracle, and so 51 prevents any serious challenge to its veracity. -202-The Azande see the sense of the argument that if one man is proven a witch, then all of his clansmen must also be witches. They do not, however, accept its conclusion. In practice, they regard only the close paternal kin of a known witch as also being witches, and extend the imputa tion to all of a witch's clansmen only in theory. Further elaborations of their beliefs enable them to escape from what we might regard as a logical deduction from the pre mise that witchcraft is transmitted by biological inheri tance. Thus, even if a man is proven to be a witch, his kinsmen may escape from the logical consequence of this discovery by denying any biological link with him. They may say that the man is not really their kinsman but a bastard, and the brothers of a witch may even extract a confession of adultery from their mother in order to prove their innocence. In addition to this, the Azande do not see the contra diction as we see it, because they have no theoretical in terest in the matter. Because a man has inherited witch craft-substance does not necessarily mean that he will be witch one. His witchcraft-substance may remain cool through out the period of his life, and will therefore be of no in terest to anyone. For when a Zande is ill, he does not normally' think of the names of all the well-known witches in the community and their paternal kinsmen. Rather, he asks himself who among his neighbours is likely to bear him a grudge and then seeks to determine which one of them, on this particular occasion, is hurting him. "Azande are in terested solely in the dynamics of witchcraft in particular -203-52 situations." This structuring of interests means that the Azande do not confront what we perceive as being a con tradiction. For they are not interested in identifying all possible witches, but only in whoever happens to be be witching a particular person at a particular time. Some Azande say that a prince would not agree to con sult the oracles on behalf of the kin of a man who had died of vengeance magic, but would inform them that the man had died in expiation of a crime. Some princes also asserted this, although Evans-Pritchard considered that they were lying and some Azande also expressed their doubts to him about the honesty of the princes in this matter. However, the silence which the Azande keep about the victims of their vengeance magic means that it is impossible to tell one way or the other. For even if a prince informed the relatives of a dea-d man that he had died as a result of vengeance magic, this would be a secret between him and them. And so that their neighbours should not know that their deceased kinsman had been a witch, his relatives would pretend that they were avenging his death. Consequently if the kinsmen of A avenge his death by magic on B and then learn that B's kinsmen have ceased mourning in sign of having accomplished vengeance also, they believe that this second vengeance is a pretence. Contradiction is thereby avoided.53 The fact that information is never pooled concerning the names of the victims of vengeance magic, means that no con tradiction faces the Azande in this sphere of their beliefs. 204-We may therefore list seven main reasons why the Azande continue to accept their beliefs in witchcraft and the efficacy of magic, manifest in the above examples: i. Magic and the oracles are mainly used against other mys tic powers, such as witchcraft and sorcery. Their action therefore transcends, and is not easily contradicted by, experience. ii. Witchcraft, oracles and magic all explain each other. Death is proof of witchcraft. It is avenged by magic. The success of the vengeance magic is proved by the poison oracle. In this way, these beliefs form an intellectually coherent system. iii. Contradictions between beliefs are not noticed because they function in different situations and are not therefore brought into opposition. iv. Individuals and kinship groups do not pool their knowl edge. Thus the same death may represent the closing of vengeance for one family, and the initiation of it for an other. v. The king's oracle, and vengeance magic, are supported by political authority. vi. Not being experimentally inclined, the Azande do not seek to test the efficacy of their oracles and magic. Thus the poison oracle is tested only within the context of the beliefs supporting its efficacy. vii. Zande beliefs are usually vaguely formulated and therefore cannot easily be shown to be contradicted by ex perience or out of harmony with other beliefs. viii. The failure of any rite is accounted for in advance 205. by a variety of mystical notions, such as that a taboo has been violated, or that witchcraft or sorcery is operative. Hence, the perception of error in one situation serves only to prove the correctness of another equally mystical notion. In addition to these reasons, Evans-Pritchard has made a summary of the main reasons why, in his opinion, the 54 Azande "... do not perceive the futility of their magic." These are: i. It is often observed that a medicine is unsuccessful, but such observations are not generalized to include all medicines of the same type, still less all magic. ii. Scepticism is absorbed in the system. It is limited to certain medicines and certain witch-doctors. Belief in others is supported by contrast. iii. The results magic is supposed to achieve are in fact usually realized. Vengeance magic is followed by the death of a man, and animals are killed after hunting magic is made. iv. The Zande is born into a culture with ready-made pat terns of belief which have the force of tradition to sup port them. Many of these beliefs are axiomatic and it 55 never occurs to the Zande to question them. v. Individual experience counts for little against this tradition. If personal experience contradicts a belief, this does refute the belief, but indicates only the pecu liarity of the experience. vi. Magic is used to secure events which are likely to hap pen in any case. Rain magic is made in the rainy season, -206-magic against rain in the dry season. vii. Magic is not usually considered essential for the suc cess of an operation, hut is regarded only as making the success greater. Thus a man will catch many termites with out using termite magic. viii. Magic is not generally expected to produce an effect by itself, but is associated with empirical action that does in fact produce the effect sought. A man makes beer by approved methods of brewing, and uses magic only to hasten the brew, not in place of it. ix. The performance of magic may be required as a part of meeting certain social obligations. Por example, the per formance of vengeance magic by the kin of a deceased man is required as a social duty. x. Successful people, such as successful hunters, acquire a reputation for good magic, even if they possess no medicines. xi. The objective knowledge of the Azande is strictly limited. Thus they do not know that the placing of a stone in a tree cannot retard the sunset. Moreover, having no clocks, they cannot perceive this fact. xii. Myths, folktales and localized stories tell of the achievements of magic. xiii. Most Zande medicines are imported from foreign peoples, whom the Azande think know much more about magic than they do. Foreign medicines are vouched for by the peoples who use them. xiv. The place occupied by the more important medicines in a sequence of events means that they cannot be proved to be inefficacious. Vengeance magic is made against unknown -207-witches. On the death of a man, the poison oracle is con sulted to see if he died as a victim to the magic. If, on the other hand, the oracle were first consulted to determine the witch responsible, and the magic was then made, its fal-56 laciousness; would soon become apparent. H. A Critique of Evans-Pritchard Although more than thirty years have passed since the publication of Witchcraft. Oracles and Magic among the  Azande, no more recent work may yet be said to have sup planted it in importance in its field. Indeed, the vast majority of social anthropological studies of witchcraft in the period following the Second World War have done no more than apply themes and suggestions already contained in the Zande study. This is not to imply that the Zande study stands above criticism. Many of the themes in it, for example, remain implicit, and there are no explicit suggestions for further cross-cultural research. More seriously, the evidence on which Evans-Pritchard based his- generalizations; is not always clear, and a great deal of useful empirical data is; lacking. Thus, little use is made of the extended case study, and no statistical data is; provided on, for example, the frequency of witchcraft accusations between persons standing in particular defined relationships;, or on the re lations between witches, victims and accusers. In this re spect, the Zande analysis compares unfavourably even to Kluckhohn1s study of the Navaho. Perhaps some of these criticisms come too easily, being akin to aecusing a scientific innovator of not developing all -208-the possible applications of his own theory. These are, moreover, criticisms of particulars rather than of funda mentals. A more fundamental criticism of Evans-Pritchard's methodological position is contained in the concluding chap ter to this thesis. -209-Notes and References 1. Gluckman, M., Custom and Conflict in Africa, (Basil Blackwell, Oxford, 1965), P. 81. 2. Ibid., Pp. 81-2. Emphasis in original. 3. Macfarlane, A., Witchcraft in Tudor and Stuart England, (Routledge & Kegan Paul, London, 1970). 4. Polanyi, M., Personal Knowledge, (Routledge & Kegan Paul, London, 1958). 5. See Evans-Pritchard, E.E., "The Intellectualist (English) Interpretation of Magic", Bulletin of the  Faculty of Arts, University of Egypt, Vol. I, Part II, (Dec. 1933), Pp. 149-82; "Levy-Bruhl's Theory of Primitive Mentality", Bulletin of the Faculty of Arts, University of Egypt, Vol. II, Part I, (May 1934), Pp. 1-36. 6. Ibid.. Peter Winch seems to have misread this essay as a polemic against Levy-Bruhl, to whom he incorrectly at tributes a racist position concerning the intellectual capacities of primitives. It is quite clear from the context of the section of this essay of Evans-Pritchard's which Winch quotes, that it is not formulated as a cri tique of Levy-Bruhl. Cf. Ibid., Pp. 20-1. See also, Ibid., P. 16, where Evans-Pritchard states explicitly that Levy-Bruhl rejected the racist thesis concerning the determinants of intelligence, and makes it clear that Levy-Bruhl*^ problem was precisely why people who are quite obviously so intelligent in some respects should accept beliefs that are so absurd (one of the questions Evans-Pritchard himself sought to answer in his study of the Azande). Of course, Evans-Pritchard makes some searching criticisms of Levy-Bruhl*s theories, in the course of this essay. These criticisms are, how ever, of a quite different order from those imagined by Winch. See Winch, P., "Understanding a Primitive Society", American Philosophical Quarterly, Vol. I, No. 4, (Oct. 1964), Pp. 307-8. 210-7. The distinction which Evans-Pritchard draws between "mystic" patterns of thought (i.e., those that attribute supra-sensible and non-existent qualities to phenomena), and "non-mystic" thought patterns (based on empirical observation and logical inference), derives directly from Levy-Bruhl. Evans-Pritchard, E.E., Witchcraft.  Oracles and Magic among the Azande, (Oxford University Press, London, 1937), P. 12. 8. Ibid., P. 4. 9. Ibid., P. 5. 10. Ibid.. 11. See Pp. 22-7 of this thesis. 12. Evans-Pritchard, E.E., Witchcraft..., op. cit., P. 63. 13. Ibid., Pp. 63-4. 14. Ibid., P. 66. 15. Quoted in Gluckman, M., Custom and Conflict..., op. cit., P. 85. 16. Levi-Strauss, C, The Savage Mind, (Weidenfeld & Nicholson, London, 1966), P. 11. 17. Evans-Pritchard, E.E., Witchcraft..., op. cit., P. 74. 18. Ibid., P. 76. 19. Ibid.. Pp. 148-82. 20. Ibid.. P. 367. 21. Ibid., Pp. 359-74. 22. Ibid., Pp. 352-7. 23. Ibid., Pp. 262-3. -211-24. Ibid.. Pp. 258-351. 25. Ibid.. Pp. 84-98. 26. Ibid.t Pp. 26-9. 27. Ibid.. P. 438. 28. Ibid.t Pp. 438-9. 29. Such a view was, of course, consistent with the tendency to view primitive magic as a system founded on deceit and fraud. See Tylor, E.B., Primitive Culture. (John Murray, London, 1920), Vol. I, Pp. 133-4; Prazer, J.G., The Golden Bough, abridged ed., (Macmillan & Co. Ltd., London, 1960), Vol. I, P. 60. 30. Evans-Pritchard, E.E., Witchcraft..., op. cit., Pp. 100-1. 31. Ibid., Pp. 102-3. 32. Ibid., Pp. 104-5. 33. Ibid., P. 105. 34. Marwick has attempted to give this suggestion a quasi-mathematical formulation. Between certain structural relationships, tension exists, which is released in the form of witchcraft accusations where witch beliefs exist and where alternative means of expressing tension are not preferred or are non-existent. This tension (T) is conceptualized as a function (f) of the value placed upon objects competed for (V), the extent to which the relationship requires total involvement of personality (P), and the extent to which the relationship is tradi tionally ascribed, for example by norms pertaining to social distance (C). This yields the formula, Marwick, M., "Some Problems in the Sociology of Sorcery and Witchcraft", in Fortes, M. and Dieterlen, G. (eds.), -212-African Systems of Thought. (Oxford University Press, London, 1965), Pp. 172-3. 35. Evans-Pritchard, E.E., Witchcraft.... op. cit.. P. 105, 1967). 37. Evans-184-5. -Pritchard, E 38. Ibid. • P. 186. 39. Ibid., P- 184. 40. Ibid., , P. 195. 41. Ibid., , P. 191. 42. Ibid., , P. 373. 43. Ibid., P. 427. 44. Ibid., P. 313. 45. Ibid., P. 193. 46. Ibid., Pp. 330-•35. 47. Ibid., Pp. 330-•1. 48. Ibid., Pp. 314-15. 49. Ibid., Pp. 339-•40. 50. Ibid., Pp. 342- 3. 51. Ibid., Pp. 343- 4. 52. Ibid., P. 26. 53. Ibid., P. 29. 54. Ibid., P. 475. -213-55. Evans-Pritchard has recently reformulated this idea. Writing more generally of the collective representations of "closed societies", he says: "Everyone has the same sort of religious beliefs and practices, and their generality, or collectivity, gives them an objectivity which places them over and above the psychological ex perience of any individual, or indeed of all individuals. ... Apart from positive and negative sanctions, the mere fact that religion is general means, again in a closed society, that it is obligatory, for even if there is no coercion, a man has no option but to accept what every body gives assent to, because he has no choice, any more than of what language he speaks. Even were he to be a sceptic, he could express his doubts only in terms of the beliefs held by all around him." Evans-Pritchard, E.E., Theories of Primitive Religion, (Oxford University Press, London, 1965), P. 55. 