My father was so Amerindian… he no went school, and he could kill and resurrect you in a day’
(Patamuna Roger Edwin, 1997)

Quien se va al Orinoco, Si no muere, se vuelve loco.
(Venezuelan proverb)

1 This essay explores the way in which shamanic complex of kanaimà becomes an articulation of the savage in both colonial and national cultures of north-eastern South America. The shamanic complex of kanaimà has become a central representation of the savage in both colonial and national cultures of north-eastern South America. Together with the cannibal sign, kanaimà is used to construct a vision of Satanic Majesty in Amazonia. This vision of a Satanic Majesty holding sway over indigenous cultures is used in turn to produce a Demonology of Development. This Demonology pictures tradition and ritual, as nothing but superstition and primitiveness, serving the ghoulish and violent desires emanating from this Satanic Majesty, who is therefore the hidden enemy of Development and Progress. This idea of Development, in western discourse, alludes to forms of both material and spiritual redemption and advancement, and is understood here as part of the colonial and national conquest and incorporation of indigenous communities under the power of the state and government.

2Kanaimà, a term which refers both to a practice and the practitioners, is a form of mystical assault that ritually requires the extensive physical maiming of its victims in order that they may be produced as a divine food. This carnal violence, primarily a mutilation of mouth and anus, renders the victim near dead but the process of slow death that occurs allows the formation of the magic substance maba within the body of the victim. The putrefaction of the cadaver in its grave then completes the cooking of the maba which the kanaimà shamans suck from the belly of the victim. This substance cools the heat of the killers, augments the senses and physical abilities of the kanaimàs, and facilitates the location, tracking, and eventual assault of future victims (Whitehead, 2002a).

3Kanaimà is currently deployed in the cultural practice of the Patamuna, Akawaio and Makushi of the Pakaraima mountains, in various ways but it is important to notice that it is not just a form of internal dialogue amongst «natives» but also enters into the self-fashioning of the Guyanese, Venezuelans and Brazilians. As a profoundly «authentic» icon of Amerindian survival, it has been appropriated by the national societies of the region as a sign of their emplacement in the wild and alien landscapes of the «interior». This idea of the «interior» posits Guayana as the end point of exploration, the counter point of modernity and the obstacle to development. Accordingly the language of conquest and occupation still suffuses the national imagining of this region and kanaimà comes to stand for that alterity. In turn the encounter with kanaimà becomes a token of the traveler’s, or anthropologist’s, or missionary’s, or miner’s penetration to that inner mystery. In this way kanaimà becomes the metaphoric equivalent of the Conradian «heart of darkness» where there is nothing but «the horror the horror».

The Demon Landscape

4This articulation of savagery through reference to kanaimà is culturally grounded, quite literally, in ideas and metaphors of physical space. Some recent work on the cartography of this cultural space in the eyes of colonial travelers such as Henry Bates (Raffles, 2002) Robert Schomburgk (Burnett, 2000) and Walter Ralegh (Whitehead, 1997a) allows more detailed connections to be made between this peculiarly Guayanese notion of the savage with the broader aims of colonial and national control of this permanently wild interior. This more detailed conjoining of the idea of Guayana with that of native savagery is not just the passing allusion of a few authors or travelers, but is a fundamental trope in the imagination of «Guayana» from the inception of European exploration through to the present day (Lézy, 2000; Whitehead, 2002b, 2004). Nor is this imaginary confined to colonial mentalities alone, since kanaimà has also fed the national imagineries of Brazil, Venezuela, and Guyana.

5The zone of kanaimà is thus both an intellectual construct and a physical space that comprises the geographical landscape of the region between the Orinoco and the Amazon rivers, defined to the west by the fluvial connection of those rivers via the Rio Negro and to the east by the Atlantic ocean. In other works (Whitehead, 1993, 1994) I have referred to this physical space as «Guayana», using the old Spanish spelling to indicate its epistemological priority over the colonial and national political territories of the Guianas (Guyana, Surinam, Guyane) and follow that usage here. In the earliest accounts of South America the land mass south of the Amazon was referred to as the «Island of Brazil», and by extension the territories to the north were thought of also as an island. The fluvial orientation of human connections in this region also underwrites this perception phenomenologically since the rainy season inundates the rivers, creating for the traveler flooded vistas from which the land only intermittently rises such that «islands» appear everywhere. In turn a political phenomenology of landscape was present in the forms of native topography such that key settlements were often designated using the suffix «-cai», a native word for island.

