Anthropology, Culture and Society
Series Editors: Dr Richard A. Wilson, University of Sussex Professor Thomas Hylland Eriksen, University of Oslo
Land, Law andEnvironment: Mythical Land, Legal Boundaries Edited by Allen Abramson and
Dimltrios Theodossopoulos
The. War of Dreams: Exercises in Ethno-Fiction Marc Auge
Orange Parades The Politics of Ritual Tradition and Control
Dominic Bryan
Identity and Affect: Experiences of Identity in a Globalising World Edited by John R. Campbell and Alan Rew
Risk Revisited Edited by Pat Caplan
Women of a Lesser Cost: Female Labour, Foreign Exchange and Philippine Development Sylvia Chant and Cathy McIlwainb
Ethnicity and Nationalism: Anthropological Perspectives Thomas Hylland Eriksen
Small Places, Large Issues: An Introduction to Social and Cultural Anthropology Thomas Hylland Eriksen
Life on the Outside: The Tamil Diaspora and Long-Distance Nationalism 0tvind Fiiglerdd
Anthropology, Development and the Post-modern Challenge Katí' Gardner and David Lewis
Anthropological Perspectives on Kinship Ladislav Holy
Locating Cultural Creativity Edited by John Lilt
New Directiotis in Economic Anthropology
susana NaROIVKY
Ethnography and Prostitution in Peru Lorraine Nencel
Anthropology of the Self: The Individual in Cultural Perspective Brian Morris
Anthropology and Cultural Studies Edited by Stephen L. Nugent and Cris Shore
Social Mobility in Kerala: Modernity and Identity in Conflict
FlLIPPO osella and caroline o sella
Being There: Fieldwork in Anthropology Edited by C.W. Watson
Human Rights, Culture and Context: Anthropological Perspectives Edited by Richard A. Wilson
POWER AND ITS DISGUISES
Anthropological Perspectives on Politics
Second Edition
John Gledhtll
Pluto
4b
Press
LONDON • STERLING, VIRGINIA
Power and Us Disguises
their objectives on a global scale. Though particular situations always reflect the interaction of the local and the global, local social and cultural histories now find expression in action in ways that are part of a common experience of modernity, as I stress in Chapter 7. Only concrete, contextualized analysis of particular situations will enable us to understand what is happening and why it is happening (in Europe and the United States as well as other parts of the world). But little that is happening anywhere can be understood without reference to the historical discontinuities produced by the rise of the modern state and modern forms of power.
2   THE ORIGINS AND LIMITS OF COERCIVE POWER: THE ANTHROPOLOGY OF STATELESS SOCIETIES
Although Clastres's polemic against traditional political anthropology would be an appropriate starting-point for a discussion of'stateless' societies, the late Stanley Diamond advanced a critique of the state which is broader than Clastres's observations about indigenous South American societies. Diamond's 1951 PhD dissertation was an ethnohistorical study of state formation in the West African kingdom of Dahomey, but his wider paradigm for a 'dialectical anthropology' reflected his New York Jewish background and experience as a volunteer with the British Army during the Second World War tn North Africa. There he met Black African 'volunteers' from South Africa. Pressed into service by the South African government through the good offices of chiefly clients of the regime, these men were sent to die in an unknown land in an unknown cause in place of Whites (Gailey 1992: 4).1 A humanistic rather than 'scientific' Marxist, Diamond focused on the repressive consequences of state formation - bureaucratic oppression, racism, marginalizatton, ethnocide and genocide - and the cultural resistances which state and colonial-imperialist domination provoked.
Diamond's analysis juxtaposes 'the kin community' and 'civil society' (Diamond 19 74). He defined "civil society' as the culture of civilization, the ideologies, apparatuses and agencies associated with political organization based on the state. A particular civil society may oppose a particular state regime. Aristocrats may use peasant unrest to displace a ruling dynasty. Persons from a different social class may wrest control of the state from an established elite, bringing about change in the ideologies associated with state power. Yet, in Diamond's view, even the most radical action of civil society against a regime does not undermine the existence of states as such. Radical resistance to the alienation provoked by the culture of civilization is mounted within the sphere of 'kinship': the world of intimate personal : relations, material reciprocity and mutual aid networks, community as the enactment of shared culture in the everyday life of the lower classes. Even in
11 have drawn on Christine Gailey's a€count throughout this discussion of Diamond's life and work.
23
24
Power and Its Disguises
the modern world, this deep level of social life remained the basis for grassroots re-creation of identities and a popular cultural creativity resistant to the increasingly powerful attempts by civil structures to repress, control and define appropriate social behaviour.
Diamond's notion of kin-civil conflict' has wide ramifications for political anthropology. Kin-communities provide a model for pre-statc societies based on consensual authority embodied in custom rather than power relations embodied in law, but Diamond's deeper purpose was to highlight the continuing existence of a contested domain even in the most 'totalitarian' of societies, such as Nazi Germany. No system of civil domination, however technically perfect, could suppress al! forms of resistance. There is much more to be said about cultural resistance to domination in modern societies, and also about its possible limits. Nevertheless, Diamond deployed his dynamic concept of the state as an apparatus seeking to impose its writ on a resistant kin-community in path-breaking analyses of the development processes of early states.
In his analysis of pre-colonial Dahomey, Diamond showed how this West African 'proto-state' strengthened its domination over the local kin-community by intervening to control the reproductive and social roles of women. Irene Silverblatt (1987) has explored similar processes in the lnca Empire of Peru. The lnca state forced the local kin-communities of the empire to surrender women to it, known as adlas. Since giving women to higher status people, hypergamy, was a basic Andean way of expressing hierarchic relations, this practice had profound symbolic significance in political terms. 'Conquerors' were conceptualized as male in the Andes and the ranking of different local kin-communities (ayllus) was expressed in terms of contrasts between female original inhabitants and male conquerors. Conquered groups provided women for secondary marriages to the conquerors. Thus, according to this ideology, only males could found conqueror lineages and intermediate lineages formed by intermarriage between conquered and conquerors.
This is an ideological discourse on status, but it had practical significance in terms of the organization of the lnca Empire. The pre-hispanic Andean power system had a different cultural logic to the European system that replaced it. The lnca demanded that their subjects provide them with women for secondary marriages, and the children of these unions became lnca. As Zuidema (1964) has shown, the structure of the empire was represented in indigenous thought as a tripartite division overlaid on a quadripartite division of the empire into four quarters (suyu) and a division of Cuzco itself into two moieties. This corresponded to a division between priestly and kingly power. The tripartite division was between a category called Collana, a second called Cayao and a third called Payan, It had various meanings: it could refer to a division between lnca conquerors, the original lords of the land, and offspring of union between the two, or it could mean aristocratic rulers, the non-aristocratic population andthelncas' assistants or servants.
The Origins and Limits of Coercive Power
25
The root meaning, however, was that Collana were the primary kin of an lnca ego. and Cayao the rest of humanity unrelated to an lnca ego, from whom Collana men could choose secondary wives, thus producing the third category, Payan.
