The Articulation Of Witchcraft & Modes Of Production Among The Nupe

Articulation Of Nupe Witchcraft

Nupe people , Nupe witchcraft , Nupe Army , Nupe guards


The political economy of occult belief in Africa can highlight hidden social and political conflict in times of transition which remain otherwise undetected.

This has been demonstrated in taking the development of witchcraft accusations over time as indicator, and the Nupe of Northern Nigeria as an example. A tentative long-term study on the growth of the Nupe state since pre-colonial times points towards a close relationship between the content and form of witchcraft accusations and the mode of production under which the stakeholders used to life and work.

Over time, witchcraft accusations among the Nupe apparently served different, even antagonistic ends, depending on the mode of production in which they were embedded.

Much confusion in literature on the apparent contradiction between ‘emancipating’ and ‘oppressive’ functions of witchcraft beliefs could be avoided by considering this articulation between modes of production, witchcraft accusations, and the underlying vested interests of the ruling powers.


Since pre-colonial times the Nupe have been renowned throughout Northern Nigeria for their knowledge of witchcraft.

This is so widely accepted among the general public that even newspapers occasionally report on it.

The virulence of occult belief among the Nupe even had serious repercussions on regional party politics.

One outstanding example was the notable shift in political power in Bida Emirate ( Niger state ), which became apparent during the gubernatorial elections in 1983 and the aftermath of violent conflicts that the local population saw as related to witchcraft 3.

Nupe witchcraft belief has already been subject to a rigorous analysis in S.F. Nadel’s pioneering studies on comparative witchcraft in Africa (1952; 1954).

Though Nadel did his fieldwork among the Nupe about seventy years ago, his subject is by no means outdated; witchcraft accusations still flourish in Nupe land, and they are an important sign of social stress and strain in Sub-Saharan Africa in general (cf. Ellis/Ter Haar 2004; Geschiere 1997; Kohnert 1996; 2003).

The Nupe are no exception in this respect.

Most Nupe informants interviewed in 1976, 1982 and 1990 insisted that the incidence of witchcraft has even increased since the early days of colonial rule 4.

In the peasants’ view, this growth of witchcraft is mainly related to the increasing ineffectiveness of traditional means of witchcraft control, an assertion which will be analysed in detail later; however, increasing social and political cleavages, caused by growing social and economic differentiation as a consequence of globalisation seem to lie at the root of this problem (cf. Comaroff/Comaroff 1999).

2 Thanks for valuable suggestions go to Lars Clausen, Georg and Karola Elwert, Max Marwick, Mike Mason, and two anonymous referees.

The responsibility for any fallacies or inaccuracies in the paper remains of course with the author.

The article is a thoroughly revised and updated version of my working paper, distributed in 1983 by the Sociology of Development Research Centre, Univ. of Bielefeld, titled: “Indicators of Social and Political Conflict in African Societies: On the Articulation of Witchcraft among the Nupe, Northern Nigeria”. 3 The surprising switch in the allegiance of the Nupe from the conservative National Party of Nigeria (NPN) to the Nigerian People’s Party (NPP) in 1983 resulted in terrible incidences of disorder, especially in Bida, were houses and vehicles were destroyed and a number of officials of the Federal Electoral Commission (FEDECO) were burnt to death.

Law and order had virtually broken down, with thugs hired by both sides and a mounting casualty list.

No wonder that in this tense political climate, witchcraft accusations flourished and were used by both parties to meet their ends.

Before the 1983 elections, the fortunes of the NPN in Niger state had already declined, as indicated by the election results, where the proportion of votes they received fell from 74.12% in 1979 to 64.72% in the gubernatorial elections, although these were probably rigged in favour of the NPN. In Bida, most NPP supporters had hoped to win, but they polled only 30.95% for the gubernatorial election in Niger state .
Disappointed Nupe as well as some minority groups in the state, for example, the Gbagyi people, protested against the Hausa – Fulani hegemony.

In addition, Sulaiman Takuma, Nupe candidate for the post of the national Secretary of the NPN was defeated by his rival Uba Ahmed from Bauchi state , last but not least, because of his unguarded remarks on the sensitive issue of zoning; although this was intended to conserve the power base of the northern Emirs, it was opposed by Takuma, who was campaigning in Niger state for the rotation of governorship between the Nupe , Gbagyi , Hausa – Fulani group of the incumbent governor and his own Nupe group. Cf. Olurode 1990:54, 64, 80; Barrett (1983:2866). 4 The author’s own field studies on Nupe witchcraft belief were conducted mainly within the framework of investigations in economic history and socio-economic differentiation of the Nupe peasantry in 1975/76 (cf. Kohnert 1982) and 1982. Semi-structured and narrative interviews were the major methods employed, complemented by research at the National Archives, Kaduna.

Further enquiries were conducted during an additional field visit in 1990.

consolidating the belief in magic and witchcraft. In African history, the occult reduction of complex social and political conflicts, as expressed by its personalisation in witchcraft accusations, was considered by the local population to be a legitimate option vis à vis external enemies, such as the colonial oppressors, and thus opened new doors for political action.

At least, this was an option pursued by (anti-) witchcraft movements directed against colonial domination, such as, for example, the Ijov-, Haakaa-, and Inyambuan movements of the Tiv in Northern Nige- ria or the Mpondo of the Transkei, South Africa, in the 1880s 8. More recent examples are the messianic grass-root movements, such as the Naprama of Mozambique, which used magic and witchcraft accusations to fight their enemies in “cults of counter-violence”, (cf. Wilson 1992); similar traits of occult thought and political action were to be found in Sierra Leone and Liberia in the 1990s (cf. Ellis/Ten Haar 2004).

Other witchcraft accusations with a strong “liberating”, “emancipative” or “egalitarian” impetus are directed against enemies within the respective community.

This is seen, for example, in the witchcraft accusations directed against rich peasants and traders in East- and West-Africa who accumulate large sums of grain or money as individuals without due regard to their obligation – under the traditional solidarity-system of the village community – to assist the poor in case of hardship. In such cases, the rich are suspected – quite correctly in the logic of a communitarian redistributionist social system – of obtaining their wealth through evil powers 9. We shall return to this question in more detail later.

Thus, witchcraft accusations made by social groups with vested political interests, and here in particular by groups which are organised in anti-witchcraft cults or secret societies, are often related to stress and anxieties resulting from the economic and social cleavages caused by the articulation of different or even antagonistic modes of production and the violation of the social laws which they represent. As will be demonstrated in the subsequent chapters, the communal mode of production in the village communities of peasant societies in Nupe land, for example, conflicted with the rules of the semi-feudal mode of production of the 19th century Nupe state , with the colonial mode of production installed in the early 1900s, as well as with modern rural.

8 Cf. Tseayo 1975:57-74; for further examples of witchcraft-control movements caused by stress and strain resulting from colonial conditions cf. Richards (1935), Marwick (1950), Goody (1957), and Lee (1976). – Early local resistance against colonial oppression in South Africa followed similar avenues, as shown by the Transkei rebellion of 1880 against the introduction of a hut tax register. This, it was believed locally, provided intimate knowledge of the taxpayers, which could be used by colonial administrators, who were believed to collude with witches and sorcerers, to inflict serious harm; cf. Redding (1996: 257): “People’s survival, as they saw it, may have depended upon the removal of a state that was prohibiting them from discovering and punishing all the ‘evilly disposed’ people.” Apparently, it was not the individual European, who was suspected of sorcery or witchcraft, but the colonial administration as such, reinforced by collaborating local chiefs with renowned occult power.

9 For similar examples cf. Salamone (1980) on ‘abujanke’ witchcraft among the Gbagyi ( Gwari / Gbari ), Northern Nigeria; cf. Tijani (1976:130) on witchcraft accusations connected with enrichment through development projects in Bornu Emirate, Northern Nigeria; and Brantley (1979) and Parkin (1968; 1970), on the “equalizing” effects of witchcraft accusations among the Giriama trading communities of Kenya. Comaroff/Comaroff (1999) analyse the accusation of migrant farm labourers in South Africa as zombies, employed by capitalist farmers, blamed to be witches.

capitalist development, induced by the Nigerian oil-boom in the 1970s.10. It is against this background that the forms and changes in the meaning of Nupe witchcraft over time will be in- vestigated and tentative hypotheses on the articulation of witchcraft accusations and modes of production in the course of Nupe History developed. The aim of this study is threefold: firstly, to test whether the analysis of occult belief is a meaningful methodology with which to uncover the origins of past and current social conflicts. Secondly, to obtain better insight into covert social conflicts during a crucial stage in the development of rural capitalism in Nigeria, taking the Nupe society as an example. And thirdly, to give a re-evaluation of Nadel’s witchcraft theory, including a reconsideration of the history of the Ndako Gboya, a secret society intended to control witchcraft, and at the same time a major pillar of occult belief among the Nupe 11. Finally, I should like to test the preliminary hypothesis of an interdependency between the form and content of witchcraft belief and the modes of production in which the different actors involved were embedded. As with most social and economic concepts intended to explain long-term historical developments, this hypothesis can be of heuristic value only; a sound database for more rigid methodological testing is lacking; this is all the more so as the scanty data and analyses available on the origins of the Nupe state and society are clouded by myth and ignorance.

The Origins of Witchcraft Accusations among the Nupe : Nadel’s Witchcraft Theory Reconsidered

As stated above, witchcraft and magic beliefs may indicate the precise nature of the social conflicts of which they are symptoms (cf. Nadel 1952:264; Marwick 1964). This holds especially for conflicts which dominant forces in a given society are obliged to camouflage because the open discussion of these conflicts may question the whole social fabric on which their well-being depends. In such a case, witchcraft accusations tend to act as a kind of safety valve which canalises existing hostilities towards a few scapegoats, rather than towards those who would be

10 For a detailed discussion of the concept of “modes of production” and its relevance for the analysis of Nupeeconomic history cf. Kohnert (1982:62-242). – On the links between globalisation, the rise of neo-liberal capitalism and the political economy of occult belief systems in Africa and elsewhere cf. Comaroff/Comaroff 1999; Ellis / Ter Haar 2004; Kohnert 2003.

11 When we talk of “the Nupe” here, this is not intended to obscure the fact that the Nupe are an extremely pluralistic society (cf. Nadel 1942:12-26). Of course, this holds even more when we speak of “the African”. For obvious reasons it is difficult to say whether the research results are representative for all Nupe or not (cf. Nadel 1942:ix/x). This is all the more valid as we are dealing with an extremely sensitive cultural and social aspect of African life. It must, therefore, be understood that where I limit my description to single events or attitudes, these can only claim to be ‘typical’ for the groups or villages which I investigated – unless stated otherwise. Although witchcraft belief in Africa is expressed in many different forms (according to region, religion, and ethnic affiliation) I have explained elsewhere (cf. Kohnert 1996) that, nevertheless, there are some essential common grounds which allow general insights.

regarded by the stakeholders as the real enemies of society. There can be no doubt that a society which needs such a safety valve has been badly constructed (cf. Nadel 1954: 206/206). However, for research purposes the study of magic beliefs provides us with a unique instrument to uncover and analyse social conflicts which otherwise might pass through undetected. In fact, this was stressed by S.F. Nadel in a well-known article on comparative African witchcraft where he applied the Durkheimian method of concomitant variances in an analysis of the differences between Nupe and Gbagyi witchcraft beliefs (cf. Nadel 1952:264). However, Nadel’s study un- derwent severe criticism by Salamone (1980) because of the alleged use of poor ethnographical data on the Gbagyi ( Gwari / Gbari )

12. This criticism certainly does not apply to Nadel’s well-founded Nupe studies, which contain a wealth of valid information. But a secondary analysis of Nadel’s data – complemented by results of my own fieldwork among the Nupe – supports a different interpretation of these data and subsequently a revision of Nadel’s witchcraft theory. It is not Nadel’s empirical base, but his witchcraft theory itself which will be challenged in the following. In light of general theories of functional anthropology and social psychology, fashionable in the 1940’s (cf. Nadel 1942:vi, vii; 1954:163-206), Nadel considered witchcraft accusations as a kind of “social illness”, analogous to psycho-pathological symptoms of mentally disturbed persons (cf. Nadel 1952:264). This “illness” is based, according to Nadel, on sex-antagonism, resulting from the specific marriage system and the sexual and economic independence of women from their husbands in Nupe society (cf. Nadel 1952:266-69; 1954:172-80, for a detailed account). Nadel holds that the symptoms are “easy to read; for in Nupe witchcraft expresses the social threats and tensions from which they spring with little disguise” (Nadel 1954:172). As proof he lists three points of reference:

    (a) The existing legends on the development of witchcraft .
    (b) The concrete accusations against witches .
    (c) The explicit reference to the market integration of Nupe women with respect to witchcraft- control .