6This meta-geography is also apparent in the continuing experience of place since the approach to the Orinoco from the north is quite dramatic for the way in which it is signaled by pronounced differences in landscape forms (Gibbs and Barron, 1993). The flat hard llanos, composed of greyish shale oil soils, are suddenly replaced by the red-earth that is indelibly Guayana. Such red laterite soils also feed ideas about the poverty of Amazon soils and are certainly of a morphological kind that challenges western notions of agricultural productivity, but they are rich in minerals and the magic of gold. To the south such contrasts are repeated in the hydrological character of the south flowing rivers that drain the hard old granitic heart of the Guiana Shield, which are filled with deep passages and narrow rapids, unlike the meandering and broad water courses that flow northwards into the Amazon channel. The deep green forest of the northern flanks of the Pakaraima mountains which form the summit or the Guyanese Shield, along with the Parima chain, also contrasts strongly with the savannas that cover the southern flanks.

7Rooted in these features of the physical landscape the idea of Guayana as insular is then understood, in both colonial and native thought, as distinct from its southern counterpart «Brazil». The Amazon river channel is thus a frontier not a center in such continental systems. The «civilized» chiefdoms to the south and west along the river, such as Tapojoso or Oniguayal are then in opposition to those of the north bank, such as Arripuna, Conori or Manoa which become the wild and exotic lands of Cannibals, Amazons and Gold in the early colonial accounts (Whitehead, 1994).

8However, it is the emergence of Guayana as a demonic landscape, incompletely imagined through the cannibal, that is the subject of this essay. Thus, the zone of the cannibal lies to the north and south of Guayana but is never clearly located within it, and then only as an invasive phenomenon. Cannibals in the Caribbean, the original site of this peculiarly American signifier, border Guayana to the north. Ancient Cannibals were hypothesized to have brought their ritual proclivities to the shore of the Barima river as part of a wider invasive moment. The excavation of shell-middens in this region by the missionary William Brett (1851, 1881), during which human bones were recovered, apparently anchored this hypothesis with the evidence of archaeology.

9Nineteenth century colonial commentators thus constructed Guayana as resistant to this cultural proclivity until the demon Caribs, plundering and killing their way through the Carribean islands and delta of the Orinoco, inflicted their warrior disciplines on the indolent indigenes, transforming the local Karinya from farmers of manioc into the «red faces» of Warao legend who ate up the peoples of the Orinoco and interior. This was in itself but a mimesis of a European occupation that had already destroyed the majority of native cultures and had consumed their remnants in the pursuit of colonial development. Firstly, through the system of plantation slavery that induced a dependency on the colonial regime through its payment for mercenary services in policing the interior for runaway slaves. Secondly, through the punitive slaving of certain Guayanese groups themselves, particularly the Makushi and other groups of the deep interior. Slaving then itself became a way of knowing and discovering and slipped easily into evangelization as the economic stabilization of the colonial regimes in the region through the plantation economy supplanted the haphazard prior dependence on individual trade partnerships with native groups.

10Likewise to the south, those gourmet cannibals the Tupi, that so fascinated the French commentators, were singular to their Brazilian location. Although early French ethnology tended to use categories derived from this Tupi encounter to describe the cultures of others in the Caribbean and even the land of the Iroquois, the attempt to implant such notions in Guayana, or to filter their experience there through this Tupian lens, resulted only in various infelicities in ethnological description. It proved all to facile to descry the ethnic dyads of Caribbean experience (Arawak vs Carib), in the dyadic logic of cannibalism and war among their allies in the region of Rio de Janeiro. In this way Guayana eluded the wider continental categories of conquest and knowledge such that native cultures in Guayana become the marginalia of evolving ethnological schema in the nineteenth century travelogues and the twentieth century Handbook of South American Indians (Steward, 1948).

A Demonology of Development

11The landscape is not a passive component of human socio-cultural processes and so the way in which the colonial regimes of the region occupied and sought to re-inscribe their own particular cultural meanings in the context of existing native practices has to be considered alongside the imaginary cartographies of the demonic that orientate colonial and national governments in their actions towards, and creation of, the indigenous (Whitehead, 2003).

12These constructions of the «native» or «indigenous», whether in public political discourse, the representations of the media, or in anthropological and ethnographic writing, have all served to enable and encourage violence against indigenous communities. From the initial charges of cannibalism made against Amazonian peoples in the sixteenth century to the supposed fierceness of the Yanomamö, there has been a continuous external discourse on Native Amazonian’s savagery and wildness, most usually demonstrated through their Satanic proclivities (shamanism) and demonic customs (cannibalism). No less relevant to the fate of native peoples today than it was five hundred years ago, this discursive production of «natives» continues to create a broad cultural framework in which violence against indigenous persons can be more easily obscured or justified.