This tripartite division was full of ideological possibilities. It could be used to talk about class and the relationship between the lnca and those they conquered. It could be used to honour male leaders of subordinate groups or it could be used in a way that emphasized lnca superiority. Lastly, the category Payan, the children of lnca and non-Inca, could be used to emphasize the hierarchic unity of the lnca state - the way lnca power encompassed other groups and made their future reproduction depend on their relationship with the superior kin category, Collana.
This brings us back to the role of the acllas. The adlas were given to the lnca as a kind of tribute, to become brides of the Sun who remained virgins unless the lnca himself decided otherwise. The political hierarchy established by conquest was thus represented symbolically by the lnca in kinship and gender terms: the category Cayao could stand for 'conquered women' and for the non-Inca population as a whole. Furthermore, the lnca himself contracted his primary marriage with a sister, thereby symbolizing his ability to marry any woman in the empire. In addition to this symbolic function, however, the aclla performed a number of distinct practical functions within the lnca power system.
Firstly, they were political pawns. The lnca king gave acllas to members of the Incaic nobility as a reward for loyalty, but he also manipulated the status significance of the alienation of women to bind non-Incaic provincial elites to the empire. Local headmen felt honoured by being asked to provide a daughter to the lnca and were seen by their communities as gaining prestige by doing so. Silverblatt gives an example of a father who hands over his only daughter to be sacrificed to the Sun. The significance of the act is that the father is now recognized by the state as the headman (kuraka) of his ayllu and governorship of the area will now pass to his male descendants. The woman was finally buried alive in the lands bordering her community appropriated by the lnca state, and so this example demonstrates the second significant function of the handing over of adlas, the symbolization of lnca domination and the relationship between conquest and hierarchy. The system thus had a subtle edge to it: its logic made ambitious local chieftains into accomplices of lnca domination.
The lnca demand for women provoked resentment in the conquered communities, however, and was quite often used as a punishment against those that rebelled against the empire. To reinforce control, representatives of the lnca state claimed the right to distribute women within conquered communities: peasant marriage became a yearly mass event presided over by state officials. But the removal of acllas from their communities was a much stronger expression of domination: women made 'Wives of the Sun' were turned directly into dependents of the state. Virginity is the key symbol here.
2 g                                                              Power and Its Disguises ;
Andean women' s sexual conduct before marriage was not rigidly controlled
in the kin-communities, but the acllas' sexuality was policed by the empire i and only the Inca and persons given his permission were permitted access
to them. So the second significant point about the aclla was that the state j
now exerted power over the demographic reproduction of the Andean -J
community in a way which also reinforced Incaic domination. . .:|
The third important function of the acllas was as priestesses of a state cult, I
and even those who were sacrificed at least had the compensation of being y
honoured and revered. Besides officiating in state rituals in Cuzco, the acllas {
performed an important role as ritual mediators between their home f
communities and the capital, distributing holy bread to provincial shrines ::j
and kurakas. All the different dimensions of the removal of women from ayllu . .^j
communities into the aclla category demonstrate the principle that whilst :;;|
/ the logic of the institution begins with the gender and marriage symbolism % j
I   of conqueror-conquered relations which was already part of Andean |
1 ideology before the Incas, the Incas used this symbolism to elaborate new :;4
1 modes of practical political domination. . [
This intertwining of class and state formation with gender transforma- \ tions of this kind is probably typical of early state formation, although European colonialism produced a further deterioration in the position of
women in these societies and gender relations are further transformed by j
modern capitalist commodity relationships (Gailey 1987). The superiority '; \
of Diamond's framework for exploring political life as a dynamic process over .. |;
formal descriptions of governmental and legal institutions is apparent. Yet '. j j
there is more to be said about those societies which remain 'kin-organized'. :
The world of 'non-state peoples' can be an extremely violent one, a world : [
of incessant warfare, killings and torture. The absence of the civil institu- :j
tions of the state and authoritative community leaders does not guarantee ■ j
equality or even balanced complementarity of social roles as far as gender .
and inter-generational relationships are concerned. Inequalities based on . [
age might be transitory, since most individuals will eventually become .[
adults and elders, but those between the sexes are a different matter j (Molyneux 1977). Stateless societies may be structured in a way which
inhibits the emergence of permanent centralized authority and social strat- ':; ification, but how far can it be argued that they rest on consensual rather
than coercive relations? ;|
Some anthropologists working on Australian aboriginal society, such as John Bern (19 79) and Peter Worsley (1992), have argued that kinship and
marriage in 'stateless societies' should themselves be seen as political ;
phenomena, concerned with gerontocratic forms of power and male ; j domination of women,2 'Gerontocracy' may be not be an innocent
2 Tonkinson (1991) has, however, argued that these patterns were not universal in Australia.
The Origins and Limits of Coercive Power 27
phenomenon, a difference between men reversed over time, since 'big men' may secure more women than others in the tribe. It may also entail a permanent coercive oppression of women in addition to the transitory measures which elders adopt to frustrate the desires of younger men to possess the women they seek to monopolize. Worsley argues that the fact that younger men generally contest the monopoly of women claimed by elders turns struggle over possession of women into the main form of political conflict in' stateless' societies (1992:44). It is a primary source of the fighting and feuding which 'primitive law' has to mediate, and the problem to which gerontocratic strategies of social control are addressed.
Before I pursue the implications of this argument, it is necessary to introduce a note of caution. Feminist scholarship in anthropology has alerted us to the pitfalls of looking at gender relations solely from a male point of view, taking indigenous male representations of women at face value. To speak of 'sex roles' is to assume that there are distinct 'male' and 'female' points of view and patterns of behaviour that characterise all persons of the same physical sex in a given culture, and that these enjoy an autonomous existence, independent of the interactions between men and women and ongoing negotiation of relations between them (Strathern 1988, Guttman 1997). 'Maleness' and 'femaleness' can only be understood in relation to each other. The assumption that persons (or gods) can be unambiguously assigned to one of two genders may be problematic even in cultures that stress the distinction between male and female and expect men and women to play different roles in everyday life. From the point of view of relationships between the sexes in 'stateless' societies, we need to be alert to the possibility that ideologies and practices of male domination may be 'complicated, if not counteracted, by women's influence in socio-economic, domestic and religious matters' (Knauft 1997: 237). It is just as androcentric to look at power relations between men and women from an exclusively male vantage-point as it is to ignore the ways in which the notion that stateless societies are 'egalitarian' might need qualification from a perspective that considered gender.
The issue of power relations based on age and gender does, however, suggest a need to scrutinize more closely the vision that Pierre Clastres offers of the political life of 'societies without states'. In the next section, I argue that Clastres himself does not fully transcend the baggage of the classical sociological tradition and its models of'social order', even if his critique of the thinking of classical political anthropology remains valid.