However, an alternative interpretation of these symptoms, quite unlike Nadel’s straightforward interpretation, might be required. This will be elucidated in analysing his arguments one by one. The first reason given by Nadel, i.e. the analysis of existing legends on the origin of witchcraft, is of course highly speculative, but it may be accepted for the sake of argument against the background of the methodological concepts of social psychology, current at his time. As far as Nupe concepts of the historical origins of witchcraft, as well as countermeasures to defeat it, are concerned, Nadel (1954:172/73) refers to two founding legends of the anti-witchcraft cult of the Nupe , the Ndako Gboya13, and takes these to be decisive indicators of the social psychology of Nupe society. One of these legends centres around people living under the reign of Etsu Shago (the founding father of the Nupe state), leading a lawless life and refusing to listen to their elders.

“Men would steal each other’s wives and commit adultery without shame. The older women, especially, caused much trouble; they quarrelled among themselves and ‘gave no peace’. The more law-abiding among the men grew angry and spoke harshly to the women, but these ‘replied with insolence’.” (Nadel, 1954:173).

    “Men would steal each other’s wives and commit adultery without shame. The older women, especially, caused much trouble; they quarrelled among themselves and ‘gave no peace’. The more law-abiding among the men grew angry and spoke harshly to the women, but these ‘replied with insolence’.” (Nadel, 1954:173).

In trying to solve this problem, a young man of great strength invented a mask (later called Ndako Gboya ) to frighten away the insolent women. But one woman refused to run away; she was caught and killed with an iron rod (sányì, a slave chain, cf. below). This legend was still widely known among the heads of the Ndako Gboya lodges to the north and south of the Niger whom I interviewed in 1982. However, the legend, used by Nadel to back his thesis of “sex-antagonism” as the principal source of witchcraft accusations among the Nupe, also lends itself to other interpretations of the origins of the Ndako Gboya . It hints at the police functions of the cult, which were aimed at maintaining the moral values of the village community, values which were disre- garded by both men and women, although the older women, in particular, were given the blame. As Nadel rightly observed, the “evilness” of the elderly women implicitly undermined the authority of the “Great Men”.

The reference to the “insolence” of women in the legend may go back to an event in a very early period of Nupe history, i.e. the transformation from a matrilineal to a patrilineal society. This transformation took place over a long period and started, according to Frobenius (1912.2.:274), during the reign of the Bini dynasty, probably during the 14th and 15th century, i.e. at about the same time as in the Hausa kingdoms further to the North 14. The transformation probably coincided with the transformation from communal rule in acephalous village societies to the semi-feudal mode of production under the despotic rule of the kings of the emerging Nupe state. Remains of matrilineal rule, especially at the village level and in the Nuperank and land tenure system, were still evident in colonial times 15. This reinterpretation of Nadel’s hypothesis seems to be backed by the second founding le- gend, rendered by Nadel, which tells of a Nupe king whose mother was an interfering woman .

13 Literally, “Grandfather Gboya”, in Nupe, i.e. a traditional “secret society”, specialised in the control of witchcraft in Nupe (cf. Nadel 1954:188-201), but used for centuries, also for many other purposes of political oppression and extortion. According to Laing, Acting Resident at Bida in 1920, who investigated in the extortion cases of the ndakógbòyá, the latter means “big father of echo”, a meaning derived from the fact “that the gongola, supporting the long body of the mask, is pierced at the bottom, and the man bearing it speaks into this cavity, his voice then appears to come out of the top of this 15’ pole”; cf. NAK, SNP 10/8, 340p/1920. The Hausa name for the ndakógbòyá was masugirro or magiro, pl. magirai, according to Laing (ibid.).
14 Cf. Smith (1978:52-57, 123), on the transformation of matrilineal rule in Dauwa Emirate, the oldest of the Hausa kingdoms, which included the Nupe as one of its seven banza bakwài, i.e. “bastard” or vassal states.
15 Cf. Frobenius 1912.2.:274; 1924.9.:87-89; Temple 1922:329; Nadel 1942:x,31/32, 51, 54, 147-49, 278; Mason 1970:194-96.

constantly meddling in his affairs (cf. Nadel 1954:172). The king in question might have been the last of the Nupe Kings before the Fulani invasion (the Nupe / Fulani jihad ), Etsu ( Ali ) Kolo (Ta- )Nagari, who lived at the end of the 18th century. Etsu Kolo had been installed by his predecessor Etsu Maazu, although the former was the heir to the throne according to the then already outdated rules of matrilineal descent; therefore he was rejected by the Nupe elders, who had, in the meantime, become accustomed to the advantages of patrilineal organisation (cf. Frobenius 1912.2.:41). One of the early European travellers who visited the court of Etsu Masaba, wrote of Etsu Kolo : “He took his mother’s advice so constantly that the Nupe people said they wanted not to be governed by a woman, and, rising against him, drove him to Yauri ” (Baikie 1867:105).

The legend, however, tells a different story, i.e. the king himself succeeded in overwhelming his mo- ther (i.e. matrilineal rule, D.K.) with the help of the Ndako Gboya mask (cf. Nadel 1954:184, 164); details from different historical events have probably been merged, a familiar feature of myth- making. However, these are not the only founding-legends and, what is more, they do not necessarily back Nadel’s theory of sex-antagonism as a source of witchcraft accusations, neither in the colonial, nor in the present social and economic order of the Nupe. For reasons subject to speculation, Nadel and others completely ignored the valuable and detailed account of Nupewitchcraft by Leo Frobenius, published about twenty years before the publication of the relevant articles by Nadel 16. Interestingly enough, the two legends given by Frobenius did not fit into Na- del’s sex-antagonism theory. According to one of these legends, the Ndako Gboya represents the spirit of Etsu Guschi, one ancient local Nupe king in Epa (in the Northwest of Nupe land) who refused to come out of his katamba to “greet”, i.e. to bow to Edegi (in Nups also called “Tsoèdè”, the mystical founder of the Nupe kingdom), who had presumably invaded his country during the 14th and 15th century. Thereupon, the Guschi died (was killed?), and his son became the follower of the new emperor. The spirit which rose out of the grave of the late Guschi was – according to this legend – the incorporation of the spirit of resistance which created the Ndako Gboya 17. According to the second legend reported by Frobenius, the Ndako Gboya , in olden times .

16 Cf. Frobenius (1912.2; 1924.9). Even Murdock and the co-authors of the “Human Relations Area Files” did not consult the works of Frobenius as a source on Nupe witchcraft. Although Frobenius’ work on the Nupe has a heavy ideological bias which may even sometimes distort the facts, it is nevertheless interesting to read because it avoids another equally serious source of confusion, i.e. the “Fulani bias” in the “official” Nupe history, on which Nadel’s accounts are mainly based (cf. Nadel 1942:76; Kohnert 1982:90,380, fn. 212). Frobenius got his information mainly from peasants in the Mokwa area who have a long tradition of resistance against “Fulani domination”.

17 Cf. Frobenius (1912.2.:377/78; 1924.9.:78,81); unfortunately, the memory of this legend was lost to all of my present informants in the Zugurma/Mokwa area, including the Etzu Zugurma, one of the descendants of a branch of the pre-Fulani Nupe Kings, living in a small palace at Eba.

named Lata, came from the East via Atagara 18 into Nupe land. It was directed against “social en- emies” of the people, and at this time the mask had greater power than most of the Nupe Kings 19. Apparently, there is no reference to any kind of “sex-antagonism” in these legends. Accor- ding to the legends, the Ndako Gboya originally represented the resistance of Nupe peasant com- munities against enemies of their society, especially against the usurpation of despotic power by outside forces. Concerning the latter aspect of these legends, there are interesting parallels with the gunnu ritual of the Nupe – the most important of all traditional Nupe cults according to Nadel – where the Ndako Gboya used to play a dominant role as dògiri nyá gunnu, literally, “policeman of the gunnu“. According to Nadel, the gunnu priest, assisted by the Ndako Gboya mask, which is also called gunnuko , literally, “great gunnu “, holds all power in his hand during the week-long ceremonies: “Neither parents nor chiefs, not even the Etsu of Nupe , retain any authority …

The legal system (introduced and enforced by the ‘feudal’ overlords of the Nupe, D.K.) is upset rather than confirmed by the gunnu organisation” (Nadel 1937:100, 113). The gunnu guards the community against intruders and periodically obliges the whole village to honour the bond of its common values (cf. Nadel 1937:120). The evidence for the new interpretation of the historical origins of witchcraft and of the Ndako Gboya (as reconstructed from the legends) given above may not be very strong. But one has to bear in mind the fact that Nadel himself admitted that the legends, although they mentioned gender, failed to corroborate his concept of “sex-antagonism” in the narrow sense, since the legends did not confirm the empirical base of his view on the weakening effects of sexual intercourse on men’s strength as a principal cause of witchcraft accusations among the Nupe. On the contrary, the first of his legends talks about the sexual dominance of the male (through adultery), and frustrations of the (male) elders which concentrate on the old women, rather than the young and attractive ones.

The same is valid for the second point of reference for Nadel’s “sex-antagonism” theory, the actual accusations against witches. Firstly, they focus on the woman’s character as a whole, which rejects the submissiveness expected of women (cf. Nadel 1954:187), and especially on the dominating character of older women, rather than on their sexual dominance in particular. Why, then, did Nadel insist on his “sex-antagonism” theses, although he himself mentioned several times the important political role played by the Ndako Gboya ? Might it not be possible that – at least at the beginning – the Ndako Gboya ritual was not a “social illness”, but a legitimate political

18 I.e. Idah, the capital of the Atta of Gara (Igara or Igala), who became the master of Edigi, when the latter was brought to Igalaland as a slave.

19 Cf. Frobenius (1924.9.: 78-81). This account by Frobenius fits well into another legend, which was related to me in 1976 by the elders of Mokwa, namely that the ndakógbòyá had already been created before there was a king in Nupe land. It originated in Tata – near Onitsha. The man who invented the cult was not a Nupe, but he was the first to protect the Nupe by means of his kuti (magic) against warriors from that area.

instrument, developed by the peasantry to fight injustice and exploitation by their overlords? An instrument which only later, during the emergence of the state in Nupe society, was taken out of the hands of the peasants and occupied by the despotic central powers whose representatives turned it against the peasantry 20 ? Secondly, if the witch in Nupe religion is correctly identified, as Nadel presumes, as a person who openly and successfully sets aside fundamental values of society, does this not suggest that the most ostensible neglect of these values by the despotic rulers would be somehow connected with such witch accusations? This is all the more probable since the Ndako Gboya was most viru- lent in Trans-Kaduna, a region where the peasants were notoriously rebellious. Such a suggestion would also be backed by far reaching analogies, within the realm of the social psychology of Nupe peasants, between the characteristics of a witch (gici, pl. gicizi in Nupe for witches of both sexes) and those of despotic rulers, namely: (a) Both have all-embracing power over life and death (b) This (evil) power is conscious, and ordinary persons are powerless against it (cf. Nadel 1954:166; Frobenius 1912.2.:262).