56. Evans-Pritchard, E.E., Witchcraft..., op. cit., Pp. 475-8. CHAPTER EIGHT KLUCKHOHN*S ANALYSIS OF NAVAHO WITCHCRAFT A. Introduction Clyde Kluckhohn*s study of Navaho witchcraft was first published in 1944, thus following the publication of Evans-Pritchard* s Witchcraft, Oracles and Magic among the Azande (1937), by fully seven years. Yet Kluckhohn shows no sign of having read Evans-Pritchard *s work, which he nowhere re fers to. His interpretation, therefore, stands closer to Malinowski*s functional analysis of magic as cathartic for tension release, than to Evans-Pritchard*s structuralism. Moreover, even had he been familiar with Evans-Pritchard*s study, it is improbable that Kluckhohn would have in any fundamental way altered the substance of his approach, for although he knew the writings of the con tributors to the Armee Sociologique, and of A.R. Radcliffe-Brown, he remained unsympathetic to the project of estab lishing an autonomous, non-reductionist sociology. For him, culture always remained an abstraction. The real unit of study was thesconcrete human individual. Thus he wrote that, ... the French sociologists, Radcliffe-Brown, and - to a lesser extent - Malinowski are so interest ed in formulating the relations be tween conceptual elements that they tend to lose sight of the concrete human organisms.... the motivations -215-and rewards which persons feel are lost sight of ... Hence, as Dollard says, "What one sees from the cul tural angle is a drama of life much like a puppet show in which •culture1 is pulling the strings from behind the scenes." The realization that we are really dealing with "animals struggling in real dilemmas" is lacking.1 Por this reason, in place of explanation on a socio-cultural level, Kluckhohn tended to prefer psycho-analytic interpretations, even if these were sometimes (as in the case of Eoheim) "extravagant and undisciplined", or based 2 on flimsy evidence. Por, whatever their inadequacies, the psycho-analytic works at least had the merit of drawing attention to "... the connection between cultural forms and impul se-mo ti va te d o rgani sms." ^ KLuckhohn's work therefore falls within the tradition of the American "culture and personality" school, together with that of such other writers as Benedict, Margaret Mead, Eardiner, Erikson and Wallace. As a result of this inte rest, KLuckhohn's work has had little influence in British anthropology, which has shown little concern for inter relating the phenomena of psyche and society, especially in the period following World War II. Thus, although often referred to, KLuckhohn's ideas are rarely discussed in depth. Marwick would seem to express the general attitude of most British anthropologists to the "culture and per sonality" approach when he dismisses the relevance of KLuckhohn's theory of Navaho witchcraft with the comment that, 216-... since it is concerned with, the dynamics of individual behaviour, i.e. with the manner in which a hu man personality generates, converts, and disposes of aggression, it is a psychological theory largely irrele vant to the problems of sociological analysis.5 Such a rejection is, I believe, over-simplistic, for Kluckhohn did not offer us a psychological theory only. As Middleton points out, he also showed, together with Evans-Pritchard, "... that beliefs in magic and witchcraft are integral parts of cultural life and can therefore be under stood only in their total social context." Kluckhohn•s analysis does encompass a psychological theory, but its ambit is broader than this alone. For, as a study in the interrelations of psyche and society, it does not seek to reduce one level to the other, but also includes elements of a sociological theory. Kluckhohn therefore operates on two levels: the sociological and the psychological. B. The Socialization of Witch Beliefs The anthropologists of the nineteenth century sought to explain the origins of magical and religious belief. With Levy-Bruhl, Rivers and Evans-Pritchard, these beliefs came to be treated as given, and examined only in their structure and in their relation to societies of a particu lar type. Of these two divergent traditions of research, Kluckhohn stands closer to the first than to the second. He considered witchcraft beliefs to be "irrational", and to stand in need of an explanation he apparently did not think it necessary to give to our own intellectual structures. 217-He differed from the progressionists, however, in seeking to found this explanation not on an evolutionary theory and a simplistic associational psychology, but in terms of a model based on a situational psycho-analytic interpretation and a theory of sociological functionalism. Kluckhohn began by asking to what extent the early experiences of the Navaho child could be used to afford an explanation for the existence of Navaho witch beliefs. He suggested that, The child, even before he is fully responsive to verbalizations, begins to get a picture of experience as potentially menacing. He sees his parents, and other elders, confess to their impotence to deal with various matters by technological or other rational means in that they resort to exoteric prayers, songs and magical observances and to esoteric rites. When he has been linguistically socialized, he hears the hushed gossip of witchcraft and learns that there are certain fellow tribesmen whom his family suspect and fear. One special experience of early childhood which may be of con siderable importance occurs during toilet training. When the toddler goes with his mother or with older sister to defecate or urinate, a certain uneasiness which they mani fest (in most cases) about the con cealment of the waste matter can hardly fail to become communicated to the child. The mother, who has been seen not only as a prime source of gratification but also as an al most omnipotent person, is now re vealed as herself afraid, at the -218-mercy of threatening forces. Other early experiences also predispose the child to the acceptance of witchcraft beliefs. The Navaho child is frequently ill, and often witnesses illness and suffering on the part of others. Because of an inadequate diet, a lack of skills for dealing with European-introduced dis eases, and draughty hogans, the Navaho have a morbidity rate higher than that characteristic of typical white com-g munities. Hunger is also an early experience. Few Navahos reach adulthood without having experienced starva tion for short periods, and hunger rations for long periods. Since such privations occur even when people work hard and skilfully, n... it is small wonder that experience has a 9 capricious and malevolent component for most Navahos.n Kluckhohn recognizes, however, the objection which im mediately springs to. mind against this argument: that there is nothing in the experiences cited which need serve as a conditioning mechanism specific to the implantation of witchcraft beliefs (with the possible exception of faeces and urine concealment). Such conditions might reasonably be expected to give rise to some sort of belief in male volent or dangerous forces. But these could as well be ghost or spirit beliefs, beliefs associated with taboo in fractions, or perhaps even bacteriological theories, as be liefs in the existence of witches. To go, therefore, be yond the demonstration of the existence of a milieu favour able to the perpetuation of beliefs in a malevolent com ponent of reality, to a demonstration of why it is witch craft beliefs in particular which are perpetuated, it is -219-necessary to cite other causative influences besides the early experiences of the Navaho child. Kluckhohn considers that this may be done by systematically examining "... the contributions which the witchcraft pattern assemblage makes to the maintenance of personal and social equilibrium."1^ C. Kluckhohn*s Functionalism To Kluckhohn*s mind, the most important question with which we are faced in relation to Navaho witchcraft beliefs, 11 is why it is that these beliefs have continued to survive. In seeking to answer this question, he started from the basic premise (which was also that of Malinowski), that "... no cultural forms survive unless they constitute res ponses which are adjustive or adaptive, in some sense, for the members of the society or for the society considered as 12 a perduring unit." "Adaptive" is a purely descriptive term referring to the fact that certain types of behaviour result in survival (for the individual or for society as a whole). "Adjust!ve" refers to those responses which bring about an adjustment of the in dividual, which remove the motiva tion stimulating the individual. Thus suicide is adjustive but not adaptive.13 Kluckhohn therefore analyses witchcraft beliefs as functional on two levels: that of the individual, and that of society. To this basic distinction between social and psychic functions, Kluckhohn adds a second: that between 14 manifest and latent functions. Thus we are presented with an analysis of four types of function of Navaho witch--220-craft belief: i. manifest individual ii. manifest social iii. latent individual iv. latent social D. The Individual Functions of Navaho Witchcraft The practice of Navaho witchcraft was seen by Kluckhohn as; having one manifest function on the individual level: that of acquiring supernatural power. Witchcraft is a means which enables its practitioner to gain; wealth; and women, dispose of enemies, and "be mean". For those per sons seeking supernatural power, witchcraft presents an alternative route to that provided by the socially approved practice of becoming a "singer" (ceremonial practitioner). If a person is unable to become a singer because of a lack of the fees necessary/ to pay a teacher, or because of am inadequate memory, witchcraft is available as an antidote 15 to deprivation. Most of the data concerning Navaho witchcraft, how ever, relates not to the practice of it, but to the beliefs of others? concerning its practice. What, then, are the manifest individual functions of these beliefs? The first function suggested by Kluckhohn is that witchcraft stories have a high dramatic; value and, as a result, partly take the place which books, magazines, films, and plays GDcxsupy in our society. Secondly, witchcraft beliefs have an explanatory func tion. They explain howv death can oecur without visible cause, and why a person should be suffering from: a stubborn illness. More specifically, witchcraft beliefs help to reinforce the individual Navaho*s credence im the efficacy -221-of the curing ceremonials. If a chant has been performed flawlessly by a singer of great reputation, but neverthe less fails to cure, a Navaho still need not question the validity of the chant. A culturally acceptable explanation of its failure to cure lies ready at hand: the illness was caused by witchcraft, and it is for this reason that the chant has proved inefficacious. This line of reasoning gives an obvious excuse to a singer who has failed in his task, since the best cure for witchcraft is not a chant, 16 but a prayer ceremonial. Since acculturated Navahos typically continue to accept witchcraft beliefs long after they have lost all faith in Navaho medicine, however, the explanatory func tions of Navaho witchcraft were not considered by Kluckhohn to constitute sufficient explanation for their persistence. Why, he asked, should these beliefs show such tenacity in surviving "... at the expense of more rational modes of 17 explanation?". To understand this tenacity, he consider ed it necessary to examine the latent functions of witch craft for the individual. The first latent function suggested by Kluckhohn is that claims of bewitchment can serve as a device for getting attention by capitalizing on the credence of fellow believers. As such, this device might reasonably be ex pected to be employed more often by those occupying a low position in the social hierarchy. According to Kluckhohn, this supposition is in fact confirmed by the evidence: a high proportion of those who faint or go into trances at "squaw dances" or other large gatherings are women, or men -222-who are neglected or of low social status. The rich and powerful, on the other hand, tend to have bewitchment an nounced, or discovered by a diagnostician, in the privacy 18 of their own homes. Kluckhohn saw this function of witchcraft beliefs in the Navaho context as also being consistent with material reported from other cultures. For example, among the Tanala of Madagascar it is only those of low social stand ing, such as younger sons or sterile second wives, who are subject to tromba (a neurotic seizure accompanied by ex treme desire to dance). A similar phenomenon is reported for participation in certain trance states in Bali. Also consistent with this is the fact that, in Kluckhohn's opinion, New England and European witch trials were often started by publicity seekers, while "It is probable that in the South frustrated women make up a considerable pro-19 portion of the accusations which get Negroes lynched." The second latent individual function assigned by Kluckhohn to Navaho witchcraft, is that of providing a socially acceptable channel for the expression of the cul turally disallowed, thus permitting certain aberrant im pulses to achieve a release in fantasy. A man might, of course, day-dream about having intercourse with a dead wo man, but he would be likely to worry about the abnormality of this kind of fantasy and feel the immediate necessity of having the Blessing Way sung over him. But if the fantasy takes the form of repeating or manufacturing a witchcraft tale involving this element, or visualizing it while listening to someone else tell the story, the psychological -223-mechanisms of identification and projection can permit an 20 outlet in fantasy without conflict. Thirdly - and here we come to perhaps the most impor tant section of Kluckhohn*s analysis - is the function assigned to witchcraft beliefs of permitting the expression °f direct and displaced antagonisms. This is not to sug gest that Kluckhohn saw witchcraft beliefs as providing the only possible way of handling the problem of aggression in Navaho society: Fights occur; aggression is express ed against dead relatives as ghosts ... But if myths and rituals provide the principal means of sublimating the Navaho individual's anti-social tendencies, witchcraft provides one of the principally socially under stood means of expressing them.21 Witchcraft accusations, in other words, objectify and alleviate displaced anxieties which arise from the general situation of the Navaho, and from the special situation of particular Navahos at particular times. According to Kluckhohn, all Navahos feel personal insecurities, and must therefore, according to the evidence of clinical psy chology and psychiatry, be expected to manifest hostilities 22 against others. What, then, are the sources of these in securities? Ill health is a major source of worry to the Navaho, not only to the individual who falls ill, but also to the members of his or her household, who must perform addi tional tasks as well as pay for the appropriate curing ceremonies. Ill health is therefore productive of tensions, -224-and probably of unconscious antagonisms. In addition to this, there are economic hazards. Lightning may destroy a relative or a part of one's stock, frost may ruin a crop, and rain may come at the wrong season. Recent population growth, with attendant over grazing, has further increased these hazards, which often give rise to real