13In Guayana successive waves of spiritual and material development have pounded native communities leading to the continuous ideological construction of indigenous peoples as obstacles to «progress». This is signaled by key cultural practices that have allowed the governmental regimes to separate out the «good» and the «bad» Amerindian. In Guayana this has proceeded along an number of axes which all reference spiritual and ritual forms of action as ciphers for the political opposition that given native groups showed towards theses plans for their redemption and development. Most notorious was the dualistic ethnic typology, originated by Columbus, of Arawaks and Caribs. I have shown elsewhere (Whitehead, 1995, 2002c) that this apparently objective ethno-linguistic distinction is in fact highly suspect and refers more to the exigencies of conquest than to native realities. However, this distinction became widespread across the region and in time a component of Amerindian identity itself. Colonial policy enshrined this scheme in legal statute - allowing plunder and slaving of those populations considered Carib. Arawak populations, both through selective alliance and the involvement of key Lokono clan leaders, were then understood as basically tractable and, given a system of raiding and warfare between key Carib and Arawak populations, such as the Karinya and the Lokono, such Arawaks were also considered to be self-interested, and so dependable, in their European alliances.

14As missionaries began systematic evangelization in the seventeenth century along the Orinoco, this ready made distinction became self-fulfilling since opposition to the missionaries was defined as Carib. In this way the Arawak, who accepted evangelization, were again seen as favorable to colonial development. But this distinction was, and still is, based on more than these competing representations of Amerindian tractability and intractability. The notions of Carib and Arawak refer also to spatial location and ritual proclivity. Consistent with the demonic nature of the colonial imagination of the space of Guayana, the Carib are pictured as interior, bush-dwellers exemplifying the secretive, dangerous and violent nature of the dark heart of the region. The Arawaks, coastal dwellers and even urbanites, in turn signal the possibility of indigenous redemption, reform and development. These competing tropes of the indigenous can then be made to fit varying political and ethnic circumstance using linguistics to bolster the idea of a fundamental difference in the cultural ontologies of Arawaks and Caribs. Despite the fact that linguistic practices do not actually conform to this scheme it has remained a passionate debate in contemporary anthropology and archaeology (Whitehead, 2002c).

15Linguistics thus anchors a demonology of development in the ethnology of the region, but this demonology is not just a linguistic distinction correlated with a political opposition but also invokes apocalyptic visions of a threat to all social order, native, colonial or national, through the linguistically inscribed association of Caribs and cannibalism, the very terms being directly related etymologically through European usages. The production of cannibals was thus an economic and political interest, given the legal provisions that allowed special violence against them. It is also a central issue for anthropological interpretation, since, as Arens (1979) rightly indicates, this is the original anthropological question. As a result the literary and ethnological production of the cannibal has gone hand in hand with the military and political domination of the native population in this region of South America. The violence of conquest in the region mimetically referenced cannibalism as its justification, and representations of the native population suppressed description of Arawak torture and cannibalism, emphasizing rather the barbarity of the Carib.

16It is therefore important to realize that in other contexts cannibalism per se simply did not function in this way as a justification for colonial or national violence against indigenous communities. The French commentaries on the Tupi of coastal Brazil from the sixteenth century thus also make a distinction between the good and bad Indian but do so with reference to the ritualization of the cannibal rite (Whitehead, 2000). This resulted from the way in which the encounter with the Tupi provoked theological debate over the nature of the Catholic Eucharist and its doctrine of transubstantiation. In this scheme the elaboration of cannibal ritual, alleged amongst the allies of the French, and denied among the allies of the Portuguese, served to produce political and military distinctions among the native population. French collusion in the cannibalism of Portuguese prisoners of war should thus be understood as a way of sustaining a distinct kind of ethno-demonology, even if it was also founded on the cannibal sign. In this way the very ritual practice that, to the north, was considered a defining characteristic, in any form, of a recalcitrance and hostility to colonial and national development, becomes an analogy of the Christian Eucharist in Brazil. The violence of the cannibal ritual was not problematic, but rather the nature of collective participation in that ritual violence. The participation of women and children implicitly undermined the colonial state’s political justifications for its violent modes of control, the redemption of the heathen and rescue of the cannibal victim.

17In the French sources (see also Lestringant, 1977) the meaning and form of the Christian Eucharist is of greatest significance and is the starting point for Montaigne’s consideration, «De Les Canibales» (in Essais, 1580). For Montaigne, both Tupi and Christian rules of cannibalism are communal, related to the worship of the dead and done in the hope of benefit to the group. But the key question for the differentiation of Catholic and Protestant liturgy then becomes - is this a literal homology present in the moment of transubstantiation or an analogy of spiritual nourishment that replenishes and feeds faith indefinitely? For Montaigne, as a Catholic, the assimilation of the Tupi practice to the Christian belief is an obstacle to the humanistic embrace of the Tupi. However, Montaigne fails finally to condemn the Tupi precisely because of the parallels in symbolic practice that other contemporary sources explicitly elaborate.