THE INTERNALIZATION OF THE POLITICAL AS THE NEGATION OF POWER
Clastres begins with the paradox posed by the aboriginal institution of chieftainship. Most South American indigenous groups possessed recognized leaders, but these were of the kind Robert Lowie termed 'titular chiefs' - chiefs
28 Power and Its Disguises \\
who possessed no ability to issue commands which would automatically be \ obeyed. In Western terras they seem practically powerless, at least in [ peacetime. Chiefly authority was stronger in wartime, approaching a \ capacity to demand absolute obedience from other members of the war band. J Yet once fighting stopped, any power an individual acquired as a war leader j evaporated, and it was common for leadership in war and peacetime to be f assumed by a different individual. History records war chiefs who sought to f perpetuate their power by trying to extend hostilities beyond the point which their communities regarded as legitimate: the South American Yanomami t warrior Fousiwe, and the North American Apache chief Geronimo, both found that no one bar a few equally egocentric young warriors would follow .: \
them (Clastres 1977: 178-9). This, Clastres contends, demonstrates that ..< chiefs were incapable of imposing their personal desires on a recalcitrant society and translating prestige acquired in warfare into permanent authoritative power. After all, 'no society always wants to wage war' (ibid.).
Even the most warlike of societies suspends hostilities periodically in order i to replenish food stocks and to undertake ceremonial activities which A demand a cessation of aggression against other groups. Yet Clastres's i suggestion that war-weariness ultimately sets limits on the development of coercive authority ignores arguments from both Amazonia and Melanesia •;) that the reproduction of war is integral to the reproduction of a maleness :}
engaged in a struggle to assert its domination over femaleness which I J consider in more depth later in this chapter. According to Clastres, power is : t
regulated by 'society' blocking the egotism of (male) Individuals'. This I framework obscures the possible structuring of 'stateless' societies by i coercive relations other than those associated with political centralization.
To explain how the separate power associated with the emergence of the state might break out of the regulatory mechanisms of the 'primitive social order', Clastres appeals to increasing demographic density as the factor which cannot be completely regulated by social practices. He is not, however, a simple demographic determinist. He argues that the transition to statehood I is socially contested as demographic density increases, using the example of the Tupi-Guarani in the fifteenth century. This group surpassed the norm for the South American lowlands in terms of demographic density and local group size. It also displayed tendencies towards a strengthening of chiefly power, which were, however, being challenged at the time of European conquest by prophetic shamans, who went from community to community urging the people to abandon their forest homes and migrate to the East in search of a promised 'Land without Evil'. Clastres argues that the millenarian vision of the Guarani prophets thwarted the dynamic of state formation by mobilizing society at large and unifying different tribes more effectively than the chiefs had done, although it resulted in terrible suffering for those who responded to the call. The Guarani case is ultimate proof of the resistance of 'stateless societies' to political centralization.
The Origins and Limits of Coercive Power 29
The Indian chief as presented by Clastres is a 'peacemaker'. His main resource as a mediator is the spoken word. Yet not only may people decide not to heed the word of the chief, but his ritualized oratorical discourse is not really spoken in order to be listened to (ibid.: 130). The chief stands up to deliver his speech in a loud voice when the group is gathered together at daybreak or dusk, but no one gathers round and no hush falls. In a sense, the people are feigning a lack of attention, because they may indeed settle the dispute afterwards, but the content of the speech is a prolix repetition of the norms of traditional life. These empty words are not the words of a man of power, but the duty of the individual selected to speak for the community, to embody its normative principles. Thus, Clastres argues, 'normal civil power' is based on the 'consent of all'. It is orientated towards maintaining peace and harmony and is itself'profoundly peaceful' (ibid.: 22).
Not only does society deny chiefs coercive power, it also denies them scope for accumulating material wealth, by insisting they display generosity at all times: in some South American groups, finding the chief is a matter of searching for the poorest and shabbiest-looking member of the community. Lowland South American chiefs did possess one privilege generally denied to others: polygyny. At first sight, the 'gift' of women to the chief might be construed as reciprocation for the chief's services to society, the flow of speech and presentations that it is his duty to provide. Yet even with the support of his wives' labour, the chief could not amass much wealth, and much of what was demanded of him consisted of things like bows, arrows and masculine ornaments that the chief himself would have to make, Clastres suggests that the notion of a quantitative exchange of equivalent values between the chief and the group is inconceivable in this case. The absence of an obvious exchange principle is reinforced by the fact that the office of chief tended to be inherited patrilineally (ibid.: 30). The same family line retained its disproportionate share of the group's women. Furthermore, the things the chief gives to the group - goods and speech - are not reciprocated in kind, since people do not give to chiefs and only chiefs can speak.
Clastres therefore suggests that it is the very denial of reciprocity in the relationship between chiefs and community that is crucial. The circulation of gifts, linguistic signs and women is constitutive of society, but chiefs are people who fall outside the web of reciprocity:
Power enjoys a privileged relationship toward those elements whose reciprocal movement founds the very structure of society. But this relationship, by denying these elements an exchange value at the group level, institutes the political sphere not only as external to the structure of the group, but further still, as negating that structure: power is contrary to the group, and the rejection of reciprocity, as the ontological dimension of society, is the rejection of society itself, (ibid.: 32, emphasis added)
This, Clastres argues, is how lowland South American indigenous peoples provided themselves with a defence mechanism against appropriation of coercive power by chiefs. As a speaker and displayer of generosity, the chief
30 Power and Its Disguises
was totally dependent on the group of whose values he was custodian. If he failed to keep the peace or provide relief in times of famine, he was removed. j Polygyny appears to be a privilege, but the chief acquires women as a pure |
gift. This places him outside the normal framework of reciprocities and . ,;.:|
expresses his dependence on the group. "|
Despite the impressive evidence provided by the millenarian revolt of the ;;|
Guarani prophets, Clastres's own material suggests that the defences of :-:'|
'primitive societies' against state formation could come under pressure. Some :f|
pre-colonial lowland societies may have been in the process of succumbing :;| to new modalities of political power (Roosevelt 1989), and Clashes does not |
take us very far towards understanding how and why states did form in the ||
highland zone of South America. Michael Mann (1986) has argued that state ;. ;.|
formation is an exceptional process in early human history, dependent on |l
unusual circumstances. Archaeologists are also increasingly drawn to the ,;d view that state formation should be seen as an historically contingent process
rather than something inherent in the political-economic and social :
structures of kin-based communities, a 'latent potential waiting to unfold as f| soon as the right conditions appear' (Patterson 1993: 103). Once this ,f unusua! event has occurred in a given region, however, the expansion of
states transforms peripheral societies, which might not otherwise have :1i developed as they did (Friedman 1994:18-22). Itis still possible to conceive of local populations actively resisting the imposition of the new forms of
domination associated with the state. Given that there were contacts ■■[
between highland and lowland societies in South America, Clashes may not :.; j
really be analysing populations that existed in some 'pristine' world of : 'society before the state'. The main problem with his analysis is, however,
its exclusive focus on the power or lack of power of 'chiefs'. It does not attend :;
to the potential existence of other modalities of domination in 'stateless j.
societies'. ■"}
This seems a strange omission in Amazonia, given the number of anthro- . pologists who have sought to make links between warfare, male bonding |
and antagonism directed against women among the indigenous peoples of .
this region (Murphy 1959, Chagnon 1988). Joan Bamberger (19 74) used :
lowland South American examples to argue that myths that speak of a time sj
of matriarchy that provokes male rebellion are ideological charters for :|
practices of male domination. Gregor (1985) has developed this argument by ,.:| arguing that Amazonian patriarchy was sustained not merely by symbolic terrorism but by ritualized gang rape. [
This line of argument has, however, been challenged by Cecilia McCallum i (1994). McCallum argues that a vision of a violent male quest for domination
over women is discrepant with the strong moral emphasis on respect for ;j
others, self-control and pacificity that she found among the peoples of the :: j
Alto Xingu (the area of Gregor's study of the Mehinaku), and contrasts :
strongly with the tenor of male-female relationships in everyday life. She ■
suggests that Bamberger and Gregor treat ritual as a kind of 'political theatre' ■■;
The Origins and Limits ofCoercive Power 31
which recreates Ideal type social relations' in a way that imposes 'debased forms of symbolism and representation' drawn from Western cultures onto complex ritual cycles that are about making, unmaking and remaking persons and the social relations between them (McCallum 1994: 91, 109-10).3 For McCallum, constructing females as objects to be consumed and controlled by males universalizes Western models of sexuality that have no place in the Xingu, where the complementary roles of male and female sexuality are integral to the construction of all personhood and sociality.