hunger. Moreover, there is the pressure exerted on the Navaho by white society. The increase of both Navaho and white populations over the past generation has intensified the impact of the whites, who have encroached on lands needed by the Navaho themselves for expansion. The activities of Indian Service representatives and of other whites operat ing within Navaho territory have also intensified. Each new school generation faces a larger problem of compromis ing between the demands of the two cultures. As a result of this, ... the Navaho have come to feel themselves in an acutely uncomfort able situation. Indifference and withdrawal are no longer effective responses. They know they must de velop some suitable form of com promise with our civilization. At the moment they feel themselves ex ploited, surrounded by more powerful forces, "on the spot".2,3 These objective hazards thus make for personal inse curity, and therefore for the intensification of inter personal conflicts. But the basis for these reactions of hostility is not restricted to external conditions. Por, although external conditions create an atmosphere conducive -225-to social mistrust, the actual centring of hostile feelings upon other persons is accentuated by the Navaho world view itself. The Navaho world view, including Navaho mythology, proclaims that the active forces operative in the world are capricious and actively malevolent. These malevolent forces strongly tend to be personalized, both as ghosts and as witches. To the realistic dangers they confront, there fore, the Navaho react with fears - reactions dispropor-24 tionate to the dangers facing them. The dangers con fronted are not conceived of as operating neutrally, and as therefore being avoidable or subject to rational manipula tion. The joint effect of the Navahos1 objective situation, coupled with the ideology in terms of which they structure reality, is to induce strong feelings of personal inse curity and therefore to generate hostile impulses towards others. These hostile impulses must find an outlet, or they will be turned inwards to threaten the aggressor him-25 self, leading to varying degrees of paralysis of action. No society can survive in which the expression of hos tilities is not restricted and channelled. But, on the other hand, unless there are some forms of hating which are socially approved, everyone will remain in an intolerable conflict situation and neuroticism will be endemic in the population. This is not to say that aggression, whether overt or masked, is the only possible response. With drawal, passivity, sublimation and conciliation are also possible mechanisms for dealing with the situation: It would be too much to say that all societies must necessarily have their "witches", I.e., persons whom it is -226-proper to fear and hate and, under defined circumstances, to behave aggressively toward. "Witches" are not very prominent in the sentiment systems of some societies. But no culture which has yet been described leaves "witches" out of its defini tion of the situation for every sec tor of life or for every group within the society. "Witches" in this very general sense of "scapegoats" have probably played some part in all social structures since Palaeolithic times. Most contemporary European societies feature such witches quite obtrusively. These "witches" may be either a minority within the society or an external society. Thus the Nazis have had the Jews, the Fascists have their Communists and their "plu tocratic democracies"; "liberals" have the Jesuits (and vice versa). For a period of time the French had the Germans. To make the broadest possible structural comparison, Navahos blame their troubles on witches instead of upon "jews" or "niggers". We should be putting the matter over simply but not altogether incorrectly if we said that a belief in witchcraft was Navaho culture's substitute for "race pre judice".... But in place of selecting its "scapegoats" by the color of their skin or by their separate re ligious tradition or by their occupa tion, Navaho culture chooses certain individuals who are supposed to work evil by secret supernatural tech niques.2^ The existence of a universal pattern of scapegoating does not, of course, provide a sufficient explanation of the -227-particular characteristics of Navaho witchcraft. For the existence of a high incidence of hostile impulses does not necessarily mean that the overt expression of aggression need be especially common. The Navaho, in fact, have many other ways of handling hostile impulses. There is, for example, a high incidence of hypochondria. There is also the "flight from reality" represented by the consumption of alcohol and, in Kluckhohn*s opinion, the use of peyote. Passivity and social withdrawal are also available responses, and patterns for the reconciliation of individuals who have quarrelled also exist. Thus a dispute between two close relatives may be brought to a close by the attendance of one at a ceremonial sponsored by the other. More direct outlets for aggression exist in the form of verbal conflicts, gossip and drunken fights. Central to Kluckhohn*s thesis, however, is the argument that none of these mechanisms is sufficient to handle existing hostile impulses. Thus, while peyote is popular in restricted areas, its use is strongly opposed by the Indian Service and also conflicts with the native religion. Similarly, the use of alcohol is sharply curtailed by Indian and white police. It is therefore not available on a ready basis except to those Navahos who live in close proximity to white boot leggers. Social withdrawal, on the other hand, is not an effective response because of the nature of Navaho types of shelter and the necessity of economic co-operation. There are strong prohibitions against the overt ex pression of aggression, which is excused only in cases of drunken fights. The helplessness of people in a scattered -228-and poorly policed society makes the playing down of overt aggression a highly adjustive response. Moreover, those against whom aggressive feelings are most likely to he har boured are the members of one*s own consumption group, with whom economic co-operation is most necessary. Given the slim margin of subsistence, pressures on the individual to co-operate with his affinal and consanguineal kin are very strong and difficult to resist. In addition to this, the small size and relative isolation of the consumption group was considered by Kluckhohn to be an extremely important factor. Often very small, ranging in size from two to six households, and situated at least one mile and often close to ten miles from other consumption groups, these permit only a very limited range of social contacts. Interaction with outsiders being thus limited, the result is n... a strong tendency toward involvement in a morbid nexus of emotional sensitivities from which there is little escape 27 through socially approved patternsy Whites have added enormously to the sources of ag gression in Navaho society, but have supplied only one out let for tension release: alcohol. The general tendency of whites has been towards the prevention of aggression re lease. This is not only the case with the taking of peyote and the consumption of alcohol, but is also true for the overt expression of hostile impulses. In the old days, if a man beat his wife she might leave him, or he might have to pay a fine to her family. Today, he might also have to spend several months in gaol. Especially important is the fact that the outlet of -229-war has heen cut off. Por several centuries, warfare was an exceedingly important part of Navaho culture. Now, for two generations, access to war has been denied, except to those few Navaho men who choose to join the U.S. Army. In Kluckhohn*s opinion, "With organized extra-societal ag gression denied, it seems probable that intra-societal ag gression has mounted."2® Taking all these factors into account, Kluckhohn1s conclusion was that witchcraft represents an eminently ad-jus tive cultural solution to the problem posed to Navaho individuals by hostile impulses. Thus, instead of directly expressing his bitter feelings towards his father-in-law, threatening his own economic security and incurring social disapproval by doing so, a Navaho man may vent his spleen against a totally unrelated witch. He will be relatively safe if he is careful in his choice of intimates to whom he speaks, and perfectly safe if he rages against a witch who lives in a locality a hundred miles or more away. According to Kluckhohn, a high proportion of gossip refers to distant witches, this making Navaho witchcraft much more adaptive than is witchcraft in those societies, like the Zuni, where accusations are centred within the local group. A witch who is seen only rarely, or perhaps even never, is sufficiently distant to make the origin of feuds unlikely. Most Navaho witchcraft gossip, then, is 29 directed against out-group members. There is, however, also some gossip and accusation of witchcraft directed against in-group members. Kluckhohn's suggestion was that this is explained by the fact that the 230-gratifications derived from gossiping about a witch living in a distant locality are too dilute to be deeply satisfy ing. Sometimes, there may even be socially sanctioned physical aggression expressed against an in-group member. The fact that the killing of witches is uniformly described as violently sadistic suggests that these acts gained huge increments of displaced aggression.... direct aggression among the Navaho, whether verbal or behavioral, seems commonly accom panied by displaced aggression. Quarrels have a fury that is often ridiculously out of proportion to the alleged grievance.'" Kluckhohn did not, however, hold the opinion that those accused of being witches are always the victims of scape goating processes. Sometimes, accusations of witchcraft may be made against the actual targets of hostile feelings. A singer may whisper accusations against a professional rival. A wealthy neighbour might be gossiped about as be ing a witch. If a man's wife runs off with another man, the jilted husband might explain this by saying that his rival used Frenzy Witchcraft. Peuds between extended family groups over land and water rights may also be formulated in terms of witchcraft accusations. Kluckhohn gives one example where a local group complained to the Indian agent about the drinking and fighting of its neighbours. The neighbours responded by threatening to kill an old man of the opposing group for being a witch. Navahos sometimes also gossip cautiously against certain relatives, such as the maternal uncle, and especially against affines. This -231-is especially true of women who are forced to live with their husbands* kin* In the reverse case, of uxorilocal residence, the husband has frequent excuses to leave his affines and visit his own family. A woman residing viri-locally, on the other hand, does not have this option open to her, since she is required to cook for her husband and care for her children. In such cases, the wife develops increasing antagonisms towards her affines. One of the few ways in which she can discharge these antagonisms is by murmuring witchcraft accusations, against her affines, to her own blood relatives.^1 According to. Kluckhohn, two important psychological mechanisms are operative in this process of releasing tensions through gossip about witchcraft: identification and projection. By identifying with the witch-aggressor, the gossip may discharge his or her own dispositions to wards cruelty. And by attributing to others the aggressive feelings which are in felt those of the gossip himself, re lief may be obtained from feelings of guilt. The fear of retaliation is expressed and objectified. As Kluckhohn himself puts it, Witchcraft tales, gossip and accusa tions supply, then, a readily avail able means of covertly expressing aggression and of objectifying fears consequent upon one's own aggres sions whether overt, symbolic or re pressed. Folk belief channels these fantasies into witchcraft patterns and documents them with culturally appropriate details.32 Kluckhohn*s analysis of witchcraft accusation and -232-gossip as providing a means for the release of hostile im pulses was not, however, confined within the parameters of a model defining witchcraft as a mechanism providing an outlet only for those tensions consequent on Navaho condi tions of life in general. He also argued that witchcraft affords a canalized relief for the specialized tensions which are a product of Navaho social structure and cul tural configurations. Thus he addressed himself to the question of why it is that the killing of siblings is so often mentioned in descriptions of witchcraft, as the price of initiation for example. His suggestion was that this may he explained by reference to the process of socialization. Kluckhohn began by suggesting that sibling rivalry is caused by the fact that new-born children displace their brothers and sisters from their previous position as the centre of affection, thus generating hostile impulses. Secondly, because of the large size of Navaho families, siblings are responsible for the socialization of their younger brothers and sisters. In this way, they act as frustrators, and are not able to compensate for this frus tration by supplying the intense gratification formerly supplied by the mother. In addition to this, the older siblings probably resent the restrictions placed upon their activities by the necessity of caring for their younger brothers and sisters. Kluckhohn further argued that sibling rivalry at the period of weaning and infancy leads to envy and hostility at the adult level between siblings who receive coveted -233-property from their parents (sisters) or who are competi tors for a desired woman (brothers). The property of a deceased relative is usually divided equally between hard working and indolent brothers. The shame of a disgraceful act falls almost as strongly on the siblings as on the perpetrator. The frugal and careful brother and sister are under strong pressure of public opinion to look after a lazy and improvident sibling. Thus, despite the exist ence of many positive feelings towards siblings, these are perhaps best characterized as ambivalent. The negative side to this ambivalence is expressed in terms of the negative relations described as pertaining between witches and their siblings.33 Kluckhohn also gave some consideration to the ques tion of why it is that old people are more often accused of witchcraft than young people, and of why it is that wo-34 men are less often accused than men. So far as the old are concerned, it was Kluckhohn1s opinion that there is a basis of resentment against them because of the fact that they are usually an economic liability. In addition to this, this particular weighting of witchcraft accusations is consistent with the Navaho cultural configuration of distrust of extremes: the very old, the very poor, the very wealthy, and very powerful singers. Thirdly, the Navaho set a very high value on long life, and feel that the aged have attained something greatly to be