(c) Both the Etsu and the witch have power over the profane and in the spiritual world (cf. Frobenius 1912.2.:261-67); personal investigations have shown that this view is widespread even nowadays (cf. below). (d) The most powerful witches are rich, they accumulate their wealth for their own selfish purposes. If they wanted money (in pre-colonial times) they would have sold their victims to another country as slaves according to the emic view (cf. Nadel 1954:165, 167). (e) Witches attack in the dark of the night – just as the Nupe slave raiders in pre-colonial times did. They suck the blood out of their victims (cf. Frobenius 1912.2.:41, 261-71; 1924.9.:68), which, in a metaphorical sense, may be said of the ruling class, too. Thirdly, if witchcraft accusations are mainly based on “sex-antagonism”, how then can the important role of male witches (eshe, pl. eshezhi , in Nupe) be explained, be it as “partners in crime” (i.e. as medicine men, cf. Nadel 1954:169), or even as male evil witches, who, Frobenius asserted prior to and in contrast to Nadel’s observations, did exist in Nupe land (1924.9.:65, 67)? In addition, the connection between magic and worldly power is not restricted to analogies which may appear more or less accidental or superficial, but is very real, manifested in the person of the Lelu. The latter is said to be the head of all witches in a village, and, at the same time, as Sagi or Nako , respected as the most powerful and rich woman in the village , held to be a “good” witch, who is expected to keep witchcraft within reasonable bounds (cf. Nadel 1954:168).

20 For a description of the ruthless exploitation of the peasantry by the institutionalised and closed lodges of the Ndako Gboya under the leadership of the Maji Dodo (lit. the “Master of the Terrible”) and the Etsu see Nadel (1935:442; 1954:197). Apparently, it reached its peak in the second decade of colonial rule, when the British Resident became afraid that no tax money would be left for the British, and therefore forbade the cult in 1921. For the exploitative aspect of the cult see also: Frobenius (1912.2:268; 1924.9.:78-80), Crowther/Taylor (1859:215), and Temple (1922:331/32).

However, we have to admit that with the emergence of the Nupe state, even the Lelu had to bow before the superior authority of the male village chief and district head. Since the establishment of Fulani dominance over the Nupe peasantry, the Etsu Nupe is alleged to have been even more powerful. He is said to control witchcraft himself, within a wide circle around his residence (cf. Nadel 1935:441; 1942:87). The alleged magic power of the Etsu served, at least until the end of the 19th century, as a well defined political means to maintain the despotic feudal order and the “king’s law”. This was done for the most part by invoking the kútí (i.e. magic) of the ègba Tsoèdè, i.e. the slave-chains of Tsoede (or Edegi), used for strangling political opponents and for uncovering crimes in the manner of an ordeal 21.

The fear among the peasants that the Emir might employ black magic to achieve his goals did not vanish after this period: Whitaker (1970:296), for example, reported that even during the electoral campaigns of 1959 local opposition to the Emir of Bida proved to be extremely difficult, among other reasons, because of the prevailing belief among the villagers, that the Emir would pronounce a curse, should they not vote for him or his candidate. Some twenty years later, in 1976, similar fears were still virulent among the peasantry with respect to Local Government elections, as I was able to observe myself. As the “open ballot system” was still in practice at that time, the above- mentioned threat proved to be very effective indeed. Altogether, major factors which constitute a witch in the eyes of Nupe peasants, i.e. illegi- timate spiritual and worldly power over life and death, and the disregard of fundamental human rights concerning health and property of their victims, seem to be very similar to the qualities of a traditional ruler in the view of the peasantry, at least in pre-colonial and colonial times. All this leads to my working hypothesis that the belief in witchcraft in Nupe land, especially the origin of the institutionalised anti-witchcraft cult ( Ndako Gboya ), was causally related to the resistance of the peasants within a communal mode of production against illegitimate power accumulation by the despotic rulers of the emerging Nupe state. At least, this thesis corresponds neatly to the second legend on the origin of the Ndako Gboya as rendered by Frobenius (cf. above).

Initially, the actual witchcraft accusations may have been directed either directly against the usurper, or they may have been the result of repressed frustrations, derived from the peasants’ impotence vis à vis illegitimate force, which was then directed against a few scapegoats. Of course, this hypothesis would have to be confirmed by sound empirical evidence. Although this hypothesis runs contrary to Nadel’s witchcraft theory, Nadel was certainly right in his general description of the effect of witchcraft accusations when he said: “Attacks against witches are thus attacks upon the successful enemies of the ideal society … witchcraft fears and accusations only accentuate concrete hostilities and in fact give them free reign” (Nadel 1952:279).

21 Cf. Nadel (1935a). Parts of this iron chain, which is similar to the old Portuguese slave chains (cf. Nadel 1954:32, 194), had been still a cult object of the Ndako Gboya in 1982, at least in its lodges south of the Niger.

Even though the Ndako Gboya was forbidden by the British Resident in Bida in 1921 for the reasons stated above, it continued to be practised, albeit on a smaller scale. In any case, the proscription applied only to Bida Emirate, and not to the Nupe districts south of the Niger . There the Ndako Gboya remained active, though it seems to have changed its character, and the cult is now even referred to as a “festival” by a tourist guide (cf. Emielu 1981:27). Whether it is really completely void of any exploitative aspect nowadays, and whether new forms of witchcraft con- trol have been developed will be explored in the following chapter. 3. Witchcraft accusation in present-day Nupe land As said at the beginning, the Nupe are still renowned (and sometimes feared) today among other ethnic groups of Nigeria, such as the Hausa or the Yorùbá, for their knowledge of powerful magic and the prevalence of witchcraft in their society 22. Nupe informants whom I interviewed during visits in several Nupe villages and towns in 1976, 1982 and 1990 insisted that witchcraft was by no means a minor aspect of Nupe cultural life, but, on the contrary, had increased very much during the past decades; however, the sensitivity of the subject makes it very difficult to confirm this by sound empirical investigation 23.

Although belief in black magic is still ingrained in both members of the ruling class in Bida, the capital of Nupeland, and the peasants in the countryside, this does not mean that there have been no changes since Nadel made his investigations among the Nupe. The most noticeable transformation probably did take place in the realm of control of the supposed witchcraft activities. The Ndako Gboya, in the 1920s one of the most powerful means in the hands of the Etsu 22 Some authors maintain that every ethnic group believes its neighbours to have more dark and potent magic than its own (cf. Parrinder 1958:196). But this certainly does not hold for the Nupe, as their “authorities” in anti-witchcraft matters are proud that they harbour the most powerful anti-witchcraft magic, a secret which is anxiously protected against disclosure to members of other tribes. In fact, the cult has even been ‘exported’ to other Nigerian provinces, e.g. to Yorùbáland. Named ” Igunnu” by the Yorùbá, the ndakógbòyá mask came to Abeokuta, Lagos and other Yorùbá towns through Nupe who migrated from the Gbado area during World War I, according to the present Lile (i.e. Village Head) of Gbado. (cf. also Parrinder (1953:58, 69) on ndakógbòyá masks in Ibadan).

23 Some of the reluctance to discuss the whole issue of witchcraft with outsiders was apparently due to the fear that other tribes could come into possession of the secrets of the Ndako Gboya as a result of the carelessness of talkative members of the cult. There is, however, also a more general suspicion among the Nupe concerning the discussion of witchcraft, as already observed by Nadel (cf. Nadel 1954:164): Since only witches can really know about their evil deeds, nobody would like to betray too much knowledge lest he or she be accused of being a witch him/herself. This is especially true of women – those most likely to be accused. It is difficult for males to interview these women, especially in an Islam dominated society like the Nupe. Thus, I hardly received any information from them, even though women such as the Nakó (lit. “grandmother”) of Dabba, a well-known herbalist and head of the women of this village in Trans – Kaduna were said to know ex officio about this craft. Needless to say, this imposes serious limitations on the value of the following analysis, which can only scratch the surface of a strong belief which is still deeply rooted in Nupe society .

of detecting witches and thereby of exploiting the peasantry, is now also used for entertainment purposes, as a masquerade, a “cultural performance”, which features on occasions of social enjoy- ment, devoid of its original social and religious meaning. Nowadays, the great masks of Kusogi 24, for example, are invited to perform their dance at Sallah, at agricultural shows of the Bida Agricultural Development Project (established in the early 1980s), or to “greet” important state guests who visit the Emir in Bida. Such events, like the performance on the occasion of the recent visit of the Oba (king) of Benin, who toured the northern Emirates in October 1982 to weld all traditional rulers of Nigeria together in a united front against attacks on their “historic privileges” in the course of the upcoming 1983 elections (cf. above), must have been experienced as degrading, especially for the mask bearers 25. After they had been waiting in vain for about two hours for the arrival of the Oba and the Etsu , who were supposed to ride through the main streets of Bida on horseback as in olden times, the latter rushed through the cheering crowd within seconds in their extravagant Mercedes and Volvo cars, concealed by the darkened car win- dscreens, and accompanied by their entourage (including the private TV-team of the Oba), hardly taking any notice of the frustrated dancing masks and drummers who tried in vain to follow them. But it may be that those concerned held their mere presence as sufficient to chase away any evil- minded spirits or witches .

Minor masks, such as the Ndako Gboya of Lade to the south of the Niger , now even dance for entertainment on the occasion of important local football matches. But again, there could also be more serious, less overt reasons for the presence of the masks, such as the protection of the players against the witchcraft of their opponents. There are several examples in African football history where violent clashes resulted from witchcraft accusations in connection with football matches allegedly “rigged” by means of black magic 26. Nevertheless, the Májin Dòdo of Kusogi admitted during my interview that the cult and its members had lost much of their former authority. Many villages and lodges had ceased altogether to perform the ritual. They had to invite the Ndako Gboya from other villages, mainly from the Nupe areas south of the Niger state , e.g. from Gbado , Tankpufu, Tsambafu, Patizuru, Etsuvun or Lade, if they had serious cases of

24 The second Kusogi was founded near Doko under Etsu Masaba (1859-73), and since that time has been head- quarters of the Ndako Gboya lodges in Bida Emirate. The original Kusogi is situated east of Pategi. It finally lost its significance when its last Majin Dodo died in about 1970. The new title holder is a teacher of Arabic who is said not to be very interested in keeping up the tradition of the ndakógbòyá. 25 The performance in question took place on October 8, 1982, just in front of and in the courtyard of the guest house, or more precisely the palace, of one of the biggest and most important businessmen in Bida, thus demonstrating the emerging alliance between the old traditional rulers and the aspiring new class of the national bourgeoisie (on the latter cf. Kohnert 1978). 26 Cf. Richards 1997. – In 2002 the Ivorian Government settled a 10-year dispute with disgruntled witch doctors who claimed to have had a hand in the country’s African Nations Cup triumph (cf. BBC News, 08.04.2002;; 04.02.05). In Tanzania there were allegations that the national team used money earmarked for players to pay a witch doctor (cf. BBC News, 20.10.04;; 04.02.05); Witchcraft was also part and parcel of Ugandan football.

witchcraft with which they could not cope themselves. One of the cases in the Mokwa area, remembered very well by a Reverend of the UMCA, happened at Kpaki (about 18 miles from Mokwa on the road to Bida) in 1963: A culmination of strange events, such as the outbreak of smallpox, a significant increase in snake bites, and the beating of young men by invisible hands which allegedly resulted in an outbreak of yellow fever, set the village in turmoil.