The Xingu Park is a special kind of environment, a show-case for the Brazilian state's claims to protect indigenous people, and McCallum is working with 'pacified' groups that have sought refuge within it. Nevertheless, she offers a powerful argument against imposing Western ideas about power, gender and sexuality onto other cultures that might also be supported by considering what happens to indigenous peoples who have been strongly drawn into 'modernity'. Knauft argues that the growing association of masculinity and male agency with trade goods and money in both Amazonia and Melanesia has increased individuation and the autonomy of the nuclear ~:j family, with a consequent commodification of female sexuality (Knauft I. 1997: 244-5). He suggests that increasing emphasis on male sexual control and increased tensions in domestic relations reflect growing insecurity provoked by the transformation of a traditional sexual licence practised with discretion into an exchange of sexual favours for goods or even prostitution, and by the fact that men are forced to leave home to acquire the commodities needed to fulfil their status aspirations (ibid.: 246). This analysis is consistent with Strathern's argument that Western models of the person, exploitation and domination derive from the logic of commodity economy and 'possessive individualism',4 contrasting with Melanesian notions of personhood as objectifications of relationships which include cross-sex as well as same-sex relationships (Strathem 1988: 338). Yet Knauft is less eager than McCallum
3 Ritual processes often have the quality of'unmaking' norma) social relations or even enacting community destruction in order to rebuild and reassert, the integrity of those relationships. McCallum argues that the Xinguano view of social life is 'person-centred' and not concerned with building corporate groups or 'society' (ibid.: 105). A good example of a ritual process that focuses on building 'community' is the Festival of Games of San Juan Chamula in Chiapas. Mexico, This complex cycle enacts the cosmic destruction of the community before reconstituting it as a collectivity able to defend its autonomy against non-Indian oppressors (Gossen 1999).
4 'Possessive individualism' reflects the idea that 'society consists of a series of market relations' (Macpherson 1962: 253-4). The crucial idea is not, however, possession of property in land or moveable goods, but property in one's own person. Those who entered a wage contract or became dependent on poor relief partially alienated property in their persons, and this is how, in the seventeenth century, political rights were linked to property in things. Such people should be 'included in their masters' as far as having a voice in government was concerned, but not be denied the civil and religious liberties that all men had a 'natural' right to defend against arbitrary monarchical rulers (ibid.: 142-8). See also Gledhill (1997).
32 Power and Its Disguises
to eliminate discussion of antagonism between men and women entirely from analysis of indigenous societies before the impact of'modernity'.
SEXUAL POLITICS IN STATELESS SOCIETIES
The people of the Western Highlands of New Guinea have also acquired notoriety in the anthropological literature both for the warlike nature of their men and an apparent antagonism between the sexes. Again, we should not assume that these communities represent fossilized 'stone-age societies' which have always lived in the manner which has been observed ethno-graphically, not only because regional conditions oflife were influenced by the expansion of colonial power long before it penetrated the area directly (Worsley 19 70: 333), but because they have their own, pre-colonial, history. Gilbert Herdt (1987) writes that war was the ultimate reality for the group he studied in the .19 70s, the Sambia, because it was a pervasive fact of daily life. We can accept that circumstance and explore its contemporary significance without further analysis of historical causes. We know, however, that the Sambia established themselves as a fringe-area group in a larger regional society dominated by more numerous tribes speaking a different language. Their fears of ancient enemies formed in a matrix of concrete historical experience (Herdt 1987: 21-3).
The Sambia lived in small hamlets based on a core of patrilineal kinsmen. Clansmen of a given hamlet were allied in ritual and military matters with those of other hamlets nearby who saw themselves as descended from a common (fictitious) ancestor. Members of clans united in these 'phratries' saw themselves as brothers or age-mates: they could call on each other for assistance at times of war, and in some cases hamlets belonging to different phratries jointly sponsored collective initiation rituals, forming an 'inter -phratry' confederacy. Yet fighting with bows and arrows could break out between hamlets of the same phratry and confederacy, even if it was less likely than fighting between hamlets of different phratries that were not linked together by marriage alliances. Although intermarriage between hamlets cross-cut phratry boundaries, and, like the interphratry initiation rituals, helped moderate violence when it flared up, it was a weak political bond, since 'bonds of blood were stronger than marital bonds' (ibid.: 50). The confederacies themselves were accordingly fragile.
Local fighting was subject to institutionalized constraint. Feuds between hamlets, premised on sorcery accusations, adultery or retaliation for an insult, would be settled by 'ritualized' exchanges of arrows on designated fight-grounds. The principle that the fighting should not be lethal could not be observed in practice on every occasion, prompting warriors to rationalize a death by saying: 'He stood in front of my arrow' (ibid.: 49). Bow-fighting was practised as a military game which was part of the process of training boys to be warriors and was also a means by which adults could compete for
; The Origins andLimits of Coercive Power 33
martial prestige, but ritualized bow-fights between hamlets could .escalate into 'big-fights'. In these, killings were more likely, war leaders prepared plans for ambushes and raids, and shamans turned lethal sorcery on the hamlet's enemies. Initial escalation usually followed an 'accidental' killing. | The original parties would mobilize allies for support, and this process tended
I to lead to further escalation as 'stranger' groups were drawn into the conflict
I through alliance networks. They arrived wielding clubs and axes, the most
j lethal weapons at the Sambia's disposal.
J This level of escalation and the accompanying bloodshed did, however,
' prejudice inter-hamlet relations to a point where the continuation of basic
! social life was threatened. Truce ceremonies, enacted between warriors on
1 the fight-grounds, might lead eventually to a restoration of peace through
| blood compensation payments, but compensation was not as significant an
i institution in the Sambia case as in other parts of the Western Highlands,
j Extended periods of fighting only tended to die down after blood revenge had
■i been exacted and food supplies had been disrupted by women's inability to
tend the gardens. Although truces tended to be short-lived, peace did permit ! elders to negotiate new marriage alliances and conduct collective initiation
rituals, allowing social life to continue in as 'normal' a manner as the Sambia could hope to achieve given their commitment to violence. Although it was ] inter-tribal warfare against groups defined as eternal 'enemies' to be slain in
which the Sambia warrior ethos reached its destructive height, assassinations and sorcery accusations abounded in the competition between warrior big-men even at the local level.