prized. It is, perhaps, felt that the aged have a power comparable to that of a singer. The aged have passed, as it were, -234-from the realm of "the profane" to that of "the sacred". Perhaps, rather ... the very old person is about to lose "something prized" and resists this - at the alleged ex pense of others.... Certainly the closeness to death is the critical factor. Navahos say, in effect, that the very old will die so soon anyway that they will take all sorts of chances with the culturally pro hibited for the sake of immediate gain where the younger person would feel he had too much at stake in the long run and therefore sticks closer "to the good side".35 Kluckhohn also believed that the Navaho feel that the aged are "almost ghosts" and that the very old, being near to death, partake of some of death's attributes. This seems to be another overlapping or linking of ghosts and witches. Ghost belief permits the expression of hostility felt toward dead rela tives, witches that felt toward liv ing relatives. Ghosts are, as it were, the witches of the world of the dead. A striking corroboration of these "psychological" interpreta tions of beliefs regarding ghosts and witches is the astonishing fact that, according to some informants, the only ghosts one can see are those of relatives....36 So far as the lesser frequency of accusations against women is concerned, Kluckhohn offered a purely culturo-logical explanation. He considered this phenomenon to be explainable by the fact that "... the solidarity of Navaho society centers in women", together with the fact that few -235-women become either singers or curers. Women witches are almost always either childless or past menopause, a fact explained, in Kluckhohn's opinion, by the unwillingness of the Navaho to attribute such evil to those who are bearing or bringing up children, "... for they are the focus of the 37 sentiment system." In conclusion, it must be noted that Kluckhohn's in terpretation did not follow the line of arguing that witch craft must exist because of the amount and types of aggres sion prominent in Navaho society: only that, given these conditions, some form of release must exist. The other mechanisms of release being inadequate, and the witchcraft patterns being historically available, the witchcraft mechanism constitutes a highly adjustive way of releasing tensions deriving from the general situation of the Navaho and the particularities of their social organization. As Kluckhohn himself summarizes his thesis, ... witchcraft is a major Navaho in strument for dealing with aggression and anxiety. It permits some anxiety and some malicious destructiveness to be expressed directly with the mini mum of punishment to the aggressor. Still more anxiety and aggression is displaced through the witchcraft pattern assemblage into channels where they are relatively harmless or where, at least, there are available patterns for adjusting the individ uals to the new problems created. Individual adjustment merges with group adaptation.3" -236-E. The Social Functions of Navaho Witchcraft The principal manifest function which Kluckhohn as cribed to Navaho witch beliefs on the social level, was that of providing a dramatic definition of all that is con sidered bad: i.e., all secret and malevolent activities directed against the health, property or lives of fellow Navahos. Thus, the solidarity of the group is affirmed by these beliefs. This sanction is made even stronger by attributing to witches the most evil kinds of knowledge and activity: nakedness, incest, the murder of siblings, inter-39 course with dead women and cannibalism, for example. In addition to this, witchcraft also has a number of latent functions which make for the preservation of the equilibrium of the society and of the local group. The first of these mentioned by Kluckhohn is that of preventing too rapid a rise in social mobility, thus impeding the un due accumulation of wealth. The wealthy Navaho knows that, if he is stingy with his relatives and with others, he is likely to be spoken of as a witch. One of the most basic strains in contemporary Navaho society is that induced by the conflict between the demands of familism and the emula tion of European patterns of capital accumulation. In Kluckhohn*s opinion, the best hope for the coherence of Navaho culture and the integrity of the Navaho way of life lies in there being a gradual transition from the familis-tic type of social organization to one more closely re sembling the European. Therefore, any pattern which tends to discourage the too rapid accumulation of wealth in the hands of a restricted number of individuals makes for the -237-40 survival of the culture. Secondly, witchcraft beliefs serve as a brake on the power and influence of ceremonial practitioners. These practitioners are, in effect, warned that their capacities for influencing the course of events by means of super natural techniques must be used for the realization of socially desired ends only. There are many specific paral lels between accounts of witchcraft and chant practice, such as the use of whistles» sandpaintings, pollen, songs and turquoise, which show that witchcraft is thought of as being bad ceremonialism. Singers are valued, but dis trusted. The possibility of witchcraft gossip, and perhaps even of trials and executions, ensures that these prac titioners will think twice before they abuse their powers. How this works in the concrete is well illustrated by an incident which occurred at Red Rock in 1937. A man went to his maternal uncle and asked him to sing over his wife. The uncle was not satisfied with the number of sheep offered and evaded performing the chant. The wife died. The ne phew first accused his uncle as a witch but was unable to muster family or community support for disciplining. He thereupon killed the uncle himself. Such events must surely reinforce the disposition of singers to be liberal to their kin and prompt in acceding to requests for ceremonial help. In fact, witchcraft functions far more widely than this as a technique of social control. Perhaps the clear est example of this is that provided by Manuelito who, con vinced that the only hope for his tribe lay in preserving 238-peace with the whites, brought about the execution, on grounds of witchcraft, of more than forty leaders, through out Navaho country, who were advocating another armed struggle against the whites. This was in 1884.*2 Usually, witchcraft operates as a technique of social control on a far less conscious level. The introduction of the Ghost Dance religion into Navaho territory was blocked by spreading the word that the proponents of the new cult were witches. Similarly, any powerful instigator of trouble in Navaho society tends to be spoken of as a witch. This tendency operates to reduce inter-group friction. The fact that anyone acting "mean" is likely to be accused of being a witch means that there exists a powerful deterrent to the perpetration of any hos tile acts. In addition to this, the existence of witch beliefs helps ensure the performance of socially required duties. The aged, the destitute - even animals - must not be ne glected or oppressed, in case they should resort to witch craft. Siblings must be aided when ill, lest the death of a sibling give rise to suspicion that a survivor is learn ing witchcraft. The effectiveness of leaders is also sometimes in creased by the fear that they are witches and will bewitch those who fail to follow them. This is important since Navaho society, lacking an institutionalized leadership, is especially vulnerable when faced by a society structured as our own is. Kluckhohn considered the survival of the Navaho to be favoured by any sanctions which assist the 239' formation of a united front behind leaders of some perma nence . Kluckhohn was even prepared to consider that the witchcraft-induced fear of going about at night has a defi nite social value. One of the principal sources of fric tion in Navaho society is sexual jealousy. The fear of witches at night operates as a deterrent to extra-marital relations, since night-time would otherwise provide favour-43 able opportunities for secret rendez-vous. F. The Cost of Navaho Witchcraft More systematic than Malinowski's functionalism, Kluckhohn's variant might also be considered a little more dialectical, in that Kluckhohn considered not only the con tributory relationship of social facts to their total con text but also gave some attention to the dysfunctional as pects of such relations. This notion of dysfunction, he expressed by the term "cost". Kluckhohn considered that, given the general condi tions of Navaho life, together with the fact that the Indian Service would always prevent a wholesale slaughter of suspected witches, Navaho witchcraft was essentially not only an adjustive but also an adaptive structure. Its cost lies in the fact that it projects aggression and causes some social disruption. In many cases witchcraft belief un doubtedly does more to promote fear and timidity than to relieve aggres sive anxieties. The fears consequent upon witchcraft tend to restrict the life activities of some persons, to curtail their social participation. -24C Perhaps the witchcraft pattern assem blage tends to be mainly adjustive for individuals who tend to be ag gressive, mainly disruptive for those who tend to be non-aggressive. Such a view would fit well with the sug gestions which have been made of the relationship between witchcraft pat terns and war patterns.** Kluckhohn further suggested that witchcraft fears some times form the basis of a reluctance to undertake or con tinue the burdens of leadership. A Navaho who is respon sible for making important decisions is certain to antago nize some people, who might resort to witchcraft in their resentment. Similarly, counter-balancing the tendency which witchcraft exerts in the direction of economic le velling, is the fact that the rich and powerful, inasmuch as they are feared and dreaded as witches, possess a power 45 and instrument of domination. G. Historical Fluctuations in Witchcraft Activity The final question to which Kluckhohn addressed him self - and here, too, his concern must be seen as an ad vance when compared with the predominant interests of con temporaneous British functionalism - was that of accounting for historical fluctuations in the frequency of witchcraft accusations among the Navaho. KLuckhohn's data indicated two periods when witchcraft activity was particularly prominent among the Navaho: i. 1875 - 1890 ii. late 1930's - early 1940«s. The first period of intensified witchcraft activity 241-followed on the defeat of the Navahos by the whites, and their imprisonment at Port Sumner. A people who for years had been the scourge of other Indians, and of the Spanish, were subjugated for the first time. A people who did not understand group captivity, and who had been accustomed to move about freely over great spaces, were taken captive and held in a limited area in a flat, colourless region quite unlike the rugged landscape to which they were accustomed. The Navaho were reduced to dependence upon others for their food. These foods were themselves unfamiliar, and at first considered to be highly distasteful. The Navaho were thus subjected to potent sources of anxiety, and therefore of hostility. At the same time, an enforced peace ended wars against other groups, thus removing the principal outlet for aggression. This subjection to white domination also weakened the internal social controls of clan organization. Indian Service officials and white-controlled courts took over some of the responsibilities formerly assumed by the clan. In the days before Port Sumner, all the members of a given clan were held to be responsible for the activities of their fellow clansmen. If there was evidence that a member of clan A had practised witchcraft against a member of clan B, the whole of clan A would have to contribute to pay a fine. This practice meant that severe punishments were dealt out to a culprit by his clan relatives, and that any connection with witchcraft was a non-rewarding business. In the period following the confinement at Port Sumner, on the other hand, although community pressures were still -242-strong enough to prevent internecine aggression, the fact that the Indian Service ignored witchcraft (except when a witch was executed), meant that the means of holding down witchcraft practice, gossip, accusations and trials were no longer effective. While some of the early Indian Service agents acquiesced in, and possibly even advised, the execu tion of witches, the recent Service and the white courts have refused to recognize the existence of witchcraft. In this way, the position of the witch has been improved by recent conditions. Witches have been placed in a position where they may practise indirect extortion, since they continue to be feared by the Navaho and yet have been plac ed by white government agencies in a position relatively immune from retribution. Many Navahos resent the fact that the white government fails to punish people "for the worst 46 crime we know".. It is important to note that the period of increased witchcraft activity did not occur at Port Sumner, nor immediately after. Kluckhohn likened this to the fact that Italian Fascism did not take over immediately following the Treaty of Caporetto, or Nazism immediately after the Treaty of Versailles. There always seems to be a period when a people are too crushed for any single response to become peculiarly frequent, when the general state is that of many varying trial and error efforts toward a new readjustment. Destructiveness always seems to ap pear in social reconstruction after a major drama.