The village authorities called the Ndako Gboya from Tsambufu to discipline the witches who were supposed to be the instigators of this sudden outbreak of evil. The reason most readily given by educated Nupe for the decreasing importance of the Ndako Gboya was the ever-growing influence of Islam in the already Muslim dominated Northern Emirates and the impact of modern education. Nowadays, fewer and fewer youngsters are prepared to undergo the harsh initiation rituals which involve floggingging, staying in the bush at night or the participation in strange customs such as exhuming of human bones in the graveyard at full moon. Another obvious reason for the decreasing incidence of witch cleansing rituals by the Ndako Gboya is directly related to the considerable costs involved in such performances. Most villages, such as Kusogi, Mwuo, and Gbado that could still afford these expenses, exercised the ritual only once a year, most often at the beginning of the dry season, in November or December; some villages, such as Gada, held their annual performance in September. Of course, the costs were heavy only in relation to the average annual peasant’s income. As the organised extortion tours of the past had apparently ceased since they were forbidden by law in 1921, and as there has also been a decreasing demand for such cleansing rituals over the past decades (due to reasons to which we shall return later), a considerable number of the lodges of the Ndako Gboya which were still in existence seemed to be at pains to cover the expenses for the ceremonies, dances, and the maintenance of the masks.

It is doubtful that under such circumstances they still contribute any considerable amount to the shadow household of the Bida emirate, in contrast to, for example, the reign of Etsu Bello (1916-26), when the lodges, besides gifts in kind, got an average of £ 20 (up to £ 100) per village in ransoms, a profit which they shared with the Etsu in Bida (cf. Nadel 1935:441; 1954:196). An investigation by Laing, the Acting Resident of Bida in 1920, revealed that the Ndako Gboya extor- tion concentrated on the districts to the West of the Kaduna river, notorious for their quest for independence 27. The range of gross family income from farming at that time was estimated at

27 According to inquiries of the colonial administration into the extortion tours of the Ndako Gboya in the early 1920s, the acting resident in Bida, Mr. Laing, came to the following conclusion: “the parties of the masugirro (ndakógbòyá, D.K.) who were found to have visited the Districts of Labozhi, Egbako, Sakpe and to a small extend Jima-Doko, the details are as follows: (37) Egbako Dist. Livestock and goats extorted £ 471.0.0. It was obvious that these depredations could not have been carried out without the knowledge of the DH and Alkalin Egbaki. (38) 11 maigirro arrested; (39) Labozhi Dist. Goods, livestock, money taken: £ 367.2.10.; (40) Skape Dist., Goods, livestock, money taken: £ 165. 1.2; (42) 13 magirro arrested at Kutegi; (43) From the foregoing it was obvious that extortions on such a large scale could not have been carried out without the knowledge of the Hakima (the DH, D.K.). After further investigation … I arrived at the conclusion that the Emir alone was the instigator and prime mover of the sending out of these masugirro parties; (45) … after a half-hearted denial the Emir acknowledged his guilt; (47) This catastrophe has been a severe blow to the N.A. and has done an incalculable amount of harm. The District Heads were witness of

between £ 2 and £ 12 per year, corresponding to the different resource positions of peasant families (cf. Kohnert, 1982:226-227). According to my own interviews among the Lílé of Gbado, as late as 1982 every villager – “even the child in the womb” – had to give the “traditional one shilling” for a performance , whe- reas the elders in Mokwa said that they themselves gave a total of about Naira 200 (2 Naira ~ 1 £ in 1982) in cash, apart from an unspecified amount paid by other members of the community. Apart from money, the villagers had to contribute in kind: a he-goat, food and drinks – even Indian hemp in one village to the south of the Niger . So the total value of the contributions for one performance of the Ndako Gboya in 1982 was estimated by the elders of Mwuo, a village in southern Mokwa District at about Naira 400 to 600 in 1982.

Thus, according to the author’s own rough estimates, the income of the cult per performance has decreased in real terms within the past six decades by about 20 to 50 per cent 28. Of course, the performance may have yielded additional income, if the Ndako Gboya discovered a witch. The accused, or her/his relatives, would have to give a ransom, like a black goat, cloth, and money, according to the seriousness of the supposed offence and the willingness to undo their spell or misdeed. But, as stated above, such incidences were few in the 1980s, compared with the past, and the deflated total annual net income from witchcraft eradication by the Ndako Gboyacertainly decreased much more than by mere 50 per cent.

The potential for systematic exploitation, based on the social structure of a semi-feudal kingdom, which did not exist any more, decreased even further. On the other hand, the cult might have guarded, consolidated or even extended its political influence into the realm of party politics, as indicated by the witchcraft accusations linked with the switch of party allegiance of the Nupe in 1983, mentioned at the beginning. In this respect, the Ndako Gboya might bear more resemblance nowadays to the Ogboni secret society of the Yoruba, or the Okija cult, i.e. the Ogwugwu-shrines in Okija, Anambra State. The gruesome spiritual and worldly powers of the latter, and its strong influence on high ranking politicians in Anambra government and

NupeProvince – Report, no 37, for half year ending 30th June 1920”) ” … (6) ‘smelling out’ witchcraft was a lucrative branch of these charlatans, who levied their tolls, at times on individuals or on the whole community.

Men and women have been tied up and beaten on the slightest pretext and only obtained their freedom on payment. It will be easily understood what scope this afforded to individuals having grudges against their fellow villagers; a word spoken to the ‘magirro’ accompanied by a douceur, would speedily occasion ‘a certain weakness’ in the household of the unfortunate accused.” (ibid.; I am grateful to Mike Mason who allowed me to quote from his copy of the relevant files). 28 The nominal growth rate of the income for one performance between 1920 and 1980 was about 4 to 4.6 per cent per annum, whereas the price per unit of staple food, like sorghum or rice, increased about 5 to 5.6 per cent p.a. during the same period (cf. Kohnert 1982:487).

29. However, again we enter the sphere of mere speculation, and further investigation would be required. Although the Ndako Gboya is still a “secret society” in the sense in which Nadel used this expression (cf. Nadel 1954:196), as the members of the lodges still guard certain important secrets on the rules and the magic of the cult, the very source of their power, and as they threaten any adept with serious punishment (even the death penalty) should he dare to betray their secrets, the cult can no longer be described as a “closed society”– with the exception of the strong gender bias which still exists.

Nupe Man , Nupe people , Ancient Nupe Weapon
According to the Heads of the lodges in Gbado and Tankpufu , i.e. the two major centres of the cult to the south of the Niger , any self-confident and reliable man who feels strong enough to stand the initiation rites and to keep the secret of the kútí (i.e. ritual or magic), may apply for membership. The high initiation fees of former times which were as high as £ 20 (or twice the customary bride price) (cf. Nadel 1954:194) have been reduced to a token payment of one cock and some 60 Kobos (1 Naira = 100 Kobos) intended for the initiation sacrifice. The attitude of the villagers towards the Ndako Gboya seems to have changed considerably, too. The terrifying influence which the dancing masks used to have on women, in particular, has almost gone, as far as I can judge from my limited experience of two performances at Tsambafu and Bida. This may be taken as another sign of the vanishing influence of the cult in general. And although one of my informants, a high ranking Nupe officer from Koro (east of Patigi), but resident in Ilorin, would not exclude that the Ndako Gboya still served the aim of upholding the authority of men over women and extracting resources from them, he also agreed that most villages which still harbour one of the lodges of the society perform their dances mainly to honour a cherished tradition.

It remains to be added that at present, as in the past, the purpose of the masquerade is not restricted to social entertainment or to fighting actual cases of witchcraft, but is still used to en- sure the well-being of the village at large, e.g. the cult pleads for rain at times of drought, helps women to become pregnant, or protects the villagers against illness (cf. Nadel 1954:195). When there was a change in the District Headship at Doko in 1976, even the compound of the D.H. (District Head or hakìmi, pl. hàkìmai) had to be cleansed by the Ndako Gboya to protect the new D.H. against any evil magic which might have been placed in his official residence during the interregnum. Although the new District Head, as usual belonging to the Majigi royal family in Bida, ridiculed this performance a little, due to the presence of a European researcher like myself,

29 Cf. Morton-Williams (1960) and Fadipe (1970) on the traditional Ogboni society; for current examples cf. The Sun, of 05. & 19.08.2004, as well as other Nigerian newspaper reports in the following month, on the Okija-affair.

he was nevertheless eager to point out that he is the major authority over all movements of the Ndako Gboya in his district. He claimed that even the Etsu Nupe had to consult him before he could send the mask to any village in his district. But, according to him, the villages preferred to handle all witchcraft cases which they were unable to settle themselves through mediation by the D.H.

Therefore, it is no wonder that during the D.H’s three years of tenure (1973-76) at Doko, the Ndako Gboya of Kusogi (which was under his jurisdiction) was not called at all for the purpose of detecting and punishing witches, although the D.H. himself had to deal with three witchcraft cases during this time, according to his own statement. The first took place in Ebagi, a hamlet south of Doko, where he was called at night because the inhabitants had accused a witch of poisoning the village well. The second and third cases occurred in Doko itself; the issues centred around marriage disputes (i.e. jealousy and adultery). The D.H. asserted that he was capable of telling at once – if necessary with the help of the Sagi – whether the accused was a witch or not. And, in the D.H.’s own words, “because the peasants still fear and respect my autho- rity, I normally succeed in convincing the evil-doers to loosen their spell on their victims, – otherwise, they are threatened with jail”. As, however, only Upper Area Courts, like that of Bida, are allowed to deal with witchcraft cases, simple Area Courts, such as those in Doko, Kutigi, or Mokwa, can make only preliminary investigations. The last of such cases in Doko occurred in 1975 and involved one local policeman who had an argument with a prostitute whom he “forgot” to pay. According to the D.H., she tried to impress the policeman with the following incident: Some days after the dispute, the policeman took a kettle as he wanted to clean his mouth and hands for prayer and suddenly heard a voice, coming from nowhere saying, “Don’t touch me, I am poisoned”.

Though he was surprised and startled he tried again, but the warning was repeated. The case was reported to the local Judge and investigated with the help of the Sagi . The detected “witch” confessed to being responsible for this incident and to wanting to teach her customer a lesson. This example may serve as an illustration that witches are supposed not just to kill randomly, but that as a rule there is a kind of “Dantesque appropriateness” between the offence of the victim and the punishment by the witch, as Salamone observed with respect to the neigh- bouring Gbagyi (cf. Salamone 1980:9). The range of such “punishments” is said to stretch from impotence to pauperisation and murder. This became evident in a case related to me by a Hausa assistant to the Irrigation Officer in Patigi: In 1976 the Irrigation Officer laid off some of his workers for reasons. Two weeks later he had a road accident under strange circumstances, escaping narrowly with his life. Some days later, my informant overheard a conversation between two of the sacked Nupeworkers in which one of them boasted of having bewitched the officer in causing the accident.

expressively to serve as a first warning, but that this “idiot” would not escape so easily the second time. It seems to be noteworthy that this time the self-confessed “witch” was a male, who, what is more, belonged to the Ndako Gboya society, according to his own words. This indicates the general ambivalence of many Nupe towards medicine-men or witch-finders, who they often equate – unconsciously or not – with witches themselves, as we shall see later in detail (cf. similar Salamone (1980:15) on Gbagyi medicine-men). How many of such or similar cases are brought to the courts annually, or whether the number has increased or decreased over the years is difficult to tell30But since most Nupe peasants are still bewildered by the ambivalent attitude of Western-educated judges towards the handling of witchcraft cases, and as they do not feel sure about the intentions of the existing law (cf. Nadel 1954:163/4), it is doubtful that significantly more witchcraft cases are handled by the courts today than in the 1930s.