Sambia men spent much of their time preparing for war and guarding against surprise attack, including sorcery attacks ascribed to women acting in league with their enemies. Sambia society was not entirely devoid of peacemaking institutions. Patterns of conflict were regulated by a degree of ritualization and the structure of alliances which linked patrilineages in one hamlet with those of another, though these cross-cutting ties could equally I well pit men of the same hamlet against each other in inter-hamlet fights in
which their hamlet as a whole was not directly involved. The Sambia i recognized different forms of warfare. They tried to limit the intensity of fights
I by specifying the weapons to be used and level of combat, but violence could
escalate even from 'controlled' and apparently harmless beginnings. Faced with constant warfare and fragile political alliances, Sambia villages put up the stockades and trained their boys to be aggressive warriors. As Herdt points out, warfare is a complex phenomenon that can have many causes, including the ambition of leaders and competition for resources. His own analysis focuses, however, on the way that violence was driven by a Sambia I 'warrior ethos', founded on a need to compete that destroyed resources and
j devalued humans, and ultimately rested on the desires of Sambia men to
j dominate women.
I There is a danger here of ascribing Sambia aggression to male psycho-
logical drives. This, and the alternative argument that 'male solidarity' and
34 Power and Its Disguises _-|
'gender antagonism' are 'functional' responses to conditions of endemic | warfare, has been addressed in an innovative analysis of the political , dynamics of Melanesian warfare in the Sepifc area by Simon Harrison (1993). Most analyses of' primitive warfare' accept the Hobbesian argument that the state imposes peaceful relations on individuals who are 'naturally' inclined to violence outside the immediate moral universe of kin communities (Sahlins 19 74). Yet Harrison notes that what the state accomplishes is not \ the abolition of internal violence but its classification as 'illegitimate' [ (Harrison 1993: 149). Melanesia refutes the assumption that peaceful t sociality is a condition that 'stateless societies' are only able to maintain I within the narrow circle of village social ties, since Melanesians are perfectly \ capable of maintaining peaceful relations with distant trading partners. i Harrison suggests that we focus instead on the way male ritual cults are :.i organizations for producing war, for negating conditions of peaceful sociality n that are assumed to be 'normal' (ibid.) He sees such male ritual as an 'imposition' that embodies male symbolic idealizations of'power' which are political in two senses: they construct a 'community' that is an enduring entity 'externally bounded against "enemies" and internally structured by inequalities of age and gender' (ibid.: 148).
[Thus] what men are demonstrating in ritual, and in warfare itself, is a kind of power and freedom from accountability that open up two interdependent spheres of action: both to kill and make war, and to act politically in their dealings with women and juniors, (ibid.: 144-5)
This account can readily be applied to the Sambia. Men assert that their dominance is necessary to defend 'society'. It rests on a peculiarly male essence, jerungdu, which is a life-force substance embodying uniquely masculine qualities of bodily and spiritual strength. Semen supplies ierungdu, and men must possess it in abundance if they are to be true men, something they must demonstrate by sexual and military prowess. This amounts to an injunction to dominate both women and other men. But jerungdu is not innate: neither women nor boys possess it, and it must be acquired from 'real men'. In Sambia cosmogony, the original couple from whom they spring are hermaphroditic age-mates, but one is bigger and older. Since both have penises, the stronger of the pair copulates through the mouth of the partner. This process transforms them into male and female, though the primal male has to cut open a vagina on the female to allow the first child to emerge into the world (Herdt 1987: 167). The first-born son is forbidden sex with his mother by his father, and told to have fellatio with his younger brother, who thereby in turn becomes masculinized. The founding myth of the Sambia thus explains what is at the core of their initiation rituals: what Herdt originally termed a 'ritualized homosexuality' that allows transmission of manhood through semen to boys from bachelors, youths who have themselves acquired maleness and become strong enough to take their place in the male society of warriors and give maleness to younger boys.
The Origins and Limits of Coercive Power 3 5
In a new preface to his first book on the Sambia, originally published in 1981, Herdt abandons the term 'homosexuality' for these same-sex practices, renaming them 'boy-inseminating practices' to eliminate the connotations of Western concepts of sexuality (Herdt 1994: xiv). They belong to a family of practices found in association with 'stratification by age, with emphasis upon the intergenerational qualities of power, knowledge, manliness, prowess in war, honor and virtue being transmitted from older to younger male' (ibid.). The Sambia did not respect same-sex relations in men who were fully socially adult. Those who continued to prefer oral sex with boys rather than vaginal sex with women which could produce children were 'rubbish men' who had failed to make the full transition to manhood. Herdt argued that the logic of Sambia notions of sexuality was determined by male models of power and what was ultimately a fear of women. Jerungdu is a form of power restricted to men that can only be acquired from men, and how much of it one has determines how much of a man one is. The role of bachelor as sperm-giver and boy as sperm-taker cannot be reversed, and the transition to sex with women is made possible only by the growing strength of one's manliness: vaginal sex with a woman threatens pollution and even depletion of jerungdu. Sambia men were often loath to start having vaginal sex with their wives, favoured the missionary position and orgasmed quicldy. Newly-wed men often continued to take semen from male partners for a while after they began to have sexual relations with women, something they kept secret from the former.
The lengthy Sambia initiation cycle subjects boys to an arduous process of physical ordeals designed both to foster warrior virtues and to end the grip of mothers over their male children. One of the most dramatic episodes in Herdt's ethnography is a moment when young men physically attack the mothers of initiated boys after the nose-bleeding ceremony, in which cane stalks are forced down the nostrils of the screaming children and the blood is collected on fern leaves. Two men try to ram the bloodied leaves down a woman's throat, and another charges into the group of mothers, with bow and arrows in hand, and curses them, provoking anger and demands for revenge on the women's part. Herdt suggests that: 'It seems that the blood and the sight of women here can create violent reactions in [young men]. Something in their gender identity is so touchy, like an unhealed wound' (ibid.: 152). The symbolism of the initiation rites in fact reveals a frequent association of key male symbols with femaieness. The phallic ritual flutes used in the initiation rituals to teach the mechanics of fellatio are supposedly hostile to women and children. Yet they also have female associations: they are called 'frog female' because they sound like a frog croaking in the forest, only women hunt frogs, the initiates' mothers fed them frog two days before and the forest edge-land is where dangerous female spirits live. Not only is the spirit of the flutes female, but the myths say they were made by women. But these are dangerous women, not the submissive, obedient women of Sambia male ideology.