^ As the increase of witchcraft activity in this earlier -243-period was explained by KLuckhohn as the result of an in crease in tensions and hostilities consequent upon depriva tions and social disorganization, so also was the period of increased witchcraft activity in the late 1930*s and early 1940's. The Wall Street Crash of 1929, and the following Depression, had a much deeper impact on Navaho life than did previous slumps in the American business cycle since, earlier, the Navaho were far less integrated into the U.S. economy. The 1929 slump, however, meant that the Navaho suffered because of the fall in the market value of sheep and wool, the loss of markets for their craft products, and the absence of employment at wage-work. Only shortly later, the U.S. government embarked on an extensive stock reduction programme, forcing the Navaho to reduce the size of their flocks of sheep in order to preserve range lands. As a result, land became a critical problem, and the number of Navahos who became dependent on the U.S. economic system in order to survive was further increased. Thus this period saw a sharp increase in the amount of witchcraft activity, as well as an increase in the taking of peyote and the emergence of new nativistic cults, and an acute dissatisfaction with the Navajo Service. In those areas where white pressure was at a maximum, witchcraft 48 activity was also at a maximum. Kluckhohn therefore saw the historical data which he analysed as being essentially confirmatory of his general thesis: that witchcraft among the Navaho represents a res ponse to conditions of deprivation where social conditions do not provide adequate means for the release of consequent -244-hostilities in other ways, and where beliefs support the efficacy of magical practices. H. A Critique of Kluckhohn It has already been suggested that Kluckhohn's variant of functionalism may be considered methodologically superior to those of his British contemporaries, Radcliffe-Brown and Malinowski, on two counts. Firstly, because Kluckhohn balanced his notion of function (the contribution which a cultural element makes to the perpetuation of a social system considered as a perduring unit, or to the adaptation of the individual to his natural and social environment) with that of cost (the "losses" such elements incur for x 49 the on-going social system). Secondly, because Kluckhohn tried to view his data in the perspective offered by a time-depth analysis. Socio-cultural theories can, in fact, be tested in only two ways: by the method of cross-cultural comparison, or by the method of historical analysis. If Kluckhohn did not offer us the former, he did attempt the latter. As far as cross-cultural analysis is concerned, Kluckhohn himself explicitly stated that he did not think that the conditions he specified were, by themselves, suf ficient to produce witchcraft, either in Navaho society or in any other: "My thesis is only that given these condi-50 tions some forms of release must exist." Kluckhohn was, in fact, compelled to assume the prior existence of the beliefs he was trying to account for: When other forms are inadequate, and when the witchcraft patterns were -245-historically available, witchcraft belief is a highly adjustive way of releasing not only generalized tension, but also those tensions spe cific to Navaho social structure." As Swanson says, however, it is possible for us to go a step further than this, and ask whether the conditions specified by Kluckhohn are in fact necessary or sufficient to produce 52 endemic witchcraft. Cross-cultural comparison does not in fact bear out the strong emphasis placed by Kluckhohn on the importance of deprivations consequent upon lack of environmental con trol. Swanson1s analysis of fifty societies indicates that no relation is to be expected between the nature and abun dance of the food supply and the frequency with which sor cery is experienced. Swanson's data suggests rather that witchcraft tends to be associated with the necessity for social interaction between people on important matters in the absence of legitimated social controls and arrange-53 ments. This implies that Kluckhohn has given an inap propriate weighting to some of the factors he has suggested are important in fostering witch beliefs. Por example, in considering the importance of the Port Sumner period, Kluckhohn should perhaps have placed less emphasis on the physical deprivations associated with the imprisonment, and examined in more depth the implications of the weakening of internal social controls as a result of surrender to white domination. This, however, must remain at the level of hypothetical suggestion. Por, as a study of statistical correlations, Swanson*s findings do not necessarily imply a -246-causal relationship (or the absence of such a relationship) between any two elements in any particular society. Kluckhohn1s Navaho analysis is not therefore essentially called into question. Perhaps one of the difficulties of Kluckhohn's model is that it is not readily testable. This is what consti tutes one of its chief defects in Marwick's eyes, who writes that: As yet we have no satisfactory measures of frustration, of aggres sion, or of anxiety that could be used for establishing a relationship between these largely subjective con ditions and their release or allevia tion in the standardized delusions of a system of sorcery or witchcraft.^4 Nonetheless, I think that Marwick is wrong to dismiss Kluckhohn's theories out of hand, as he does here, for there are precious few hypotheses in the social sciences which conform to the rigorous requirements of testability that he demands. The fact that a concept cannot immediately be quantified does not mean that it may not serve as a useful tool for social research. Moreover, while Marwick may in tend to be disparaging when he refers to the phenomena of frustration, aggression and anxiety as "largely subjective", he does not thereby diminish their relevance for analysts of human affairs. We may conclude that Kluckhohn's model accounting for the persistence of witchcraft beliefs among the Navaho re mains hypothetical. He has indicated an important series of factors which need to be investigated in studying witchcraft. -247-But so far as his own interpretation of hos those factors are structurally related in Navaho culture is concerned, we have no real way of checking its validity. Basically, our acceptance or rejection of Kluckhohn's analysis is likely to he determined less by whether or not we think the analysis makes sense of Navaho material, than by our general attitudes towards the methodological bases of the analysis: i.e., functionalism, and the culture and personality approach. Parsons and Vogt have referred to Kluckhohn's "eclecticism 55 in theoretical matters". Applying this insight to Kluckhohn's study of Navaho witchcraft, we may say that he did not so much provide us with a theory in explanation of a problem, but rather, that he suggested some elements, from different theories, which could be considered in relation to a problem. -248-Notes and References 1. Kluckhohn, C, "Myths and Rituals: A General Theory", Harvard Theological Review, Vol. XXXV, (Jan. 1942), P. 58. Bobbs-Merrill reprint, S-146. 2. Ibid., P. 59. 3. Ibid.. 4. Thus, in a recent anthology entirely devoted to the analysis of witchcraft and sorcery, Kluckhohn's work is accorded only two - passing - references. See Douglas, M. (ed.), Witchcraft: Confessions and Accusations, (Tavistock Publications Ltd., London, 1970). 5. Marwick, M., "The Study of Witchcraft", in Epstein, A.L. (ed.), The Craft of Social Anthropology, (Tavistock Publications Ltd., London, 1967), P. 238. 6. Middleton, J., "Introduction" to Middleton, J. (ed.), Magic, Witchcraft, and Curing, (The Natural History Press, Garden City, New York, 1967), P. x. 7. Kluckhohn, C, Navaho Witchcraft, (Beacon Press, Boston, 1967), Pp. 77-8. 8. Kluckhohn quotes "a prominent officer of the Indian Medical Service", who estimates the incidence of ill-health to be about three times higher among the Navaho than in average white communities. Kluckhohn, C, "Myths and Rituals...", op. cit., P. 72. 9. Kluckhohn, C., Navaho Witchcraft, op. cit., P. 79« 10. Ibid.. 11. Ibid., P. 77. 12. Ibid., P. 79. According to Kluckhohn, there are no functionless cultural elements. Even the buttons sewn on the sleeve of a coat or jacket serve the "function" -249-of maintaining a tradition and constitute an adjustive response, since people feel more comfortable if they feel a continuity in behaviour. "Similarly, saying 'five houses' when 'five house' would be sufficient at the level of manifest 'function' is to be understood as fulfilling the latent 'function' of bolstering the indi vidual's security through adherence to a familiar and established tradition." Ibid., P. 81. As Robert Merton remarks, in thus seeking to ascribe "function" to seem ingly functionless items, Kluckhohn has fallen back upon a type of function which would be found by definition rather than by empirical investigation. Thus the impu tation of a "function" to the wearing of buttons on the sleeve of a coat adds little or nothing to a direct de scription of the cultural behaviour in question, while Kluckhohn's argument about the "function" of conforming to a tradition "... is equivalent to saying that the 'function' of conformity to any established practice is to enable the conformist to avoid the sanctions other wise incurred by deviating from established practice." Merton, R.K., Social Theory and Social Structure, (The Free Press, Glencoe, 1949), Pp. 32-3. Emphasis in ori ginal. It may be added that to hold that every cultural element must be functional is to divert attention away from the possible dysfunctional aspects of such elements, and makes the understanding of endogenous social change almost impossible. Moreover, Kluckhohn's example of the "function" of saying "five houses" in place of "five house" ignores the fact that it is only by following such linguistic conventions that communication is at all possible. Functional analysis of this type therefore diverts attention away from the understanding of meaning. 13. Kluckhohn, C, Navaho Witchcraft, op. cit., P. 79. 14. Ibid., Pp. 80-1. Kluckhohn borrows the terms "manifest function" and "latent function" from Robert Merton. In Merton*s usage, "manifest function" refers to the ob jective consequences of an item of cultural behaviour for a specified unit (person, group, social system, etc.), which contribute to the adaptation or adjustment of that unit, and which were so intended. "Latent function" refers to unintended and unrecognized con--250-sequenees of the same order. Merton*s purpose in intro ducing this distinction was to clarify the analysis of "seemingly irrational social patterns": ile., those patterns of behaviour which continue to persist even al though they clearly do not achieve their manifest pur poses. As an example of such a pattern, he cites the Hopi Rain Dance. Merton argues that the persistence of this institution is not explicable in terms of the ir rationality of the Hopi, but in terms of its functions for the Hopi social order. Kluckhohn*s use of these terms is far less clear than that of Merton, and he seems to employ "manifest function" in the sense of "common-sense explanation", and to use "latent function" to refer to functions below the threshold of superficial observation. Merton, however, seems to accept Kluckhohn*s usage as conforming to his own. Cf. Merton, R.K., Social Theory..., op. cit.. Pp. 62-8. 15. Kluckhohn, C., Navaho Witchcraft, op. cit., P. 82. 16. Ibid., Pp. 82-3. 17. Ibid.. P. 83. 18. Ibid., Pp. 83-4. 19. Ibid.t P. 84. 20. Ibid., P. 85. 21. Ibid.. Emphasis in original. 22. Ibid., P. 88. 23. Ibid., P. 87. 24. Ibid.. 25. Ibid., P. 88. 26. Ibid., Pp. 89-90. Emphasis in original. 27. Ibid., P. 94. -251-28. Ibid., P. 95. 29. Ibid., Pp. 96-7. But see note 18 on Pp. 45-6 of this thesis. 30. Kluckhohn, C, Navaho Witchcraft, op. cit., P. 98. 31. Ibid., P. 100. 32. Ibid., P. 101. 33. Ibid., Pp. 102-3. 34. Of a total of 222 accusations, where other information was also available, Kluckhohn found that: i. all of the accused were adults ii. 184 of those accused were men iii. 138 of those accused were women iv. 131 of the men accused were definitely old (i.e., described as "white-haired", etc.) v. all of the women accused were old vi. no women were accused of Wizardry or Frenzy Witchcraft vii. 140 of the men accused were ceremonial practitioners viii. 12 of the women accused were ceremonial practi tioners ix. 21 of the men accused were "headmen" or "chiefs" x. 115 of those accused were "rich" or "well-off" xi. 17 of those accused were poor or very poor. For 90 of the group, there was no data on economic status. Ibid., P. 59-35. Ibid., Pp 36. Ibid., P. 37. Ibid., P. 38. Ibid., P. 104-5. 105. 106. Emphasis 110. in original. 39. Ibid.. 40. Ibid., P. 111. -252-41. Ibid., P. 112. 42. Ibid.. 43. Ibid.. Pp. 112-13. 44. ;Ibid., P. 121. 45. Ibid.. 46. Ibid., P. 116. 47. Ibid., P. 114. 48. Ibid., Pp. 117-18, 120. 49* A pertinent question here, however, is whether or not Kluckhohn gave sufficient attention to the dysfunctional aspects of Navaho witchcraft. Por a much more negative evaluation of the implications of witchcraft beliefs for a society, see Nadel, S.P., "Witchcraft in Pour African Societies: An Essay in Comparison", in Marwick, M. (ed.), Witchcraft and Sorcery, (Penguin Books Ltd., Harmondsworth, 1970), P. 279. 50. Kluckhohn, C., Navaho Witchcraft, op. cit., P. 106. 51. Ibid., Emphasis mine. 52. Swanson, G.R., The Birth of the Gods, (University of Michigan Press, Ann Arbor, 1964), P. 144. 53. Ibid., Pp. 151-2. 54. Marwick, M., "The Study...", op. cit.. P. 238. 55. Parsons, T. and Vogt, E.Z., "A Biographical Introduction", to Kluckhohn, C, Navaho Witchcraft, op. cit., Pi xiii. CHAPTER NINE THE IMAGE OP SCIENCE AND THE UNDERSTANDING OP WITCHCRAFT A, Introduction Ih the course of this thesis, I have traced the gradual development of anthropological theories concerning the nature of magic and witchcraft up to and including Kluckhohn*s sophistication of the psycho-functional model, and Evans-Pritchard's analysis of Zande witchcraft in terms of what proved to be the starting point for a structuralist analysis of social life. Always present in the background of this discussion, has been the question of the relation ship of magic and witchcraft to Western modes of thought -Western science in particular - and the image which the anthropologist has entertained of the nature of science and scientific: method. In this chapter, I propose to examine this question a little more? explicitly, to suggest some of the inadequacies in the way in which anthropologists: have conceptualized scientific method, and to indicate some of the shortcomings of Evans-Pritchard's approach. B. The Progressionists? Tylor did not attempt to discuss the question of the relationship of magic to science, although implicit in his position was a differentiation between a subjective, fal lacious association of phenomena (magic), and an associa tion of phenomena between which the link is objective or -254-real (science). A more explicit discussion of this rela tionship had, however, to await Prazer. Frazer's attitude towards science is somewhat more complex than that implicit in Tylor's position, and it is my belief that, as Jarvie and Agassi have pointed out,1 Prazer adhered to two contradictory notions of science. In places, Prazer seems to have advanced a highly sophisticated view of science, as consisting in the ac tivity of formulating explanatory hypotheses ("the laws of nature are merely explanatory hypotheses"), which are then tested against experience and rejected; if falsified. There; is no final truth in science: truth is only the hypothesis that is "found to work best". Thus, "the advance of knowl edge is an infinite progression towards a goal that for ever recedes."^ Such a view of science is, in essence, identical with that elaborated by Sir Karl Popper, and it is unfortunate that anthropologists could not have developed it further. Por although ultimately unsatisfactory (at least as de veloped by Popper), it is far superior to that actually accepted by the vast majority of anthropologists; since Prazer's time.