As in the past, the courts are most likely to be involved only if witchcraft cases assume strong political overtones, or if serious offences, like murder, were committed. The following two case histories may serve as an illustration. The first case, which received nationwide attention, took place in spring 1973 in Mwuo, a remote village on the western border of Nupe land 31. As usual, there are different versions of the story. According to one version, the victim was a 28–year-old patent medicine seller, born in a Christian family and a staunch member of the UMCA (United Missionary Church of Africa) in a village whose population was almost equally made up of Moslems and Christians. The medicine seller, G., had a quarrel with his mother and other co-villagers, possibly because of his recent marriage with an 18-year-old girl, and was warned by his brother, a pastor at the Salaka Bible School, to leave his village over New Year. However, G. disregarded this warning and subsequently fell seriously ill.

Two days later he died. The villagers suspected that he died as a result of some “evil practices”, they buried him in the village cemetery after a funeral service con- ducted at the local church. Four months later, the dead and buried man was found alive again – though distressed and confused – in the bush near Ndafu, a hamlet some 18 miles from Mwuo. The story of this apparent rise from the dead spread rapidly, and the Local Administration – fearful of the religious and political disorder that might ensue from this story – instructed all religious leaders in Mwuo to tell their congregation that no human being could ever arise from the dead. The case of the “resurrected” man was later investigated by the police of Mokwa , a .

30 Both the former Chief Àlkali of Bida, who was among the members who drafted the Nigerian Penal Code (including section 214 to 219 on ordeal, juju, and witchcraft) and who was one of the chief members of the Emir’s Council at the time of the interview, and the Àlkali at the Bida Upper Area Court in 1976, were very reluctant to provide any useful information on this issue. 31 The incident received nationwide publicity, as one Nigerian magazine carried a photo-story on this witchcraft case; cf. “Mystery man of Mwuo”, in: “Spear – Nigeria’s National Magazine”, July 1973:15-18.

medical officer from Bida General Hospital, and even the Etsu Nupe himself. A skeleton of doubtful origin was dug up from G.’s grave, which some villagers had found unusually depressed shortly after the burial. It was suspected that the Christian community of Mwuo, or at least its re- presentatives, took part in a religiously motivated campaign to destabilise the traditional (Islamic) authorities. The former Village Head and others were tried in the courts for inciting religious unrest, first at Mokwa, then at Bida and Minna (the division and state capital, respectively). The accused Village Head was deposed, whereas the victim was apparently able to turn the newly acquired fame into money. At least, G. owned a Toyota van and ran a prosperous medicine sel- ling business in 1982 32.

The second case took place long ago, in 1953, but is still of interest because it throws light on the close relationship between Nupe and Gbagyi witchcraft beliefs which, quite contrary to Nadel’s theory (cf. Nadel 1952:267), the peasants, at least, believe to exist (cf. similar Salamone 1980:4). The incident took place at Yidna, a Gbagyi village in Paiko District (neighbouring Badeggi District in the former Bida Division). It centred around accusations of unlawful trials by ordeal administered by three Nupe medicine-men 33. At that time mass ordeals were also endemic in Nupe land. This was probably due to the fact that, in view of the official prohibition of the witch hunts of the Ndako Gboya , the people thought they had no alternative means to detect witches than to take recourse to individual witch finders. The people of Yidna feared that their village was troubled by an evil spirit, which among other things had caused illness and an attempted murder of the Nákorjí, the Nupe title of the Village Head of Yidna, when the latter was riding through the bush.

In November 1952, the Nakorjí sent for three travelling Nupe “witch doctors” (bocizi) 34 who had been seen practising in some neighbouring villages (Gawu and Buku). As the bocizi had never before been in that area, they first asked the Sarkin Paiko, the District Head, for permission to come to Yidna, which was granted after they had sent him a tribute of £ 5, as well as a calabash of cola nuts. When the Nupe bochizhi , two brothers and one of their friends from Bida, who earned their living by such extortion tours, arrived at the village, they told the population that the Sarkin Paiko had given them authority to dispense a liquid potion for everybody in all the hamlets of his district and that this would reveal any witch among the population. Anyone who should refuse to drink their “medicine” would be arrested. The next morning, the whole village assembled in front of the door of the Nákorjí’s compound, and everybody, including babies, had to drink from the potion, mixed from water and different .

32 However, according to another local source of information, G. was later accused of being a witch himself because two children died shortly after he treated them with an injection. 33 Cf. here and in the following: NAK, MINPROF. 2478, “Preliminary inquiry: Regina vs. Moh. Kusogi Bida and 5 others”, trial at the court of Minna, Jan. 16, 1953. 34 Medicine men or “witch doctors” are called boci (pl. bocizi) or cigbeci (pl. cigbecizi) in Nupe ; the suffix ci indicating the actor or owner of the medicine ( Chigbe , in Nupe ). The Gbagyi equivalent is beki.

powders by Moh. Kusogi, the leading boci. Some of the villagers fell down immediately after drinking the potion, while three of them – all women – died on the spot and were buried the same day. This caused some concern among the bocizi because, according to their own statement, all their victims had recovered from similar ordeals in the past, if treated with an antidote (in this case lemon juice). But since the villagers were convinced that the women who died were indeed witches, the travellers could nevertheless collect ransom money or fines from the victims and their relatives, respectively.

One unlucky man, whose mother and wife died, had to give £ 3, two lengths of cloth, and a chicken. Others gave a big gown, three lengths of cloth, and five chickens or 35 shillings and a chicken, etc. It is interesting to note that there was little opposition to the dealings of the bocizi, even among the relatives of the deceased victims. When one of them dared to protest, the elders replied: “Why has your mother died, others drank and are still alive?” The husband of another woman who died during the ordeal even agreed that “she died as a consequence of her evil ways” as revealed by the ordeal. And yet another villager confessed to one of the bocizi that he was an evil person and had attempted to kill the Nakorji by witchcraft. Eventually, in spite of the attempt by the villagers to cover up the incident, the colonial administrators heard about it, albeit by chance. At about the same time, another famous boci called Ndasogba, from Badeggi, used to tour the southern and eastern Districts of Nupe land on the invitation of villages who sent for him to cleanse their villages from witchcraft. He was known as the leading herbalist in the Cis-Kaduna area in the last two decades of colonial rule and was in the habit of organising public trials to prove the innocence of persons who were suspected of being witches. Reasons for such suspicion were described by the Nupe as follows: If a person is bewitched and falls ill, there is a chance that he (or she) will signify the name of the person who has bewitched him in his dreams. The dream figure will be only dimly recognizable through a column of smoke 35. However, the victim himself will not be able to remember either his dream or the name he uttered.

Therefore, the cooperation of the boci or the watcher at the sick-bed is required to note down the name of the accused – usually, a person with whom the victim is in close contact. The relatives of the bewit- ched or the boci himself will then visit the accused person and demand the release of his or her spell. Should the accused refuse or deny the accusation, he or she will be forced (normally moral coercion will suffice, as the whole story will be spread round the village) to prove innocence by

35 Cf. “Note on witchcraft in Badeggi District”, A.H.M. Kirk Greene, Badeggi, September 24, 1954; included in: NAK, Bida Div., Non-Current Papers, B118, 1933, “Witchcraft cases and anti-witchcraft in Nupe society”; cf. similar, Frobenius (1924.9:65). – The reference to smoke in this report probably indicates the use of a certain anti-witchcraft medicine, called toràri, which is put beside the sick person on a fire and is supposed to cause the bewitched person to utter the name of the evil spirit. According to the Etsu Zugurma, the toràri was one of the most powerful anti- witchcraft medicines in the 1980s, and had replaced the ndakógbòyá, at least in the Eban/Zugurma area of Nupe land.

ordeal. Two well-known Nupe potions were applied: ukpa and kpakangichi , made from the powdered bark of a poisonous tree 36. Similar kinds of witch finding in other parts of the country were already used in the twenties as a means to extort money, as a report by H.M. Irwin indicated. The report was included in the above-mentioned memo on witchcraft prepared for the Resident of the Niger state Province 37. Whether this report, together with the sad experience of the extortion tours of the Ndako Gboya described above, caused the British Resident to exert some moral pressure on the Etsu Nupe is not known.

But it is likely that the Colonial Officers persuaded the Etsu and his council to take a closer look at such mass ordeals. In the above- mentioned case of Ndasogba, at least, the Etsu told his District Heads that such public trials would spread only worries and fears among their people and would cause disunity among them. He ordered the hakimai (D.H.) to announce that any community inviting witch doctors, as well as those who would apply the ordeal, would be seriously punished. Apparently, this order was implemented immediately.

After Ndasogba was caught a third time, the Àlkalin Àlkaki of Bida fined him as well as the inhabitants of Tswakoko, the village which had invited him, £ 5 38. However, this reflects only the official handling of the subject. Similar to the prohibition of tribute payments (under the pretext of zakkat) – which the British tried in vain to enforce because it ran contrary to the interest of the ruling class (cf. Kohnert 1982:99-119) – the traditional autho- rities apparently were not very eager to stop the business of “witch-finders”. Even after the death of the Ndasogba, other travelling bochizhi ( Nupe Native Doctors ) continued to perform “ witch trials ” in Nupe land unmolested.

One well-known boci in western Trans-Kaduna, who continued to earn his living in 1982 by touring villages on invitation, was the Ndagena of Ebi. His junior brother, a boci himself at Mokwa, told me that both of them, assisted by other villagers, conducted a witch hunt in Mokwa just four weeks before my arrival. Altogether, four witches were detected and punished. They were all females, one old and three younger women, married, but from different families. None of them could be called rich, though each of them had her sideline occupation, e.g. petty trading in foodstuffs. All cases centred around conflicts within the extended family. In two of the cases, the

36 In the court case against Moh. Kusogi, mentioned above, the potion, analysed by a Government chemist, revealed a lethal dose of aworoso, which is sometimes used as a purgative in a smaller dose. 37 “Sometimes, however, this belief in witchcraft was only used as an excuse for raising money by the chiefs. When someone died, the relatives when bearing away the corpse would approach as if voluntary the door of a certain man, who would at once be accused of causing the death of the deceased.

The victim would always be a man of means which would be appropriated by the Chief. The latter would previously give his instructions to the deceased’s relatives and probably a present also for their pains… The natural effect of such a custom was to militate against the acquisition of wealth, as its possession was a source of temptation to the accuser. Even today no pagan will make a display of wealth…” H.M. Irwin, D.O., “Extract from Kumbashi Assessment Report, 1920”, NAK, MINPROF, M. 1228, 1934. 38 Cf. the correspondence of Etsu Nupe , Muh. Ndayako, on “Ndasogba Mai-Maganin Maita” and related subjects, Nov. 1952 to Feb. 1954; included in: NAK, Bida Div., Non-Current Papers, B 118.