3 6 Power and Its Disguises ~-[
The rites separate boys from the world and power of women, but Sambia seem unable to deny the necessity of the female contribution to social reproduction even as they engineer this separation: semen is likened to mother's „1 milk (ibid.: 150). Women, it is conceded in the heart of these male rites, are naturally fertile and reproductive, and men are not (ibid.: 190). The ritual , process aims to remove female residues from the body of the initiate and all \ 'femine feelings' and behaviour. Yet its repetitiveness and the aggression ! towards women manifest throughout the cycle suggest that men never truly f believe that their cultural mimicking of women's natural fertility is sufficient | to establish the superiority of male power over 'natural' female power. If t the women knew that men acquire their power only by playing the woman to an older man, the basis for male gender domination would be threatened. I The secret fear of Sambia men is that women are really more powerful than they are.5
The secret society organization of Sambia men is therefore a political orga- ! nization in Harrison's sense. It involves a transfer of control over young men from their mothers to the elders. The latter need to control the products of women's labour and of their bodies, including the female children who will be needed to obtain wives for the bachelors. They also need to control the bachelors themselves, who cannot yet be allowed sexual access to the ; women around them. To accomplish the last objective, elders resort to repetition of the nose-bleeding ordeals. Herdt suggests this is a kind of .■: t
'symbolic castration', steering bachelors' sexual impulses away from women i towards younger initiates (ibid.: 185). Ritual beliefs about the polluting nature of women's bodies reinforce bachelors' avoidance of women, and j warfare redirects any anger youths might feel towards the elders towards i external enemies.
Yet the power claims of Sambia masculinity remain fragile. Sambia ' recognize the power of female and hermaphrodite shamans, and women who are shamans can participate in the healing of men. Once again, notions of motherhood and the natural fertility of women surface, and this time not simply in myth but in social practice. That coercion and violence enter into relations between the sexes among the Sambia reflects the cracks in the power structure over which male elders preside. Here power is negated in Clastres's sense because the male part of society is dedicated to the negation of the power of the female part, which it is nevertheless forced to acknowledge, in its myths and in its dreams.
5 Some New Guinea peoples went further in exorcizing the spectre of female power. The Iqwaye cosmogonic myth denies the necessity of the female role in sexual reproduction by having man and the cosmos created from a primal self-creating being with his own penis in his mouth: as the creator and first man vomits his own semen and blood, thus were created all things in this world. The creator Omalyce was at once his own father, mother and son. The Iqwaye therefore close their ideological system more successfully than the Sambia (Mimica 1988: 74-87).
The Origins and Limits of Coercive Power 3 7
Herdt concedes that he did little work with women in his first fieldwork. His analyses were written from his own male point of view as well as from that of his male informants. As Gutmann points out, following Young (1983), the assumption that women are 'muted' or inarticulate in 'traditional societies' is in danger of becoming a self-fulfilling prophecy (Gutmann 1997: 848). Herdt's ethnography shows women as present throughout the ritual sequence which lies at the core of his analysis. They actually confront men at various points, He argued, however, that the assumption by women of a complementary role to men in performing the ritual indirectly reinforced the male position, despite the fact that women and men ended up confronting each other, verbally and physically, as a group (Herdt 1987: 134). Yet though the Sambia men said that, they regarded the women merely as spectators, their role as an audience was neither trivial nor passive: women acted in a way which sanctioned the necessity of initiation, because their antagonism to men as husbands in the rites reinforced the separation of the male world into which the initiate was moving.
There is a point in the initiation rite, during the moonlight rituals of the dance-ground, at which the women perform their own ceremony, the Firewood Ritual. Herdt describes the rhetorical chastising of initiates and adult men by women in these nocturnal events as 'rituals of rebellion' (ibid.: 132), following Gluckman' s analyses of ritualized confrontations that allow the tensions provoked by domination to be expressed and released in a way which allows the system to continue functioning (Gluckman 1954,1955). The ultimate effect of the ritual dramas enacted is to realize the separation of the initiate from the female power domain of motherhood.
Yet men worry about hiding the 'secret' of the flawed nature of their power from women. How can we be sure that the women do not know the secret, and are Sambia men themselves convinced of their ignorance? Female contestation of male dominance claims in Sambia society might run deeper than the opposition visible in the ritual process. Even that ritualized opposition seems less than trivial given that it provokes violence that is not entirely ritualized. Ritual is, however, not 'everyday life', and as McCallum suggested, we should also take everyday relations between men and women into account in framing our interpretations. Her own analysis suggests that the dangerous states of being created in the making and unmaking of an order of things in ritual contexts can produce 'unscripted' reactions from the participants. Yet although the Sambia case is consistent with the kinds of generalizations Strathern has made about the role of both male and female in creating 'personhood' in Melanesia, these same generalizations are integral to Harrison's model of the political role of male cults as an effort to establish the kind of freedom of accountability that is evident in a different form in the behaviour of war chiefs (Harrison 1993: 144). The elimination of all notions of 'domination' from 'stateless societies' therefore seems unwarranted. The tensions that result from the attempt by older men to privilege their own status through ritual cults seem, however, of a quite different order to the effects of contemporary forms of male domination,
3 8 Power and Its Disguises
expressed in the commodiftcation of women's bodies, the divide that Knauft suggests separates the Melanesian past from the present.
CIVILIZATION, MOTHER OF BARBARISM
The issues raised by an exploration of ways in which the workings of 'stateless societies' inhibit centralization of political power make it more difficult to explain how the first states in human history formed. We cannot, however, assume that the 'stateless societies' we know about from ethnography represent the kinds of societies which existed before there were any states or civilizations at all {Fabian 1980). No anthropologist has been able to study a human community unaffected by Western civilization by ethnographic methods, let alone the societies which actually preceded the formation of the earliest civilizations of the Old and New Worlds.
The problem of 'the Origins of the State' is therefore one to which political anthropology can make a contribution only in collaboration with archaeologists. The nature of that contribution, in my view, lies primarily in using ethnographic data to problematize theorizing about the social mechanisms that might have led to the centralization of power and to pose questions about the nature of power and inequality in early state societies (Gledhill 1988a). There are areas of the world, such as Polynesia, where close cooperation between archaeologists and anthropologists has already shown its potential (Sahlins and Kirch 1992). In most cases, however, the interface is indirect and theoretical. It is seldom possible to combine even ethno-historical and archaeological data in an investigation of state origins in most regions of the world for the simple reason that indigenous or early colonial written sources usually relate to later episodes in cyclical processes of state formation and collapse, the beginnings of which lie in a remote past.
Much of the vast literature on early state formation is not, however, about state 'origins' in the literal sense, but later episodes in such long cycles. Seen in terms of the 'ecological paradigm* that stresses the role of demographic pressure as the underlying motor of social evolution, all processes of social and political change result from essentially similar causes. If, however, we stress the importance of cultural strategies of power and the actions of social agents in forming alliances and negotiating relations of domination as central variables in these processes, explanations are likely to be much more complex (Brumfiel 1992, Patterson 1993). Each successive episode will reflect the impacts of previous historical experience and transformations in social practices and world-views associated with the civilizational process.6
6 As Diamond stressed, 'civilization' is a cultural system in which the hierarchic principle becomes a presupposition of social life. It is, however, also important to note, as Edmund Leach showed in his classic Political Systems of Highland Burma, that hierarchical and non-hierarchical models of local politics can continue to coexist in an unstable and oscillating state on the margins of more stably hierarchized polities (Leach 1954).
The Origins and Limits of Coercive Power 39
All societies in a region may be changed by state formation, including those on the periphery of expanding centres. This area of research is important to political anthropology in another sense. It encourages anthropologists to think about societies studied ethnographically as components in larger regional systems, and to reflect on how their organization might have been affected by these wider relationships before the era of Western colonialism. If we fail to do that, then we not only treat the West's 'others' as 'peoples without history', but miss opportunities to understand why particular human communities organized themselves politically in particular ways.