^ Most anthropologists have in fact adhered to the other view of science present in Prazer's writings. This is a view of science as representing an accumulated mass of em pirical observations. Prom the sum of these observations, scientific laws are derived by a process of induction. Be ing thus descriptive of empirical reality, scientific laws -255-are necessarily true. They may be found to be restricted in their application to certain cases only, or to be only a particular example of a more general principle operative in nature. Laws qua laws, however, are by definition true. What are thought to be laws, and are later found not to hold true, are misinterpretations of the empirical data. In this view, science is to be identified with the facts, theories and methods- collected in scientific texts, and scientific development is the piecemeal process whereby all these elements have been added to this ever-increasing body of technique and knowledge. Science is; therefore descrip tive and cumulative. As Prazer was consciously engaged in the process of comparing magic to science, and trying to establish the similarities and differences between the two, it is obvious that his criteria of magic must be ambiguous to the extent that his criteria of science also were. Nevertheless, this ambiguity does not prevent us from perceiving the elose re lationship between magic and science as described by Prazer. According to him, both are made up of hypotheses seeking to reduce the universe to order, and both assume that the: uni verse is governed by unconscious, impersonal forces, not subject to variation. Both therefore constitute a single type when opposed to religion, "... a propitiation or con ciliation of powers superior to man which are believed to 4 direct and control the course of nature and of human life." Yet what is the dividing line between magic and science? This is far less clear, precisely because of the ambiguity -256-of Frazer's notion of science. Acceptance of the first view offered of science would seem to indicate the neces sity of some criteria different from those implicitly ac cepted by Tylor as constituting the essential difference between magic and science. For if science is not naively equated with truth, but is regarded as merely a set of pro visional hypotheses, it follows that the distinguishing criterion of agreement or non-agreement with empirical reality is inadequate. What criteria, then, could be used in place of this? Presumably the classifications offered by the native observers themselves, although Frazer does not suggest this, and does not even seem to have been aware of the problem. The criterion Frazer actually proposed to distinguish magic from science was, in fact, identical to that assumed by Tylor. Magic assumes the existence of non-existent re-5 lations; science describes actual relations. It is ... a truism, almost a tauto logy, to say that all magic is neces sarily false and barren; for were it ever to become true and fruitful, it would no longer be magic but science. From the earliest times man has been engaged in a search for general rules whereby to turn the order of natural phenomena to his own advantage, and in the long search he has scraped to gether a great hoard of such maxims, some of them golden and some of them mere dross. The true or golden rules constitute the body of applied science which we call the arts; the false are magic.^ -257-Quite apart from the philosophical presuppositions in herent in this distinction, Frazer*s argument encounters an important dilemma: what to do with ideas, usually accepted as being in some way scientific, which are later repudiated by the scientific community. The phlogiston theory of com bustion is presumably "dross" in terms of Frazer's categori zations, as no doubt is Ptolemaic astronomy, Laraarckian evolutionary theory, and even Newtonian dynamics. Are we to term these ideas "magical" in consequence? Similarly, how are we to classify those actions, usually described as "magical", which are plainly efficacious, even if we choose to explain their efficacity in terms of a theory which differs from that accepted by the native observer? I am referring, of course, to such phenomena as shamanistic cures 7 and the deaths which sometimes result from ensorcellment. Frazer, in fact, made no attempt to discuss phenomena such as these, and had he done so', he would necessarily have had to reject the simplistic objectivist framework he accepted. C. Malinowski For Frazer, magic was essentially akin to science, pre ceding it in the course of human evolution, but governed by principles of reasoning at root identical. This equation was not explicitly challenged by Malinowski, who went so far as to repeat Frazer*s characterization of magic as being a pseudo-science. Implicit in the theory Malinowski developed, however, was a complete rejection of Frazer*s position. Malinowski began by asserting the existence in every primitive society of a clear dichotomy between the sacred and the profane, the latter dealing with the realm of every--258-day experience, and the former with those events and actions not completely subject to the control of rational human in telligence. Magic belongs to the realm of the sacred, and the closest equivalent to modern science is therefore the empirical knowledge which belongs to the realm of the pro fane. Magic, and this empirical knowledge, exist together of course. But they control different aspects of human be haviour, and differ in substance, form and function. More over, magic differs from empirical knowledge in being of a primarily emotive character. In discussing the empirical knowledge of primitives, Malinowski offered a number of criteria for deciding whether or not this knowledge could be considered genuinely scien tific. In doing so, he provided us with a means of under standing his conceptualization of the nature of science. These criteria were: i. A body of rules and conceptions, based on experience and derived from it by logical inference, embodied in material achievements and a fixed form of tradition, and carried on by some sort of social organization. By itself, however, this criterion is insufficient, since any art or craft could qualify as scientific by virtue of it. ii. A body of explicit rules, open to criticism by reason and control by experiment. In other words, not simply rules of practical behaviour, but also theoretical laws of knowl edge. iii. The scientific attitude, which consists in the dis interested search for knowledge and for the understanding of 8 the causes and reasons for phenomena. -259-On the basis of all these criteria. Malinowski argued that one could demonstrate the presence of science in primitive society. Merely to survive, a community must embody empirical experience in rules and conceptions, these rules and conceptions often taking the form of theoretical laws of knowledge. For example, The native shipwright knows not only practically of buoyancy, leverage, equilibrium, he has to obey these laws not only on water, but while making the canoe he must have the principles in his mind. He instructs his helpers in them. He gives them the traditional rules, and in a crude and simple manner, using his hands, pieces of wood, and a limited tech nical vocabulary, he explains some general laws of hydrodynamics and equilibrium. Science is not detached from the craft, that is certainly true, it is only a means to an end, it is crude, rudimentary, and in choate, but with all that it is the matrix from which the higher develop ments must have sprung.^ Moreover, even the "scientific attitude" - the disin terested search for knowledge - is not entirely absent from primitive society, according to Malinowski, although circum scribed by the traditional world of native culture. There exists in primitive society, ... the antiquarian mind passionately interested in myths, stories, details of customs, pedigrees, and ancient happenings ... the naturalist, patient and painstaking in his observations, capable of generalization and of con necting long chains of events in the -260-life of animals, and in the marine world or in the jungle.... the socio logist, the ideal informant, capable with marvellous accuracy and insight to give the raison d'etre, the func tion, and the organization of many a simpler institution in his tribe.10 Between this incipient science, and magic, Malinowski denied any fundamental similarity, since the former was based on observation and reason, the latter on specific emo tional states. Scientific theories are governed by the dic tates of logic; magic is based on the association of ideas under the influence of desire. Each is located in a dif ferent social setting and is associated with a different type of activity. Moreover, these differences are clearly recognized by the native observers. "The one constitutes the domain of the profane; the other, hedged round by ob servances, mysteries, and taboos, makes up half of the 11 domain of the sacred." The important question we must ask at this point, though, is whether or not Malinowski's criteria for dif ferentiating science are adequate. In fact they are not, since they fail to take account of an absolutely crucial distinction: that between science and technology. The tech nological success of primitive peoples in certain spheres of behaviour does not show that the ideology underlying this behaviour is necessarily sound, since technical success can often be based on far-fetched theories, or unintegrated rules of thumb. The example of Ptolemaic astronomy should make this clear. As Kuhn explains, the Ptolemaic system: ... was admirably successful in pre--261-dicting the changing positions of both stars and planets.... for the stars, Ptolemaic astronomy is still  widely used today as an engineering  approximation; for the planets, Ptolemy's predictions were as good as Copernicus'.12 The only criterion to be applied to technical knowledge is whether or not it works. Theoretical knowledge, on the other hand, demands to be evaluated in terms of its truth or 13 falsity, regardless of its utilitarian value. Primitive technology rests on a knowledge of scattered facts about natural processes. As Nadel remarks, primitive peoples: ... have nowhere reached the notion of "scientific laws" or any other concep tion meriting the name of science. At least, they have not done so in the  "phase" of profane behaviour."^ But what of the "phase" of the sacred? Malinowski denied that primitive magic had anything to do with under standing the universe, and set out to prove that primitive man is never misled, in his everyday life, by its spurious axioms and classifications. Magic constitutes an emotional response, not something carefully articulated and elaborated. In consequence of this position, Malinowski never undertook a discussion of magic and science as theories about the world, principles discovered in phenomena, or deductions made from them. Yet, as Nadel says, it is in fact quite possible that the earliest speculations about the universe occurred, not in the sphere of practical technology, but in 15 the realm of magic - as Prazer had suggested. Malinowski, however, categorically denied this. Por him, magic was -262-primarily an action-system, lacking any theoretical or specu lative side. But while the theory that magic is emotive and tends to arise spontaneously in situations of uncertainty may perhaps explain why people use it to further their own ends, and are convinced of its efficacy at the time of per forming it, it utterly fails to explain the persistence in society of a body of rules and prescriptions in which every body trusts. Once, however, we acknowledge that magic has a theoretical side, it is easy to come to grips with this. Por this knowledge must be of such a type that it convinces the dispassionate person by way of an objective appeal. "It must contain something in the nature of a theory, persuasive because of the principles and arguments on which it rests, 1 and akin, however remotely, to scientific theories." Magical conceptions may indeed, therefore, be the primitive analogue, "... not of applied science, nor perhaps of experimental science but of its theoretical and speculative side, however 17 crude and 'metaphysical'." It is worth noting that Malinowski himself was aware that the body of technical and practical activities existing in primitive society might not constitute a genuine equi valent to science. But he dismissed this question as being epistem©logical rather than anthropological, and did not 18 therefore examine its possible implications. Doubtless, this is because his crude and positivistic conception of science meant that the question was already settled for him. To the end of his days, he believed that the fundamental 19 truths in the sciences are never sophisticated. -263-D. E.E. Evans-Pritchard Evans-Pritchard attempted a far more systematic discus sion of the interrelation of magical and scientific thought than did any other of the authors considered in this thesis to this point. Yet many of his notions seem to stand close to those of Prazer and Malinowski, and the adequacy of his view of the nature of science is questionable. Evans-Pritchard derived his notions of the nature of 20 science from writers like Mach, Pearson and Poincare\ Both the first and the last of these writers exercised an 21 important influence on the Vienna Circle, and it is there fore not surprising to find in Evans-Pritchard, as in Prazer and Malinowski, an essentially positivistic conception of 22 science. This conception is not elaborated in his best-known writings, but is mainly to be found in the three articles published by him in the Bulletin of the Faculty of  Arts at the University of Egypt. In these articles, Evans-Pritchard presented a view of science as being essentially incremental, inductive, and descriptive of empirical reality. He also drew a firm distinction between the scientific and the logical, on the grounds that while the former must necessarily conform to the rules established by the latter, not every type of thought which is logical is necessarily scientific. We therefore have a differentiation of four types of thought: Scientific notions are those which accord with objective reality both with regard to the validity of their premises and to the inferences drawn from their propositions. Unscientific -264-notions are those which are invalid either in their premises or in the in ferences drawn from them. Logical no  tions are those in which according to the rules of thought inferences would be true were the premises true, the truth of the premises being irrelevant. Illogical notions are those in which the inferences would not be true even were the premises true, the truth of the premises again being irrelevant. Much of the confusion that has arisen by use of such terms as non-logical and pre-logical will be avoid ed by making a distinction between logical and scientific. In making pots all grit must be removed from the clay or the pots will break. A pot has broken during firing. This is probably due to grit. Let us examine the pot and see if this is the cause. That is logical and scientific thought. Sickness is due to witchcraft. A man is sick. Let us consult the oracles to discover who is the witch respon sible. That is logical and unscien tific thought.23 Similarly, the social content of our thought is scien tific, since it is in accordance with objective reality, whereas the