Mother of the family head was accused of having bewitched her son and husband “to give his mother money”, i.e. to guarantee the social security of his mother. In another case, a young schoolboy fell seriously ill. He was allegedly bewitched by a barren older woman living in the same compound and said to be jealous of the fertility of her female co-residents. In both cases, the accused women “confessed” their evil deeds and loosened their spell, i.e. their “victims” reco- vered from their illness 39. Whether these confessions were made without coercion is doubtful, last but not least, because my informants in Mokwa agreed that if an accused person tried to argue, he or she would be forced to prove his or her innocence by an ordeal, applied either by the Ndako Gboya or a boci. In the Mokwa region, the accused was given a certain potion, called fien or tagan, which is meant to cause him to confess immediately if he or she is a witch. Innocent people, they say, do not feel anything if they drink the medicine 40.

As late as 1982, witches were also punished by flogging, and in more serious cases they were exposed naked in the streets of the village in broad daylight, beside payments in cash and/or kind to the boci or the Ndako Gboya . First of all, however, they were ordered to repair the harm they were accused of having caused. The failure or agreement to obey this order was a decisive element in determining the degree of the punishment. Although Nadel’s informants, at least, maintained that a witch was never killed, since that would only entail revenge by other witches (cf. Nadel 1954:188), this information was apparently falsified by the evidence of violent clashes connected with the “witchcraft riots” in Bida in 1931, when troops even had to be called out to quell the riots (cf. Nadel 1942:127; 1954:163). Also, one of my informants, a Nupe from Kata eregi , confessed that as a young boy he had taken part in the stoning of a witch in his village.

The witch in question was an old woman who confessed to having bewitched her co-wife out of jealousy. This case, however, took place in the late thirties and it seems unlikely that such an incidence of stoning would occur today, although one should not rule out the possibility, last but not least in view of the heated international controversy around the stoning verdicts of Sharia courts, related to cases of adultery, in other Northern Emirates in 2003 and 2004. Generally

39 According to the elders of Mokwa there are seven different kinds of illness caused by witchcraft in Nupe land: (1) shamu, i.e. stomach trouble which causes severe pain to the victim right from birth; (2) zana-zana which can be seen in the eyes; (3) lan-jwan-jwa which is witnessed through continuous headache; (4) esu, i.e. severe pains under the ribs; (5) wuregi, an illness which paralyses the whole body; (6) kparagi, i.e. convulsions which are lethal for children; (7) bogun, which will cause the body of the victim to swell up so that he loses his power. 40 Yet another ordeal, known as wasa, was applied in the Doko area.

This is a snake bite medicine, the belief being that a true witch would be killed or at least frightened and subdued by a snake after drinking the medicine (cf. also Nadel 1954:188). As the former D.H. of Doko told me, the wasa medicine was still applied in Jima/Doko District in 1982, also to detect thieves. The same is valid for the sòrògi dance, mentioned already by Nadel (1954:188), which was still applied as recently as 1982 in villages south of the Niger state , to detect and to punish witches, who are supposed to get a swollen throat from watching the dance, according to the Lile of Gbado. The ordeal where the suspect had to scratch the ground with his finger-nails until blood appeared from under his nails (cf. Frobenius 1924.9.:66; Nadel 1954:188), however, disappeared some generations ago, at least in the Mokwa region.

speaking, according to the elders of Mokwa, the most serious cases will nowadays be handed over to the police and the courts. The bocizi are often assisted in their search for evil magic by the Village Head and the Lelu of affected villages. The use of the word Lelu is ambiguous in Nupe . In pre-jihad times (i.e. before 1806) it denoted the office of the Head Woman of a village, and it still refers to the supposed spiritual qualities of this woman (cf. Nadel 1954:167/68).

But as many Nupe villagersused to call any woman who is suspected of being experienced in witchcraft Lelu, in a rather abusive sense 41, the existence of a Lelu in a village will often be denied by the inhabitants. Therefore, it is doubtful whether the assertion of the Majin Dodo of Kusogi that today the Lelu is found only in Patigi Emirate is valid. Other informants, natives of Doko, insisted that they still have a Lelu in this village, which is only a few kilometres from Kusogi. In Dabba, a medium- sized village in Trans-Kaduna, they even have four Lelu, one for each of the four village wards.

These women were so much respected by the Nkó, the Head Woman of Dabba and a great her- balist herself, that she had to consult them first before granting me an interview, out of fear that otherwise, these women would take her to task because of single-handed decisions on such an important matter as providing information to strangers. It is, therefore, evident that many Nupe villages still have their Lelu, but as witchcraft remains such a sensitive issue, people would rather avoid calling her by that title. However, everybody agreed that a Lelu is not sufficient to check evil magic at the village le- vel. Despite the relatively dense network of “witch-finders” described above, most interviewed Nupe authorities insisted that the number of witches in Nupe land must have increased over time. Three reasons were given to explain this development. The first two of these explanations seem to be quite obvious with regard to the inner logic of occult belief systems: Firstly, population growth caused a growing effective demand for witchcraft medicine.

The growth of population meant that other figures being equal, the absolute number of witches would double every 25 years 42. In addition, it was said that the incidence of witchcraft per village increased in relative terms as well. This was explained by the growing impact of the market economy and the aftermath of the Nigerian oil boom (both in terms of economic growth and the spread of corruption and social differentiation), which had its side effects in Nupe land, too; more money became available in the villages, leading to more opportunities and a greater effective demand for witchcraft medicine 43. The Nupe , as a rule, believe that witchcraft is not necessarily hereditary since there are various

41 Cf. memo of M. Aliyu on witchcraft and related subjects” NAK, Bida Div., Non-Current Papers, B118, 1938. 42 Population increase in Nigeria was estimated at about 2 to 2.5 per cent per annum in 1982 in the countryside and 3 per cent in the urban areas. This would imply that the population – and ceteris paribus with it, the absolute number of witches – doubled within one generation (or every 23 to 35 years). 43 The Nupe have various names for such medicines ( chigbe ): They call them ega (n), like witchcraft in general, eshe, like the name for male witches, or bàdufù, literally, “dark place” (cf. Nadel 1954:158/59).

methods of acquiring it (cf. Nadel 1954:165/66, 170). The most common method, they say, is to buy the cigbè, either from other witches or from a cigbèci or boci, the crucial implication being that the evil deed of the witch is intended. In view of the legal situation, no boci would ever admit openly to selling such witchcraft medicine, but nevertheless some offer “guestimates” on the costs involved, namely some £ 5 in the “olden days”, probably Naira 400 – 600 in 1982 44. Therefore, my informants reasoned, someone who is really poor could hardly be a witch. But, as they said, you never know, since a young girl could have been initiated into this sinister oc- cupation free of charge by her mother, who was already a witch.

The growth of the market economy also played a crucial part in the closely related reasoning of some of my informants, that with the oil boom and its socio-economic effects spreading well into the countryside, it became easier to accumulate capital. In fact, the impact could be easily seen in the growing farms of rich peasants and capitalist absentee farmers and the flourishing business of big traders all over Nupe land. But the growing social differentiation, even within the village (cf. Kohnert 1979; 1982), raised the envy of the poorer villagers, and some of them accused their wealthy neighbours of having enriched themselves by evil means. However, as we have already seen, this widely reported “natural effect” of witchcraft accusations to militate against the unequal acquisition of wealth, is nothing new. It was already noted by British colonial officers in the 1920’s (cf. the touring notes of H.M. Irwin, D.O., quoted above), and it is also common in other African societies 45. However, there is no empirical evidence available to back the working hypothesis that witchcraft accusation as a means of defending traditional egalitarian community structures against the intrusion of rural capitalism in Nupe land increased in importance over time, relative to other factors such as jealousy or greed.

The most worrying local explanation for the growing incidence of witchcraft accusations in Nupe land was given by “Papa Angulu”, a former Reverend of the Church Missionary Society (CMS) at Doko. His account was confirmed later by the similar reasoning of other informants and runs as follows: In view of the official ban of the Ndako Gboya and of witchcraft ordeals, many Nupe were afraid of being left unprotected against the evil intentions of the witches, now supposed to be spread all over the country without any effective control. In addition, the power and influence of the Ndako Gboya and the Lelu deteriorated due to other reasons listed above. Under these adverse circumstances, it might happen that a desperate family head or a mother would feel pressed to resort to witchcraft to protect the family against evil magic from outsiders. The argument behind this reasoning is that it is possible to drive out the devil with Beelzebub (similar to the concept of the Lelu and the Ndako Gboya , cf. Nadel 1954:191). However, this could

44 Frobenius (1924.9:68) mentioned that a “Boschi” would ask a price of 30,000 cowries in 1909, i.e. roughly the equivalent of the price of a slave in pre-colonial times. 45 Cf. Salamone (1980) on the Gbagyi ( Gwari / Gbari ); Parrinder (1953:54) on the Yoruba ; Brantley (1979:131/32) and Parkin (1979:223) on the Giriama of Kenya ; Comaroff/Comaroff 1999 on South Africa .

lead to the tragic situation that a mother, in her despair about an actual incidence of witchcraft, might see no alternative other than to commit a ritual murder, either to appease the witches, or as a precondition for acceptance into the witch order as the Nupe see it 46. Since, for a mother, the most easily available victims are likely to be young children (notably young girls) of her own extended family, especially the kin of her husband or co-wives, she was said to be faced with the tragic choice of either killing one of her own relatives in order to be able to protect the rest of the family against the evil intentions of co-witches, or of leaving the family unprotected 47.

This indicates that altruistic reasons for practising witchcraft also existed beside the usual selfish reasons given, such as greed or jealousy. Regarding the logic fabric of Nupe religion, this certainly had serious repercussions on the way the relatives and neighbours of the witch as well as the bocici, and maybe, even the Alkali courts evaluated the witchcraft committed. Furthermore, it is an additional explanation of why in the view of the _, the evilness of the witch is so often directed against his or her husband’s or co-wives’ kin. And as the woman is considered to be the traditional keeper of the family by the Nupe , this was given by my informants as a final explanation of why more women than men are accused of witchcraft 48.

Towards A New Conception Of Nupe Witchcraft

In summary, a consistent local concept of Nupe witchcraft, if existent, would have to distinguish between at least two different kinds of witchcraft: Firstly, the employment of black magic – e.g. the sacrifice of one’s own family members – for good ends, i.e. to protect the rest of the family49. And secondly, the use of black magic to harm anyone in order to satisfy unsocial, selfish desires,

46 Cf. Nadel 1954:165; a more prosaic, but no less tragic, situation would arise if a family head or a mother, in desperate need of protection against witchcraft, were requested by the consulted witch-doctor to commit ritual murder in order to get parts of the human body considered to be necessary for the preparation of an anti-witchcraft medicine. Cases of ritual murder are reported periodically in the newspapers, not only from Nupe land , but from all parts of Nigeria, one of the outstanding cases being the Okija-cult incident in Anambra state, mentioned before. The spread and commercialisation of so-called muti-murder, i.e. ritual murder for the purpose of enhancing the business success of greedy businessmen, has been widely commented in Southern Africa in recent years (e.g. in South Africa, Tanzania, Zimbabwe); it has already been subject to scholarly analyses (cf. Comaroff/Comaroff 1999). 47 Precisely this protection of the family against witchcraft might be the strange “special” and “unintelligible pleasure” which Frobenius (1924.9:67) narrated 80 years ago from the Nupe in Mokwa in the following, somewhat confused statement on the reasons and methods by which the male (!) would-be Nupe witch acquired his evil craft from the boci: “If a man has many children, he may very well fear for them, or else entertain the hope, by means of their help, namely in sacrificing some of them, to gain a special pleasure, which is, of course, unintelligible to us …