In his path-breaking studies of the nomads of Inner Asia, Owen Lattimore (1962) suggested that the social and political structures of these pastoral societies should be understood in the context of their long historical interaction with the agrarian heartlands of China, The nomad chiefdoms were organized into a structure of clans whose segments were ranked. Although the power of the chiefs was limited in peacetime, the hierarchic order of a chain of command was present in embryo in this political organization. At one level, this seems merely another case of the distinctive powers accorded to leaders in war and peacetime mediators, but Lattimore suggested that there was a deeper logic to nomad organization. Imperial China liked to think of the barbarians as its inferiors and subjects, but practised a complex diplomacy on the borders of its domain, accepting the nomads' gifts as tribute paid by those who acknowledged Chinese hegemony, while carefully distributing resources of far greater value in return. These buffer mechanisms were indices of Chinese weakness: periodically the nomads erupted across the border as the imperial state grew weaker and exhausted its energies putting down peasant rebellions and unruly provincial lords. Lattimore suggested that these moments revealed the underlying rationale of the nomadic clan organization: it enabled the nomads to achieve rapid consolidation of administrative control over the territories they conquered, including re-establishment of tribute payments by the local peasantry.
In a general overview of the relationships between 'centres' and 'peripheries' in the processes of state formation, Gailey and Patterson (1988) suggest that: 'the emergence of state societies has immediate effects on the stratification and production relations of surrounding societies' (ibid.: 86). Not all states are the same, however, and these writers draw a distinction between 'strong' and 'weak' tribute-based states. An example of the former is the Inca state after the 1430s, and the latter category would include African kingdoms like Dahomey, which relied on external slave-raiding to supplement the tribute the centre extracted from the subject population within its borders. Gailey and Patterson focus on how the development of each kind of state influenced the development of peripheral societies. They classify peripheral societies whose nature has been changed by articulation with states into two broad groupings.
Heavily militarized strong tribute-based states promote the development of societies based on what Marx described as the 'Germanic mode of
40 Power and Us Disguises
production'. The organization of production is atomized at the household level, although households' access to common lands and other resources depends on their participation in larger communally organized activities, such as age-grade ceremonies and military raiding. The relationship between the peripheral society and the neighbouring state has several dimensions. The periphery supplies the state with exotic prestige goods and with slaves captured from neighbouring groups. Any members of the community who achieve privileged positions in this trading system may not only become wealthier, but be drawn into the state system itself, as retainers, military officers or functionaries. They are, in fact, peculiarly suited for such roles: as 'strangers' they can be used in place of members of the state elite who may harbour political ambitions of their own, particularly the collateral kin of the existing ruler. Over time, differences in wealth emerge between households in the peripheral society, but this alone does not necessarily lead to their transformation.
That comes where slave-ownership becomes the basis for the emergence of an internal elite or when militarism is turned inward on the peripheral society itself, and poorer households are forced to turn to their more powerful neighbours for protection and come to accept the dependent status of feudal retainers. Should the tribute-based state itself collapse, the peripheral society may remain household-based without continuing its raiding or attempting to subjugate neighbouring peoples, revert to a less militarized and economically atomized 'communal mode of production', or undergo a transformation into a weak tribute-based state on the fines of the Mongol state that emerged on the periphery of China after the collapse of the Sung dynasty in the thirteenth century.
The second type of peripheral social formation Gailey and Patterson discuss is that associated with the 'lineage mode of production', This concept owes its origin not to Marx but to the French Marxist anthropologists Claude Meillassoux and Pierre-Philipe Rey (Rey 1975). These societies are defensively organized in non-military ways to resist the encroachment of weak tribute-based states on their lands and people (through slave raiding). Use-rights to land, labour and products are restricted to members of corporate kin groups (which are not necessarily unilineal descent groups). Control of resources and people remains communal, but higher-status persons within the corporate kinship group exercise a differential control. This has implications not merely for inter-generational relations but for gender: tighter kin control over resources not only weakens cognatic claims to labour and produce, but, in patrilineal-patrilocal settings, the role of sister is diminished in importance relative to wife-mother roles by the diminished contact with natal kin. Male control over marriage is strengthened, and with it, gerontocratic authority. Gailey and Patterson suggest that lineage societies may themselves transform into weak tribute-based states, when the state neighbouring them itself collapses. Alternatively, they may revert to less restrictive communal relations, and can also transform into 'Germanic'
The Origins and Limits of Coercive Power 41
systems and militarize themselves where they compete with neighbouring Germanic societies (or communal control over resources at the corporate group level weakens). If that proves impossible, the population may simply migrate out to more remote 'regions of refuge'.
Gailey and Patterson present 'Germanic' and 'lineage' societies as examples of how state formation and expansion transform kin communities on their periphery. They are both transformations of a 'communal mode of production' where 'property needed for subsistence is held by the group as a whole and rights to use it are available to all on the basis of gender, age/life status, and kin connections' (1988: 80). Differences in status and wealth may exist within the political community. It may be one in which there is a permanent institution of 'chieftainship' even if chiefly power is unstable, chiefs are removed from office by assassination and usurpation, and different kin groups rise and fall in status as individual chiefdoms develop, expand and collapse. Societies that belong to the communal mode of production do not, however, display the atomization of control of resources by households found in the 'Germanic' mode, nor the restrictions on access to resources to members of kin corporations foundin the 'lineage mode'. Claims to usufruct of resources can be made bilaterally, so that the boundaries of kinship groups remain flexible. People can choose to affiliate to mother's or father's natal group, as was the case with the Andean ayllus (Spalding 1984).
For some, this attempt to rework Marxist concepts for new purposes may seem an unappealing reduction of ethnography's rich variety into a tight classificatory scheme. Nevertheless, it is hard to deny the importance of the principle that motivates it. The forms of human social and political organization cannot be seen simply as the unconstrained exploration by 'people' of a series of logical possibilities, as if every human community sat isolated on an island in the midst of a limitless ocean, It has also proved difficult to explain social structures simply as 'adaptations' to local techno-environ-mental conditions. Other societies are part of any human group's environment. The political communities we term 'states' are expansionary organizations that draw in human arid material resources from beyond their borders. In some cases, drawing in resources and people from the 'external' periphery is essential for the reproduction of the state itself, since its elite cannot extend their exactions from the population they rule beyond a certain point without provoking revolt (Carneiro 1981). Thus centres both create and (unintentionally) transform peripheral societies. State and stateless societies develop in ways that are fundamentally interdependent.
'STATELESS SOCIETIES' UNDER THE MODERN STATE
It would be inappropriate to end a discussion of ethnographic 'stateless' societies without underscoring the need to understand their transformation by colonial and post-colonial states. To do so, I will return to the classic
42
Power and Its Disguises
The Origins and Limits of Coercive Power
43
African case of the Nuer and Sharon Hutchinson's updating of Evans-Pritchard's work (Hutchinson 1996).