social content of primitive thought is unscien tific, since it does not accord with objective reality, and may also be mystical, in so far as it assumes the existence 24 of supra-sensible forces. Later, in the earlier sections of his study of the Azande, Evans-Pritchard was to refine these ideas: MYSTICAL NOTIONS ... patterns of thought that attribute to phenomena supra-sensible qualities which, or -265-part of which, are not derived from observation or cannot be logically in ferred from it, and which they do not possess. COMMON-SENSE NOTIONS ... patterns of thought that attribute to phenomena only what men observe in them or what can logically be inferred from observation. So long as a notion does not assert something which has not been observed, it is not classed as mystical, even though it is mis taken on account of incomplete obser vation. It still differs from mys tical notions in which supra-sensible forces are always posited. SCIENTIFIC NOTIONS. Science has developed out of common-sense but is far more methodical and has better techniques of observa tion and reasoning. Common-sense uses experience and rules of thumb. Science uses experiment and rules of Logic. Common-sense observes only some links in a chain of causation. Science observes all, or many more of, the links.... Our body of scientific knowl edge and logic are the sole arbiters of what are mystical, common-sense, and scientific notions. Their judge ments are never absolute. RITUAL BE HAVIOUR. Any behaviour that is ac counted for by mystical notions. There is no objective nexus between the be haviour and the event it is intended to cause. Such behaviour is usually intelligible to us only when we know the mystical notions associated with it. EMPIRICAL BEHAVIOUR. Any be haviour that is accounted for by com mon-sense notions. Such behaviour is usually intelligible to us without ex planation if we see the whole of it and its effects.25 -266-In thus distinguishing between common-sense and scien tific notions, Evans-Pritchard undoubtedly went a step further than Malinowski, who failed to discuss the question of the relationship of technology to scientific theory. Yet how satisfactory is the idea of science here presented? Be fore proceding to answer this question, it is necessary to first complete this brief account of Evans-Pritchard1s con ceptualization of the nature of science by making one more quotation from his writings: All scientific theory is eclectic, for a scientist takes the hypotheses of his predecessors and examines them by logical tests and checks them by ob servation. By these means he selects what he finds to be valid in each hypothesis and works them into a co ordinated system. He adds his own observations and inferences and these in turn serve as hypotheses till they are verified by independent workers and are recognized as true by the con-Oft sensus of specialized opinion.*° Evans-Pritchard*s description of the nature of science shows that he conceptualizes its development as if it were governed by the results of empirical observation and logical inference alone. Of course it is true that these processes occupy a place of central importance in the evolution of scientific theories. It is equally true, however, that scientific development is not explicable in these terms alone. Evans-Pritchard therefore presents an overly rationalistic account of scientific development, which fails to take into consideration such important psychological pro cesses as "the flash of insight" which plays such an im--267-portant part in the formulation of novel scientific theories, not to mention the impingement on the scientific domain of epistemological and religious conceptions, and 27 even aesthetic preferences. Evans-Pritchard's characterization of the nature of science also seems to imply that he imagines there to he a full and complete, objective and true account of nature, and that science is scientific to the extent to which it ap proximates this ideal account. But if we are to concep tualize the universe as being infinite in nature, and the human mind as being finite, what justification do we have for assuming that we can ever fully comprehend this infinity, or that its comprehension is given in terms of one particu lar methodology - the scientific - alone? More than this, Evans-Pritchard*s approach tends to identify science with the facts, theories and methods collected in current scien tific texts. In such an identification, scientists become the men who have added one or other element to that particu lar collection, while the history of science becomes the gradual process whereby these elements have been added to the ever-growing stock-pile of truth. Yet what sort of treatment is one to extend to Aristotelian dynamics, earth-centred astronomy, or phlogiston chemistry on the basis of such a view? One may deny that such theories are scientific (for, surely, they do not correspond to objective reality). Alternatively, one may accept such theories as being scien tific, but only if one recognizes that science encompasses bodies of belief quite different from those held today. In that case, one must attempt to devise some criterion other -268-than that of agreement or non-agreement with objective reality in order to differentiate the view that the placing of a stone in the fork of a tree delays the setting of the sun from, say, the view that the earth rotates on its axis every twenty-four hours. Perhaps two final criticisms might be made of Evans-Pritchard* s definitions of scientific and of common-sense notions. Firstly, Evans-Pritchard incorrectly treats em pirical observation as an invariable, not subject to change 28 from one cultural context to another. Secondly, by de scribing scientific notions as corresponding to objective reality, he over-looks the fact that there is almost never a perfect ''fit" between empirical evidence and elaborated scientific theories. Thus even Einstein's Theory of Relativity, the cornerstone of the most exact of the natural sciences, can be shown to rest on somewhat uncertain em pirical evidence. For example, the Theory of Relativity has still to be reconciled with the results of the Michelson-Morley experiments of 1887 (since confirmed by later investi gators), which seem to show that the speed of light is not independent of the motion of the observer, contrary to 29 Einstein's theory. The correspondence of scientific ideas to objective reality, is therefore always partial. In Popper's words, The old scientific ideal of ... abso lutely certain, demonstrable knowledge - has proven to be an idol. The demand for scientific objectivity makes it inevitable that every scientific statement must remain tentative for  ever. It may indeed be corroborated, -269-but every corroboration is relative to other statements which, again, are tentative. Only in our subjective ex periences of conviction, in our sub jective faith, can we be "absolutely certain". E. Witchcraft and Science as Paradigmatic Activities Probably one of the most important books to be publish ed in recent years on the subject of scientific methodology 31 is that of Thomas S. Kuhn. Repudiating the positivistic image of science, Kuhn argues that the most characteristic attribute of scientific activity is the fact that it is governed by paradigms. Paradigms are "... universally re cognized scientific achievements that for a time provide model problems and solutions to a community of practition-32 ers." They define the problems available for scientific scrutiny, and the standards by which the scientific pro fession determines what may, and what may not, count as an admissible problem or a legitimate problem-solution. In this way, paradigms define the field of research for suc ceeding generations of practitioners, and give rise to par ticular traditions of coherent scientific research. This is possible for two reasons: i. The paradigm constitutes an achievement sufficiently im pressive to attract an enduring group of adherents away from a competing mode of scientific enquiry. ii. The paradigm is sufficiently open-ended to leave all sorts of problems for the re-defined group of practitioners to solve. Examples of paradigms are ready at hand: Ptolemaic and -270-Copernican astronomy, Aristotelian, Newtonian and Einsteinian dynamics, Darwinian evolutionary theory: perhaps we could add Classical, Marxian and Keynesian economics. According to Kuhn's view, normal science develops in the following manner: i. A set of phenomena attract attention, but no single para digm emerges to govern and direct the work of the community of scientific practitioners. The same ground is covered many times and new investigators, starting from scratch in the field, find themselves at no disadvantage. Examples of pre-paradigmatic scientific activity are the study of optics prior to Newton, and the study of electricity in the period preceding the formulation of Franklin's theories. ii. A single paradigm succeeds in becoming accepted through out the field, and thus comes to direct subsequent research. iii. This inaugurates a period of "normal science", during which scientists tend to unquestioningly accept the paradigm as true and limit themselves to developing and checking its implications. iv. The expectations provided by the paradigm are not, how ever, always fulfilled. Anomalies and contradictions occur, and problems are therefore posed for the paradigm which are attacked with the object of encompassing them within the terms of its explanatory powers. Further study of these anomalies may lead to a reinterpretation of them, or to a secondary elaboration of the paradigm. v. Frequently, however, a set of anomalies is particularly difficult to deal with in terms of a paradigm, and a period of crisis ensues in which repeated unsuccessful attempts are -271-made to modify the paradigm to accomodate them. Eventually, another paradigm is proposed which succeeds in winning over a number of practitioners, and a "scientific revolution" takes place. It is important to note that a new paradigm may be accepted, not because of any demonstrably superior explanatory power to the old, but simply because a number of practitioners welcome the opportunity for another attempt at explanation. vi. A new period of normal science then ensues, and the textbooks are rewritten in order to convince the neophyte of the truth of the paradigm he is being initiated into. Kuhn's statement of the nature of scientific activity might be regarded as being, in its field, a comparable work to that of Evans-Pritchard in his. Instead of focusing upon the content of various forms of belief, Kuhn begins the process of analysing sets of relations. The implications of this should be apparent for us, and the question naturally arises of the relationship of the sets of relations analysed by Kuhn to those analysed by Evans-Pritchard (and forgetting the equivocal distinction between "scientific" and "mystic"). The important problem then becomes, not the content of the relevant beliefs, but the manner in which they are structur ed. This being so, we are led to seek the differences be tween, say, Einsteinian physics and Zande witch beliefs, on a different level from that of truth and falsity. In fact, many similarities can be drawn between the witch beliefs of the Azande, and scientific paradigms. As S.B. Barnes says, the behaviour of the Azande in consulting their oracles seems to fall within "... the definition of a -272-problem-solving activity conducted within a set of rules given by social consensus but applied in an open ended way."^ Recognition of this fact leads Barnes to extend Kuhn's notion of paradigm to cover: "The notion of a set of categories, theories and procedures learnt in connexion with concrete examples, accepted by the entire reference group and applied to deal with problems in concrete situa tions.. . ".^* But where, given this extension of the notion of para digm, can the specificity of science be said to lie? Funda mental to the notion of paradigm is the idea that it is a set of concepts and procedures which are used to govern be haviour. Surely, then, the specificity of science must lie in the nature of the behaviour which its paradigms govern. It is strange that anthropologists have not appreciated this, and that instead of examining the contexts within which scientific thought operates they have placed their emphasis on the content of that thought. What difference is there, then, between the behaviour governed by scientific paradigms and that governed by social paradigms (Barnes1 term for such paradigms as the witchcraft beliefs of the Azande)? Barnes' suggestion is that, ... a scientific paradigm governs ac tivities of an esoteric and restricted nature and activities which have no bearing on the general pattern of the scientists' social life. Social para digms on the other hand tend to be ex tremely pervasive and to structure ac tivities which it would be greatly disadvantageous to alter, or even ac tivities which the individual is in--273-capable of altering. Thus for the actor the social paradigm governs more action and more significant action than the scientific one. Abandoning, say, the molecular orbital theory of chemistry means a lot less than aban doning the notion of responsibility or, for example, abandoning belief in poison oracles if you are an Azande.35 The specificity of science is therefore to be found in the fact that it governs only a restricted set of activities, and does not govern the way of life of the scientist con sidered as a whole. Science is therefore linked to the growth of a social structure permitting a more specialized role differentiation than is to be found in societies such as the Zande. Western science is therefore a product of permissive and differentiated social structural features, and the difference between the mode of thought of the indi vidual scientist and the primitive believer in magico-religious ideas is far less significant than has generally been appreciated by those operating within the context of traditional social anthropology up until the present time. -274-Notes and References 1. Jarvie, I.C. and Agassi, J., "The Problem of the Rationality of Magic", British Journal of Sociology, Vol. XVIII, (1967), P. 68. 2. Prazer, J.G., The Golden Bough, abridged ed., (Macmillan & Co. Ltd., London, 1960), Vol. II, P. 932. 3. See Popper, E.R., The Logic of Scientific Discovery, (Harper & Row, New York, 1965). For a critique of Popper^s views, see Ayer, A.J., "Editor's Introduction" to Ayer, A.J. (ed.), Logical Positivism, (The Free Press, Glencoe, 1959), P. 14; also Kuhn, T.S., The  Structure of Scientific Revolutions, (University of Chicago Press, Chicago, 1962), Pp. 145-6. 4. Frazer, J.G., The Golden Bough, op. cit., Vol. I, P. 67. 5. The further difference between magic and science in Frazer's view, should be noted. This is that the as sumption that nature is governed by impersonal and im mutable laws remains implicit in magic but is rendered explicit in science. Ibid.. It should also be pointed out that an important school of modern physics (Bohr, Heisenberg, von Neumann, etc.) no longer accepts the notion of strict physical causality which Frazer re garded as basic to science. According to these scien tists, the notion of strict causality in physics must be replaced by that of statistical probability, because subatomic events are indeterminate and unpredictable. Koestler, A., The Act of Creation, (Pan Books Ltd., London, 1966), P. 242. 