The man (asks the boci, D.K.): ‘Give me a good medicine which I can take home … and which I can use for me and for my children … Give me a good Tschibe (i.e. cigbe, D.K.) so that I can prove my power to everybody.'” The medicine which the family head eventually gets from the ‘Bassa-tschi’ or ‘Boschi’ has to be eaten by the witch himself ‘and all his children’. (Translation from German, D.K.). Probably, Frobenius confused and mixed the protective and the greedy reasons for acquiring witchcraft medicine, which, however, had significantly different repercussions in the logic of Nupe religion. 48 At Mokwa, my informants estimated the proportion of male to female witches to be one male (eshe) to 20 female (ega(n) or gaci, as witch is called in general in Nupe ) in a gang of witches in 1976. But as the female cannot kill without the help of the male (cf. Nadel 1954:169/70), the latter was considered to be more dangerous. However, as nowadays the males are said normally not to kill themselves, they are hardly ever detected and punished. 49 It goes without saying that even in this case the end does not justify the means, killing of family members or co- villagers, for whatever reason, is regarded by any Nupe as a capital crime.

such as greed, envy, jealousy, private redress, or other sinister reasons (killing for fun is also held to be possible, according to my informants). Apparently, there are – contrary to Nadel’s comparative witchcraft theory (cf. Nadel, 1952) – close similarities between these two types of Nupe witchcraft and the nwando and abujanke witchcraft of the neighbouring Gbagyi ( Gwari / Gbari ). My Nupe informants agreed to this, though they do not distinguish the two types by separate names. However, among the Gbagyi the nwando witchcraft is restricted to personal gains resulting from the witchcraft of males within the family, and abujanke is restricted to female witchcraft against outsiders or social enemies for reasons of revenge (cf. Salamone, 1980:6-8, 12). But it also seems necessary to point out some differences between these two indigenous witchcraft theories. Firstly, among the Nupe there is not such a strong correlation between the type of witchcraft and the sex of a witch, the victim, or the client, respectively, as among the Gbagyi50. Secondly, the Nupe , unlike the Gbagyi , do not differentiate between the “real evil-doer” (i.e. the client of the witch) and the witch, who performs the evil for him. Again, this separation between the witch and his or her client in Salamone’s witchcraft account seems to be more of a theoretical nature, as the Gbagyi hold that in practice the real evil-doer is the person who intends to harm other people with the help of either a witch or a witch-doctor (cf. Salamone, 1980:10). This is si- milar to the Nupe belief where the would-be-witch has to consult the bochi or another witch before her/his evil intentions can become effective.

The most sinister effect of witchcraft, the murder of one’s enemies or rivals, depends in each case anew on the co-operation of the boci (cf. Nadel 1954:169/70). On other occasions the Nupe witch may act directly, without the help of others, provided that he/she has been initiated into the witch order, but again dependent on the cooperation of the boci 51. Thirdly, whereas in Gbagyi society, according to Salamone, witchcraft itself acts as a means of social control, it seems that among the Nupe, witchcraft accusation, i.e. the anti-witchcraft activities, are meant to fulfil this function. This thesis may need some explanation: According to Salamone, the Gbagyi witches are “tragic” figures. They get their evil power already with their mothers’ milk, and their victims have been selected for them by their clients. These clients again seek revenge for some harm previously caused by the victim. So the witch’s victims are not innocent, on the contrary, certain social norms of Gbagyi society have been violated by every victim 52.

50 In Gbagyi witchcraft theory the client is the “real evil-doer” who consults the witch (or the witch-doctor) and selects the victim for him/her. However, it may be that Salamone overstated this point, as his case histories underline that this correlation between sex and type of witchcraft is a vague “general tendency” (cf. Salamone, ibid.). 51 However, the question remains open as to why – contrary to the Nupe – the Gbagyi should punish not the “real evil-doers” but their partners in crime (i.e. the consulted witch) as Salamone (1980:10/11) maintains.

This is even more puzzling as the Gbagyi say – contrary to the Nupe – that one becomes a witch not by intention but by birth, unconsciously and involuntarily (cf. Salamone, 1980:9). 52 Cf. Salamone (1980:5, 13). Those violations are “jealousy, hatred, refusing to share good fortune, withholding love, injustice, and resorting to magic to advance your fortune to the detriment of another. All can be reduced to the formulation that violators have not acted like true Gbagyi ” (ibid., p. 13). – This statement raises some doubts again: According to the case histories that Salamone provides as examples to clarify his point (ibid., p.9), at least jealousy and hatred were motives of the witches and their clients, not of their victims. In fact, the latter interpretation would appear more reasonable and is common not only among the Nupe but also in many other African societies. However,

As shown above, under certain conditions the Nupe are also said to employ black magic against victims who openly set aside the customary social norms of Nupe society, (or are at least accused of having done so), thus punishing the social deviant. This happens especially in cases of rich people who refuse to share their resources in times of hardship. Even in such cases, where witchcraft against outsiders is employed to serve good ends, the witches would be considered as evil persons because they seek to settle their grievances individually and by illegitimate force.

This egoistic, impudent use of individual power, however, is contrary to the social norms of Nupereligion. As in most African societies, the prerogative of punishing the enemies of the society is restricted to the community, the state, or its representatives. In this respect, there is no difference between the Nupe and the Gbagyi (cf. Salamone, 1980:15). For the Nupe , however, the important point is that the use of power is not considered as such to be evil ( as maintained by Salamone (1980:15) for the Gbagyi ).

On a theoretical level, the evil embodied in the illegitimate use of individual power for selfish purposes is a necessary, but not sufficient precondition of witchcraft within the realm of the occult belief system of the Nupe .
Other things being equal, anti-witchcraft cults are a means for a society to effect sanctions against egoistic private redress, which represents the illegitimate use of individual force; illegitimate, of course, only from the standpoint of the dominant ideology, which reflects more often than not the ideology of the rulers. However, in different socio- economic formations, this could have a completely different meaning which may be camouflaged by seemingly insignificant changes in the appearance of witch purges.

Under the conditions of the communal mode of production, e.g. among the stateless village communities of the Nupe zányi (i.e. the “pure” Nupe) in pre-Edegi times up to the 15th century (cf. Nadel, 1942:19/20, 25, 73), such socially organised witch accusations might have been used as a direct defence against the usurpation of private power by despotic invaders (such as Edegi, cf. the first legend quoted above), similar to witchcraft accusations in other parts of Nigeria, as among the Tiv of the “Middle Belt” (cf. Tseayo, 1975:57-65). Under the slave (or semi-feudal) mode of production of the 18th and 19th century they were used by the Nupe state, first as a method of securing and consolidating the appropriation of power from the gentile organisation 53.

Nupe Army , Nupe Jihad , Nupe invader , Ancient Nupe people
Afterwards, during a period stretching from the Nupe jihad (i.e. during the 19th century) well into the colonial mode of production (i.e. the first half of the 20th century), the witch hunts were used as Ndako Gboya extortion tours 54. This might have been welcomed by the ruling class in Bida as a

this interpretation would fail to back Salamone’s thesis that among the Gbagyi the witches act as agents of social control. 53 According to narratives of the Nupe , as recorded by Frobenius in 1912, the Head of the Ndako Gboya and the Etsu Nupe already shared the extortion money, which is said to have been paid in cash (200,000 cowries, i.e. the price of 4 – 7 slaves) and in kind (e.g. cola, goats, chicken) per village, at the time of Edegi (cf. Frobenius 1924.9:79-81; cf. also Kohnert (1982:62-90) on the articulation of the slave mode of production in Nupe land). 54 Crowther and Taylor (1859:215), who were among the first missionaries to arrive in Nupe land, wrote about the big mask ( Igunnuko ) of the Ndako Gboya in the Rabba-Mokwa-Region of Trans-Kaduna: “These ( Gunnuko ) dance

substitute for the expropriation of the peasantry through tribute and slave hunts, which were first restricted and finally forbidden by the British colonial power (cf. Kohnert 1982; Mason 1981). Thus, the anti-witchcraft cult in this period reconfirmed the power of the ruling class in threefold respect: Apart from acting as a substitute for the punitive expeditions against disloyal or resistant tributary villages, especially in Trans-Kaduna, it was an open sign of the power of local rulers in the profane as well as in the spiritual world and signified their judicial authority even over the occult enemies of the society, as they themselves defined them. And thirdly, by promoting the witch craze in Nupe land, it served to disguise their true goals of aggression, of exploiting the peasantry. In this respect, it is revealing that not only individual women but whole villages were accused, and the Village Head had to pay a heavy ransom to the members of the lodges.

Quite often, the villagers appealed to the Etsu Nupe , to order the Maji(n) Dodo to recall his servants (cf. Nadel 1935:441; 1954:196); but in fact, the Etsu himself was considered to be the prime instigator of the extortion tours of the Ndako Gboya , as mentioned above. This lends further support to the thesis that anti-witchcraft cults among the Nupe were principally based on class antagonism and only to a lesser extent on gender-specific cleavages, although the latter should by no means be excluded from consideration as another determinant.

However, I do not think that it is necessary to take sides, either for or against cognitive, political, feminist, or any other “pure” witchcraft theory. All of these, when taken alone, are likely to be too restrictive, although each of them may contain valid arguments. The point is, that anti-witchcraft movements may change their content and meaning over time, although their outer appearance remains the same; a change which is closely related to and interdependent with the different modes of production in which they are embedded. It seems equally important to note that witchcraft accusations themselves may not only have served opposite ends under different modes of production but might also have done so in quite different manners, i.e. the societies concerned, in adhering to the Ndako Gboya , used quite different social or socio-psychological mechanisms to achieve their goals.

As Nadel (1954:4-5) rightly observed, Nupe peasants, similar to peasants all over Africa, do not face the choice between faith and knowledge concerning the explanation of the realities of their life. They generally believe that hard work and good weather will provide a good harvest, but may hold that under certain conditions, good or evil spirits would bring the same results. Notwithstanding the fact that in African as well as in Western societies, religious or magic and scientific knowledge are complementary to rather than exclusive of one another, this absence of conflict between the supernatural and natural science becomes relevant on different levels of

about from village to village and receive cowries. They exercise some tyrannical influence over the people during the time of their appearance.”

explanation. It is against this background that I would like to forward three working hypotheses which would challenge Nadel’s witchcraft theory if they should prove to be valid. Firstly, the anti-witchcraft cult among the Nupe was created in the beginning as a direct means of defence against an aggressor, i.e. the Nupe openly pointed at the true enemy of the community, whom they tried to eliminate; therefore, at this stage of development of Nupesociety, anti-witchcraft accusations would be misinterpreted as symptoms of mental stress in the society. Secondly, even if one were to agree that witchcraft accusations of the Nupe were developed to cover up social frustrations and anxieties, it does not necessarily follow that they are to be interpreted as pathological symptoms of the society, i.e. as a (social) illness which had to be cured. If, as I maintained at the beginning, under certain circumstances (e.g. anti-colonial witchcraft movements, accusations against selfish accumulation of resources), witchcraft accusations act to attack directly or indirectly (through scapegoats) true enemies of the society, they might prove to be a powerful antidote – serving to prevent the destruction of the valid social structure.

The fact that, from the Western point of view of rational science, better medicines (i.e. social strategies) exist, may not be of great help to the stakeholders as long as the actual relations of production in which they are embedded do not favour the development of such a superior social medicine. Thirdly, socially organised witchcraft accusations (e.g. organised by anti- witchcraft cults) are to be interpreted as pathological symptoms of social conflicts only if they are mainly used (consciously or unconsciously) by dominant forces in a society to repress or exploit other social groups, i.e. if they serve the same purpose as ideologies 55. In this respect, the degree to which the witch hunt is ritualised may serve as an indicator of the strength of the ideological component of such a purge (cf. Bergensen 1978:127-28): a high degree, such as that of the Ndako Gboya performances described by Frobenius (1924.9.:65-68, 78-82) and Nadel (1954:188-201), indicating a strong exploitative character, and a comparatively low degree, like that of the current performances of the Ndako Gboya lodges (see above), pointing to a diminishing exploitative content. Only in this context do witch hunts tend to canalise social antagonism by diverting the interest from the real source of the conflicts.