Today the Nuer experience national state power with a vengeance in the form of the Arab-dominated, Islamic regime in Khartoum, which renewed a brutal civil war against the peoples of the south of the country after a decade's hiatus in 1983. Civil war brought Nuer more violence than they had ever known and their sufferings drove many of them into a life of poverty as migrant workers and refugees in urban slums. Adjustment to the new order of things proved traumatic. Many migrants from the south were forced to seek a livelihood doing the most degrading land of work, such as emptying latrines. Nuer initially considered that a cow purchased with 'the money of shit' could not live, and extended this idea of the polluting consequences of degrading work to other types of labour (Hutchinson 1996: 84), In a delicious example of how women selling beer tried to keep money polluted by association with degrading work separate from the rest of their income, Hutchinson records how one woman remarked, with a smile, that 'It's going straight to the government' (ibid.). 'The government' was hardly a novelty for Nuer by the 1980s. The state was already part of their lives at the time when Evans-Pritchard presented them as a paradigmatic case of an 'acephalous' political organization, since it was the colonial state that made the world safe for anthropologists to construct such models.
Hutchinson traces Nuer experience of national state power through the colonial period to the 1990s. Nuer themselves talk about this experience as 'the age of government' - and also as 'the age of the gun'. Nuer themselves made efforts to get guns from early colonial times (ibid.: Ill), and their history is one of both resistance and accommodation to the state. Although Nuer hate and despise the Arab government in Khartoum even more than they hated the British, they do not entirely reject another kind of state-like authority today, that of the Sudanese People's Liberation Army (SPLA). Hutchinson shows that the SPLA has in fact been extending administrative policies begun by the Anglo-Egyptian colonial administration (ibid.: 146-9).
Hutchinson argues that Evans-Pritchard's model of Nuer 'ordered anarchy' was undermined by developments in the colonial period. In part, this was because the British did actually manage to create new kinds of leaders, although the result of these efforts was to fragment the Nuer political system even more. Far from fostering social peace, colonial policies actually accentuated the centrality of feuding in Nuer life, and the old systems of ritual mediation became less effective (ibid.: 131-2), When a man died in traditional inter-community disputes, the kin of the killer had offered a woman in marriage to the dead man's kin, so that he could live on.in the children born to her. The British thought it would be a better idea to ramp up the rates for blood compensation, removing the old incentives to restore peaceful relations through marriage. But the most fundamental change Hutchinson charts through the colonial and post-colonial is that in Nuer ideas about the polluting effects of homicide.
Traditionally, a killing among Nuer could kill the slayer through a mysterious transmission of blood between them, the form of pollution called nueer. The role of the 'leopard-skin chiefs' centred on the removal of this pollution, and one of Hutchinson's informants told her that nueer pollution itself should be seen as 'one of our chiefs' in the past (ibid.: 124). The danger of pollution underpinned the rules of violence in Nuer society, an ethical code that prevented use of the most lethal weapons in fights between kinsmen and neighbours. The use of guns and the role of colonial courts began to undermine these ethical codes. Government-appointed chiefs began to pervert the blood compensation rituals by taking most of the sacrificed animals for themselves and the police who supported them. The major change lay, however, in the differences between guns and spears as killing instruments, and the fact that so many people were killed in Sudan's first civil war (1955-72). Gun killings were at first compared to death by lightning (ibid.: 108). People who died in this way were initially seen as killed by God, and became col wic, potentially dangerous lineage guardians who had to be propitiated and whose kin had to give priority to getting them a 'ghost wife' (ibid.: 138-40). But the number of people dying from guns increased whilst the number of cattle available for bridewealth payments fell radically because of the devastation caused by war. The SPLA commanders therefore decided to promote the idea that people killed in the war were not col wic, and that killing in a war with the government did not involve pollution dangers (ibid.: 140). There were precedents for this in past history, since Nuer who had acted as government chiefs and policemen had been involved in judicial executions, but did not hold themselves personally responsible for the deaths.
In saying that killing for the SPLA was morally different from feuding, the SPLA was effectively making its own claim to be recognized as a legitimate government. But the implications were radical. Before, Nuer had located the meaning of gun-related violence in a concept of a divinely ordered world: now they were losing confidence in supplicating God as a way of dealing with violence, and seeing violence itself as increasingly inevitable as the Sudanese post-colonial state disintegrated. The gun became a symbol of masculine power in an increasingly individualized Nuer society, a way of winning back self-esteem (ibid.: 153). Gun symbolism became a fetish, replacing the cattle and spears of Evans-Pritchard's day. The problem is, however, that each successive state regime commands more fire-power than Nuer do, and men have great difficulty seeing themselves as successful in fulfilling traditional roles as defenders of women, children and herds. Because guns are too expensive for individuals, the gun has ironically become a means of promoting a continuing collective solidarity at a time when the older binding force of shared claims on cattle acquired from marriages is declining (ibid.: 150-1). Yet, on balance, the overall picture Hutchinson paints is bleak. Although there is still some continuity of old Nuer cultural concepts which were central to the construction of sociality and regulation of violence, the
44 Power and Its Disguises
experience of successive state regimes has undermined those systems without replacing them with alternatives that work.
This is, of course, the opposite of what classical anthropology, with its faith in the civilizing mission of the Pax Britannica and the ultimate superiority of the modern liberal state and its institutions, assumed would be the future of colonized peoples. It is also not a particularly good advertisement for the supposed benefits of 'government' in general.
3   FROM HIERARCHY TO SURVEILLANCE: THE POLITICS OF AGRARIAN CIVILIZATIONS AND THE RISE OF THE WESTERN NATIONAL STATE
Debates about the 'origins of the state' may seem more relevant to political philosophy than to contemporary political anthropology. Yet it is less obvious that the same can be said about the analysis of the great 'agrarian civilizations', the combination of a stratified, agrarian-based society with one of the world 'religions of the book', such as Buddhism or Islam. The agrarian civilizations of the Near and Far East were world-historical rivals of the Latin Christian civilization of the West, and the multi-ethnic religious communities they established continue to be a force in modern global politics, as Europeans were so sharply reminded by the tragic events in Bosnia in 1993. The European response to the Bosnian problem suggests that the shifting frontier between Christendom and Islam remains salient to the very identity of 'Europe' and 'the West*. The later Kosovo crisis reminds us that that Western Europe's identity is also entangled in the division between Catholic and Eastern Orthodox Christianity, a consequence of the collapse of the Roman Empire in the West and its partial survival in the East in the form of the Byzantine Empire. Although Islamic governments ultimately failed to make a decisive collective intervention in the Bosnian conflict, individuals from Islamic countries volunteered for service with the Muslim forces. The presence of nationalist Russian volunteers in the Bosnian Serb forces was followed by the celebrated dash by a Russian column to take control of Pristina airport before the advancing NATO ground forces. Although the motivations of the actors involved in the break-up of Yugoslavia should be sought in the present, and ethno-nationalism is clearly a more general global phenomenon, history, and not simply twentieth-century history, remains important for understanding the deeper meanings with which contemporary actions may be invested (van de Port 1999).
This underscores the point I stressed in Chapter 1, the usefulness of trying to understand 'modernity' at a global, cross-cultural, level. Such a perspective not only sheds light on the contemporary politics of religion but illuminates other aspects of the division of the world into geo-political blocs which are based on essentially similar forms of political and economic orga-
45