6. Frazer, J.G., The Golden Bough, op. cit., Vol. I, P. 65. 7. See Cannon, W.B., "•Voodoo* Death", American Anthropo  logist, Vol. XLIV, (April-June 1942), Pp. 169-81. 8. Malinowski, B., "Magic, Science and Religion", in Malinowski, B., Magic, Science and Religion and Other -275-Essays, (Doubleday & Co. Inc., Garden City, New York, 1954), Pp. 34-5. 9. Ibid.. 10. Ibid., P. 35. 11. Ibid., P. .87. 12. Kuhn, T.S., The Structure..., op. cit., P. 68. Emphasis mine. 13. Jarvie, I.C., The Revolution in Anthropology, (Routledge & Kegan Paul, London, 1967), P. 96. I follow Jarvie in regarding the demand for true explanations as being unsatisfiable, but as nevertheless constituting an "indispensable regulative idea." Ibid., P. 16. 14. Nadel, S.F., "Malinowski on Magic and Religion", in Pirth, R. (ed.), Man and Culture, (Routledge & Kegan Paul, London, 1957), P. 198. Emphasis mine. 15. Ibid.. 16. Ibid., P. 197. 17. Ibid., P. 198. 18. Malinowski, B., "Magic...", op. cit., P. 26. 19. Nadel, S.F., "Malinowski on Magic...", op. cit., P. 208. 20. Evans-Pritchard, E.E., "Science and Sentiment: An Exposition and Criticism of the Writings of Pareto", Bulletin of the Faculty of Arts, University of Egypt, Vol. Ill, Part II, (Dec. 1935), P. 188. 21. Ayer, A.J., "Editor's Introduction", op. cit., P. 4. 22. It is significant, in this respect, that Winch considers there to be a close similarity between Evans-Pritchard's philosophical position and that of the early Wittgenstein - perhaps the most important single in--276-fluence on the Vienna Circle. Winch, P., "Understanding a Primitive Society", American Philosophical Quarterly, Vol. I, No. 4, (Oct. 1964),. P. 313. 23. Evans-Pritchard, E.E., "Science and Sentiment...", op.  cit., P. 188. 24. Evans-Pritchard, E.E., "Levy-Bruhl's Theory of Primitive Mentality", Bulletin of the Faculty of Arts, University of Egypt, Vol. II, Part I, (May 1934), P. 22. 25. Evans-Pritehard, E.E., Witchcraft, Oracles and Magic  among the Azande, (Oxford University Press, London, 1937), P. 12. 26. Evans-Pritchard, E.E., "The English (Intellectualist) Interpretation of Magic", Bulletin of the Faculty of  Arts, University of Egypt, Vol. I, Part II, (Dec. 1933), P. 149. 27. Barber indicates that scientists may reject novel theories and discoveries beeause of the professional status of their originator or discoverer, or because these theories and discoveries are inconsistent with current philosophical and religious views. Barber, B., "Resistance by Scientists to Scientific Discovery", Science, Vol. CXXXIV, (Sept. 1961), Pp. 596-602. It would be wrong, however, to view such influences in negative terms only. Thus, according to Koestler, Franklin's formulation of the law of the conservation of matter represents a transposition to the scientific domain of the Pythagorean doctrine of the transmigra tion of souls. Koestler, A., The Act of Creation, op.  cit.. Pp. 691-2. Similarly, Kepler's astronomical theory was inspired by the analogy he saw between the Holy Trinity and the sun, stars and space. Ibid., Pp. 125-7. On the role of aesthetic considerations in science, see Ibid.. Pp. 246-7. 28. On this point, see note 20, Pp. 125-7 of this thesis. 29. Koestler, A., The Act of Creation, op. cit.. P. 245. -277-30. Popper, E.R., The Logic.... op. eit., P. 280. Quoted in Koestler, A., The Act of Creation, op. cit., P. 247. Emphasis in original. 31. Kuhn, T.S., The Structure.... op. cit.. 32. Ibid., P. x. 33. Barnes, S.B., "Paradigms - Scientific and Social", Man (n.s.), Vol. IV, No. 1, (1968), P. 97. 34. Ibid.. 35. Ibid., P. 100. BIBLIOGRAPHY Ayer, A.J. (ed.). Logical Positivism. (The Free Press, Glencoe, 1959). Barber, B., "Resistance by Scientists to Scientific Discovery", Science. Vol. CXXXIV, (1961), Pp. 596-602. Barnes, S.B., "Paradigms - Scientific and Social", Man (n.s.), Vol. IV, No. 1, (1968), Pp. 94-102. Bell, D.R., "The Idea of a Social Science", Aristotelian  Society Supplement, Vol. XLI, (1967), Pp. 95-114. Bunzel, R.L., "Introduction" to Levy-Bruhl, L., How Natives Think, (Washington Square Press Inc., New York, 1966). Burrow, J.W., Evolution and Society, (Cambridge University Press, London, 1966). Cannon, W.B., "•Voodoo1 Death", American Anthropologist, Vol. XLIV, (April-June 1942), Pp. 168-81. Cantoni, R., II pensiero dei primitivi. (II Saggiatore, Milan, 1963). Cardan, P., The Pate of Marxism, (Solidarity, Clydeside, n.d.). Cardan, P., History and Revolution, (Solidarity, London, 1971). Childe, V.G., Social Evolution, (Watts & Co., London, 1951). Crawford, J.R., Witchcraft and Sorcery in Rhodesia, (Oxford University Press, London, 1967). Douglas, M., Natural Symbols: Explorations in Cosmology, (Barrie & Rocklife, London, 1970). -279 Douglas, Iff., Purity and Danger; An Analysis of Concepts of  Pollution and Taboo. (Penguin Books Ltd., Harmondsworth, 1970). Douglas, M. (ed.), Witchcraft: Confessions and Accusations. (Tavistock Publications Ltd., London, 1970). Durkheim, E., "Examen critique des systernes classiques sur les origines de la pensee religieuse", Revue philo- sophique de la France et de 1*Stranger. Vol. LXVII, (Jan.-Feb. 1909), Pp, 1-28, 142-62. Epstein, A.L. (ed.), The Craft of Social Anthropology, (Tavistock Publications Ltd., London, 1967). Evans-Pritchard, E.E., Essays in Social Anthropology, (Faber & Faber, London, 1962). Evans-Pritchard, E.E., "The Intellectualist (English) Interpretation of magic", Bulletin of the Faculty of  Arts, University of Egypt, Vol. I, Part II, (Dec. 1933), Pp. 149-82. Evans-Pritchard, E.E., "Levy-Bruhl's Theory of Primitive Mentality", Bulletin of the Faculty of Arts. University of Egypt, Vol. II, Part I, (May 1934), Pp. 1-36. Evans-Pritchard, E.E., "The Morphology and Function of Magic: A Comparative Study of Trobriand and Zande Ritual and Spells", American Anthropologist, Vol. XXXI, (1929), Pp. 619-41. Evans-Pritchard, E.E., "Science and Sentiment: An Exposition and Criticism of the Writings of Pareto", Bulletin of the  Faculty of Arts, University of Egypt, Vol. Ill, Part II, (Dec. 1935), Pp. 163-92. Evans-Pritchard, E.E., Social Anthropology. (Cohen & West Ltd., London, 1951). Evans-Pritchard, E.E,, "Sorcery and Native Opinion", Africa, Vol. IV, No. 1, (1931), Pp. 23-8. -280-Evans-Pritehard, E.E., Theories of Primitive Religion. (Oxford University Press, London, 1965). Evans-Pritchard, E.E., "Witchcraft (mangu) Amongst the Azande", Sudan Notes and Records. Vol. XII, (1929), Pp. 163-249. Evans-Pritchard, E.E., Witchcraft. Oracles and Magic among  the Azande. (Oxford University Press, London, 1937). Firth, R., Human Types. (Mentor, New York, 1958). Firth, R. (ed.), Man and Culturet An Evaluation of the Work  of Bronislaw Malinowski. (Routledge & Kegan Paul, London, 1957). Firth, R., "The Sociology of 'Magic'", Sociologus. (n.s.), Vol. IV, (1954), Pp. 97-116. Firth, R., Tikopia Ritual and Belief. (Allen & Unwin, London, 1967). Fortes, M. and Dieterlen, G. (eds.), African Systems of  Thought. (Oxford University Press, London, 1965). Fortune, R.F., Sorcerers of Dobu, (Routledge & Kegan Paul, London, 1963). Frazer, J.G., The Golden Bough, abridged ed., 2 Vols., (Macmillan & Co. Ltd., London, 1960). Frazer, J.G., Psyche's Task, (Macmillan & Co. Ltd., London, 1913). Giglioli, P.P. (ed.), Language and Social Context, (Penguin Books Ltd., Harmondsworth, 1972). Gluokman, M., Custom and Conflict in Africa, (Basil Blackwell, Oxford, 1956). Gluokman, M., "The Logic of African Science and Witchcraft", Human Problems in British Central Africa, Vol. I, (1944), Pp. 61-71. J -281-Goldmann, L., The Human Sciences and Philosophy. (Jonathan Cape, London, 1969)* Goods, W.J., Religion Among the Primitives, (The Free Press, Glencoe, 1951). Gough, K., "New Proposals for Anthropologists", Current  Anthropology, Vol. IX, (Dec. 1968), Pp. 403-7. Haddon, A.C., History of Anthropology. (Watts & Co., London, 1949). Harris, M., The Rise of Anthropological Theory, (Thomas Y. Crowell, New York, 1968). Hogbin, H.I., "Sorcery and Administration", Oceania, Vol. VI, (1935), Pp. 1-32. Homans, G.C., "Anxiety and Ritual: The Theories of Malinowski and Radcliffe-Brown", American Anthropologist. Vol. XLIII, (April-June 1941), Pp. 164-72. Horton, R., "African Traditional Thought and Western Science", Africa. Vol. XXXVII, Nos. 1-2, (Jan.-April 1967), Pp. 50-71, 155-87. Hymes, D., "Toward Ethnographies of Communication: The Analysis of Communicative Events", in Giglioli, P.P. (ed.), Language and Social Context. (Penguin Books Ltd., Harmondsworth, 1972). Jakobson, R., Essais de linguistique generale, (Les Editions de Minuit, Paris, 1963). Jarvie, I.C., "Academic Fashions and Grandfather Killing -In Defence of Frazer", Encounter. Vol. XXVI, (April 1966), Pp. 53-5. Jarvie, I.C., The Revolution in Anthropology, (Routledge & Kegan Paul, London, 1967). -282-Jarvie, I.C. and Agassi, J., "The Problem of the Rationality of Magic", British Journal of Sociology. Vol. XVIII, (1967), Pp. 55-74. Kardiner, A. and Preble, E., They Studied Man, (Mentor, New York, 1963). Kluckhohn, C, "Myths and Rituals - A General Theory", The  Harvard Theological Review, Vol. XXXV, (Jan. 1942), Pp. 45-79. Kluckhohn, C, Navaho Witchcraft, (Beacon Press, Boston, 1967). Koestler, A., The Act of Creation, (Pan Books Ltd., London, 1964). Kuhn, T.S., The Structure of Scientific Revolutions, (University of Chicago Press, Chicago, 1962). Leach, E.R., "The Epistemological Background to Malinowski's Empiricism", in Pirth, R. (ed.), Man and Culture: An  Evaluation of the Work of Bronislaw Malinowski, (Routledge & Kegan Paul, London, 1957). Leach, E.R., "Prazer and Malinowski", Encounter, Vol. XXV, (Nov. 1965), Pp. 24-36. Leach, E.R., Genesis as Myth and Other Essays, (Jonathan Cape, London, 1969). Leach, E.R., Rethinking Anthropology, (The Athlone Press, London, 1961). Levi-Strauss, C, The Savage Mind, (Weidenfeld & Nicholson, London, 1966). Levi-Strauss, C. Structural Anthropology, (Basic Books, New York, 1963). Levi-Strauss, C, Totemism. (Merlin Press, London, 1962). -283-Levy-Bruhl, L., Les carnets de Lucien Levy-Bruhl. (Presses Universitaires de Prance, Paris, 1949). Levy-Bruhl, L., How Natives Think. (Washington Square Press Inc., New York, 1966). Lowie, R.H., The History of Ethnological Theory. (Holt, Rinehart & Winston, New York, 1937). Lowie, R.H., Primitive Religion. (Boni & Liveright, New York, 1924). Macfarlane, A., Witchcraft in Tudor and Stuart England. (Routledge & Kegan Paul, London, 1970). Maclntyre, A., "The Idea of a Social Science", Aristotelian  Society Supplement. Vol. XLI, (1967), Pp. 95-114. Mair, L., Witchcraft. (McGraw-Hill Book Co., New York, 1969). Malinowski, B., "Anthropology", Encyclopaedia Britannics. 1st. Supplementary Volume, (London, 1926). Malinowski, B., Argonauts of the Western Pacific. (E.P. Dutton & Co. Inc., New York, 1961). Malinowski, B., Coral Gardens and their Magic, 2 Vols., (Allen & Unwin, London, 1966). Malinowski, B., Crime and Custom in Savage Society, (Routledge & Kegan Paul, London, 1926). Malinowski, B., Magic, Science and Religion and Other  Essays, (Doubleday & Co. Inc., Garden City, New York, 1954). Marwick, M., "Some Problems in the Sociology of Sorcery and Witchcraft", in Fortes, M. and Dieterlen, G. (eds.), African Systems of Thought, (Oxford University Press, London, 1965). Marwick, M., "The Study of Witchcraft", in Epstein, A.L. (ed.), The Craft of Social Anthropology, (Tavistock -284-Fublications Ltd., London, 1967). Marwick, M. (ed.), Witchcraft and Sorcery. (Penguin Books Ltd., Harmondsworth, 1970). Marwick, M., "Witchcraft as a Social Strain-Gauge", Australian Journal of Science. Vol. XXVI, (1964), Pp. 263-8. Merton, fi.K., Social Theory and Social Structure, (The Free Press, Glencoe, 1949). Middleton, J. (ed.), Magic, Witchcraft and Curing. (The Natural History Press, Garden City, New York, 1967). Middleton, J. and Winter, E.H. (eds.), Witchcraft and Sorcery in East Africa. (Routledge & Kegan Paul, London, 1963). Miller, A., The Crucible. (Bantam Books, New York, 1959). Nadel, S.P., "Malinowski on Magic and Religion", in Firth, R. (ed.), Man and Culture: An Evaluation of the Work of  Bronislaw Malinowski. (Routledge & Kegan Paul, London, 1957). Nadel, S.P., "Witchcraft in Pour African Societies: An Essay in Comparison", American Anthropologist. Vol. LIV, (1952), Pp. 18-29. Pareto, V., The Mind and Society. (Harcourt Brace, New York, 1935). Parsons, T. and Vogt, E.E., "A Biographical Introduction" to Kluckhohn, C, Navaho Witchcraft, (Beacon Press, Boston, 1967). Pocock, D.F., Social Anthropology. (Sheed & Ward, London, 1961). Polanyi, M., Personal Knowledge: Towards a Post-Critical  Philosophy, (Routledge & Kegan Paul, London, 1958). -285-Polanyl, M., The Tacit Dimension. (Doubleday & Co. Inc., Garden City, New York, 1966). Popper, K.R., The Logic of Scientific Discovery. (Harper & Row, New York, 1965). Popper, K.R., The Open Society and its Enemies, 2 Vols., (Routledge & Kegan Paul, London, 1962). Popper, K.R., The Poverty of Historicism. (Routledge & Kegan Paul, London, 1961). Radcliffe-Brown, A.R., Structure and Function in Primitive  Society. (Cohen & West Ltd., London, 1952). Redfield, R., "Introduction" to Malinowski, B., Magic,  Science and Religion and Other Essays. (Doubleday & Co. Inc., Garden City, New York, 1954). Rivers, W.H.R., Medicine. Magic and Religion. (Kegan Paul, Trench, Trubner & Co., London, 1927). Roszak, T., The Making of a Counter Culture. (Doubleday & Co. Inc., Garden City, New York, 1969). Swanson, G.E., The Birth of the Gods, (University of Michigan Press, Ann Arbor, 1964). Turner, V.W., "Witchcraft and Sorcery: Taxonomy versus Dynamics", Africa, Vol. XXXIV, No. 4, (1964), Pp. 314-25. Tylor, E.B., Anthropology: An Introduction to the Study of  Man and Civilization, (D. Appleton & Co., New York, 1898). Tylor, E.B., Primitive Culture, 2 Vols., (John Murray, London, 1920). Tylor, E.B., Researches Into The Early History of Mankind  and the Development of Civilization, (John Murray, London, 1870). -286-Wilson, M.H., "Witch Beliefs and Social Structure", American Journal of Sociology, Vol. LVI, No. 4, (Jan. Winch, P., The Idea of a Social Science and its Relation to  Philosophy, (Routledge & Kegan Paul, London, 1958). Winch, P., "Understanding a Primitive Society", American  Philosophical Quarterly. Vol. I, No. 4, (Oct. 1964), Pp. 307-24. Wittgenstein, L., Philosophical Investigations, (Basil Blackwell, Oxford, 1953). 


Citation Scheme:


Usage Statistics

Country Views Downloads
United States 90 13
United Kingdom 54 2
Canada 27 2
India 22 6
China 10 20
France 8 1
Germany 7 1
Pakistan 6 0
Morocco 6 0
Switzerland 5 0
Australia 4 0
Philippines 3 2
Jamaica 2 0
City Views Downloads
Unknown 69 15
Mountain View 13 3
Brighton 12 0
New Delhi 11 0
Vancouver 9 2
Highland Park 8 0
Shenzhen 7 20
Cambridge 7 0
Geneva 7 0
Durham 6 0
Islamabad 5 0
Waterville 5 0
San Mateo 5 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items