But to stress the point again, “canalise” does not mean that they solve the conflict. On the contrary, in the long run, they tend to build up social tensions even further. Finally, it goes without saying that the Durkheimian methodology of “concomitant variations”, i.e. detecting relevant social divergences by further, concomitant divergences, a method which, according to Nadel (1952:264) “any enquiry concerned with social facts must employ”, will be meaningful in the latter two cases only. However, let us be clear about it. Much further and deeper empirical investigation is neces- sary before we can arrive at any conclusive theory on the social function and meaning of Nupe

55 Here, ideology is not meant in the totalitarian Mannheimian sense, but in the stricter sociological meaning of wrong or dependent consciousness.

witchcraft, especially in its present function and its relation to changes in the past. Currently, available information bears more resemblance to some pieces of an incomplete puzzle than to a meaningful picture. The shift from a subsistence to a market economy provoked not only the dissolution of the production unit of the extended family (efakó, in Nupe) in favour of smaller consumption units, which could hardly afford to support the sick and “unproductive” elder members, it also led to the deterioration of the economic and social independence of the female members of the farming household (cf. Kohnert 1982:232-42).

Could it be that the remarkable increase in witchcraft accusations in Nupe land is related to an individualisation of the socialisation process caused by the growth of the market economy, rural-urban migration, and class differentiation within the peasantry that promotes the personification of evil powers and the apostasy from traditional beliefs in collective actions of gods or one’s ancestors 56? Could it be that the increasing prominence of females as targets of witch accusations in Nupe land over the past two or three generations was caused by the combined effect of the above mentioned socio- economic development patterns, just as during the witch craze in medieval Europe (cf. Ben- Yehuda 1980:17-18)? Was the source of witchcraft accusations against females, thus, not the jealousy of the male regarding the independent economic and sexual position of the Nupe women in the 1940s (as Nadel 1954:172-81 maintained) but, on the contrary, the growing dependence of the women on their men, whereby the latter were tempted to consolidate their newly acquired power by witchcraft accusations, which may have compensated them for the dim prospects of their own economic future? Could the predominance of accusations against old women as murderers of young children result from a fear of the guilt-ridden young, male and female alike, concerning imaginary (or real) threats of their elders who felt neglected by their kin? Was the projection of this hatred into witch fantasies promoted by the fact that, still, the only legitimate method to question the authority of senior relatives in cases of social conflict within the family was to press them into the role of an outsider? All these, and many more questions await an answer, which only painstaking further grass-root research would be able to provide.

56 A similar argument was put forward by Elwert (1980:458) concerning witchcraft movements in rural Benin, and by Comaroff/Comaroff 1999 concerning Southern Africa.


Awolalu, J. Omosade (1979), “Yoruba beliefs and sacrificial rites”, Longman, London Baikie, W.B. (1867), “Notes on a Journey from Bida in Nupe to Kano in Hausa”. Journal of the Royal Geographic Society, 37.1867:92-109. Barret, Lindsay (1983), “Nigeria’s Elections – The methods and the implications”, West Africa, No. 3461 (12 December 1983:2865-68) and No. 3462 (19-26 December 1983: 2924-27) Ben-Yehuda, Nachman (1980), “The European witch craze of the 14th to 17th centuries: A sociologist’s perspective”, American J. of Sociology, 86.1980.1:1 –31 Bever, E., (2000). “Witchcraft fears and psychological factors in disease” Journal of Interdisciplinary History 30 (2000) 4: 573-590. Bierschenk, Thomas / Elwert, Georg / Kohnert, Dirk (1993),”The long-term effects of development aid: empirical studies in rural West Africa”, Economics (Tübingen), 47.1993.1:83 – 111 Clausen, Lars (1978), “Tausch-Entwürfe zu einer soziologischen Theorie”. München, Kösel, 1978 Comaroff, J.; Comaroff, J.L., (1999); “Alien-Nation: Zombies, immigrants, and millennial capitalism. in: CODESRIA Bulletin 3-4, pp. 17-27 Crowther, S. / Taylor, J. (1859), “The Gospel on the Banks of the Niger , Journals and Notices of the Native Missionaries Accompanying the Niger Expedition of 1857-59″, 1859, reprinted London 1968. Elwert, G. (1983), “Bauern und Staat in Afrika”, 1980, Frankurt, Campus, 1983. Emielu, S.A. (1981), “Guide to Kwara”. Ilorin, 1981. Evans-Pritchard, E.E. (1937), “Witchcraft, Oracles and Magic among the Azande”, Clarendon Press, Oxford Fadipe, Nathaniel (1970): The sociology of the Yoruba. Ibadan University Press Fohrbeck, Karla / Wiesand, A.J. (1981), “Wir Eingeborenen – Magie und Aufklärung im Kulturvergleich”. Leske und Budrich, Opladen, 1981 Frobenius, Leo (1912), “Und Afrika sprach”, Vita, Berlin, 1912, vol. 2, “An der Schwelle des verehrungswürdigen Byzanz”. Frobenius, Leo (1924), “Volkserzählungen und Volksdichtungen aus dem Zentral-Sudan – Der Geist des Quorra”, Jena, 1924 Geschiere, Peter (1997), “The modernity of witchcraft. Politics and the occult in postcolonial Africa”, London Goody, J. (1957), “Anomie in Ashanti!”, Africa, 27.1957.4:356-63. Kohnert, D. (1978), “Unternehmer und Grundherren Nord-Nigerias im Kampf um die politische Macht”. Afrika-Spectrum, 13.1978.3:269-86. Kohnert, D. (1979), “Rural Class Differentiation in Nigeria – Theory and Practice”. Afrika-Spectrum, 14.1979.3:295-315. Kohnert, D. (1982), “Klassenbildung im ländlichen Nigeria. Das Beispiel der Savannenbauern im Nupeland”. Institute of African Affairs (IAK), Hamburg Kohnert, Dirk. (1996), “Magic and witchcraft: Implications for Democratization and poverty-alleviating aid in Africa”, World Development, 24.1996.8: 1347 – 1355 Kohnert, Dirk. (2003), “Witchcraft and transnational social spaces: witchcraft violence, reconciliation and development in South Africa’s transition process”, Journal of Modern African Studies, vol. 41.2003.2: 217-245 Kohnert, Dirk. (2004), “Local Manifestations of Transnational Troubles: Different Strategies of Curbing Witchcraft Violence in Times of Transition in South Africa”, in: Ossenbrügge, Jürgen / Reh, Mechthild (Eds.): Social Spaces of African Societies. Applications and Critique of Concepts of ‘Transnational Social Spaces’. Lit, Münster; 2004:175-198 Lee, A.A. (1976), “Ngoga and Six Theories of Witchcraft Eradication”. Ufahamu, 6.1976.3:101-17. Marwick, M.G. (1950), “Another Modern Anti-Witchcraft Movement in Central Africa”, Africa, 20.1950.2:100-12. Marwick, Max (1964),”Witchcraft as a social strain-gauge”, Australian J. of Sc., 26.1964:263-68; reprint in: Marwick (ed.), 1975:280-295 Marwick, Max (ed.)(1975), “Witchcraft and Sorcery. Selected Readings”, Penguin, Harmondsworth Mason, Michael (1970), “The Nupe Kingdom in the Nineteenth Century – A Political History”, Ph.D. thesis, Birmingham, Centre of West African Studies, 1970 Mason, Michael (1981), “Foundations of the Bida Kingdom”, Zaria, Ahmadu Bello Univ. Pr. Meyer, B. (1998), “The power of money. Politics, occult forces and Pentecostalism in Ghana”, The African Studies Review 41(3), pp. 15-37. Morton-Williams, Peter (1960): The Yoruba Ogboni cult in Oyo. Africa, 30 (1960) Nadel, S.F. (1935), “Nupe State and Community”, Africa, 8 (1935) Nadel, S.F. (1942), “A Black Byzantium. The Kingdom of Nupe in Nigeria”, London, OUP, 1942, reprinted 1973. Nadel, S.F. (1952), “Witchcraft in four African Societies”, The American Anthropologist, 54.1952:18 – 29; reprinted in: Marwick (ed.), 1975:264-279 Nadel, S.F. (1954), “Nupe Religion”, Routledge/Kegan, London 1954, reprinted 1970. Nadel, S.F., (1935a), “The King’s Hangman; a Judicial Organisation in Central Africa”. Man, 35.1935.143:129-32. National Archives Kaduna, (NAK), various files. Olurode, ‘Lai (1990); “A political economy of Nigeria’s 1983 elections”, John West Publications, Ikeja, Lagos Parkin, D.J. (1968), “Medicine and Men of Influence”, Man, 3.1968:424-39. Parkin, D.J. (1979), “Politics of Ritual Syncretism: Islam among the Non-Muslim Giriama of Kenya”. Africa, 40.1979.3:217-33. Parkin, D.J. (1979),”Politics of ritual syncretism: Islam among the non-Muslim Giriama of Kenya”, Africa, 40.1979.3:217-33 Parrinder, G. (1953), “Religion in an African City”. London, OUP, 1953, 211p. Parrinder, G. (1958), “Witchcraft”, Penguin Books, Harmondsworth, 1958, 208pp. Redding, S. (1996). “Sorcery and sovereignty: taxation, witchcraft, and political symbols in the 1880 Transkei rebellion”, Journal of Southern African Studies 22, 2: 249-70. Richards, Audrey (1935), “A Modern Movement of Witch-Finders”. Africa, 8.1935:448-61. Richards, Paul (1997), “Soccer and Violence in War-Torn Africa: Soccer and Social Rehabilitation in Sierra Leone”, in: Armstrong, Gary / Giulianotti, Richard (eds.) (1997), “Entering the Field: New Perspectives on World Football. Oxford: Berg, 1997: 141–57 Salamone, Frank A. (1980), ” Gbagyi Witchcraft . A Reconsideration of S.F. Nadel’s Theory of African Witchcraft”, Afrika und Übersee, 63.1980:1-20. Schöck, Inge (1978), “Hexenglaube in der Gegenwart – Empirische Untersuchung in Südwestdeutschland”. Tübingen, 1978 Smith, A. (1976), “A Wealth of Nations”, Cannan Ed., 1976. Temple, O. (ed.) (1922), “Notes on the Tribes, Provinces, Emirates and States of the Northern Provinces of Nigeria”, Cass, 2nd ed. 1922; new impr., 1965 Tijani, Kyari (1977), “Research priorities for agricultural development. A critique of the irrigation project south of Lake Chad (Nigeria)”, in: Research and Public Policy in Nigeria, Report of a Conference, unpublished, ABU, Zaria, 1977:117-143 Tseayo, J. I. (1975), “Conflict and incorporation in Nigeria”, Zaria, ABU Whitaker, C.S. (1970), “The Politics of Tradition Continuity and Change in Northern Nigeria 1946-1966”, Princeton Univ. Pr., 1970 Wilson, K.B. (1992), “Cults of violence and counter-violence in Mozambique”, J. of Southern African Studies, 18.1992.3:527 – 582

Keywords: witchcraft , modes of production , informal politics , social conflict , occult belief , Nupe , Northern Nigeria .

Previous Post

Black Sherif Biography, Net Worth, Age, Phone Number

Next Post

Rahama MK Biography, Net Worth, Age, Phone Number