18Without the possible religious significance of the parallels between Tupian and Christian Eucharist the subsequent fate of the cannibal is to be superfluous and so anachronistic, fit only for eradication, as with that «exterminating angel» Robinson Crusoe (Lestringant, 1997: 11). And so the cannibals come to be seen as responding to their bio-ecological conditions not their culture, or, in the act of erasure that Lestringant (1977: 12) rightly resists, the materiality of their acts are supplanted by the evanescence of a cultural discourse which «shifts the noise of teeth and lips towards the domain of language».

19André Thevet, in his earliest work on Brazil, Les Singularitez de la france Antarctique (1558), delineated a geography of man-eating that located the uncultured «cannibals» to the north towards the Amazon river, and the ritualized «anthropophages» in the orbit of French experience. Thus the former are characterized as cruel eaters of human flesh as a matter of diet, the latter as exponents of certain rituals of revenge. Not surprisingly this cannibal cosmography also conforms to the patterns of French trading and military alliances in the region (Whitehead, 1993a).

20Thevet notes the progress of the cannibal rite in great detail and seems unable to resist an analogy with the preparation of roasting pig in French peasant culinary traditions. However, elaborate though this ritual becomes in Thevet’s subsequent account in the Cosmographie (1575) and a manuscript work Histoire de... deux voyages, it is clear that vengeance is the hermeneutic key to understanding the meaning of the cannibal act, such that description of the careful distribution of the victims body parts and the embedding of the ritual in myth, becomes central to these later works.

22For Léry cannibalism can also symbolize cruelty, usury and a lack of charity and, as the meaning of the cannibal sign universalizes, so debate over the Eucharist bloats Léry’s account of the Tupi materials to the point that even Villegagnon, leader of the French colony of Rio de Janeiro during Léry’s time there, is identified on account of his Catholic views as a soi-disant Ouetaca. So if the Tupian cannibal rite recapitulates at all points the Christian Eucharist, to prevent his readers turning away in ethnocentric disgust, Léry goes on to detail the unspeakable cruelties of European tyrants - including Dracula, by way of Vlad the Impaler. In this way Jean de Léry de-territorializes the cannibal who thus freed from the Brazil shore haunts the European and American imaginary (Lestringant, 1997: 71). In the same way as cannibalism, kanaimà is disembodied in nineteenth century ethnological writing through a refusal to accept the materiality assault sorcery, the violence being seen as vengeful not mystical. This allowed a safe occupation of the demonic landscape from which such ghoulish remnants of native savagery have finally been exorcized.

23By the end of the eighteenth century, however, as is evident from the French sources themselves, the «cannibal» had begun to fade away into that mist of nostalgia and remorse for the pre-modern. The conquest and control of native societies had led to the virtual extermination of most autonomous native polities, leaving only relatively isolated remnants in the deep interior, or the emergent neoteric groups grounded in the social and cultural relations of the colonial world. In either case the spiritual practice of cannibalism, directly connected to the creation of political power, was therefore itself in decline. The previous ritual sponsors of such events were now engaged in the politics of incorporation within the burgeoning national societies of the region.

24As a result the violence and barbarity of the intractable Indian became occluded and hidden, a matter of inner orientation and belief, no longer an aspect of public culture. Thus, with the suppression of native warfare, and the loss of autonomy that implied, the uncontrolled and unknowable realm of spiritual and mystical assault emerges as the site of demonization. Whereas before the warrior or cannibal-killer was the object of colonial nightmare, in the new world of nineteenth century progress the infrastructure of native autonomy was absent and so the mimesis of conquest becomes the skulking assassin, the vengeful and lone killer. In the demon landscape of Guayana the prominence given to the shamanic practices in the colonial literature, particularly kanaimà, also begins at this precise moment.

25Kanaimà, depicted as vengeful assassination in the colonial literature, has subsequently become widespread in the cultural imagination of both national and indigenous populations of northeastern Amazonia as a violent marker of both «tradition» and «indigeneity». This process begins in the mid-nineteenth century as scientific explorers and missionaries wrote reports and travelogues of the relatively unknown interior peoples.

26But, in the imagination of the Patamuna and other peoples of the highlands today, kanaimà is seen as emanating from the first time - a primordial force that has structured the universe and formed the world as we know it. Kanaimà is more ancient than warfare and more ancient than society itself. As a result kanaimà itself almost becomes outside history, or at least it is a universal and constant human possibility that is only contingently realized in the course of human history. In this way one may say that, while there is a history to the ritual practice of kanaimà, kanaimà as a way of being in the world is beyond time.

27The temporal and spatial dislocation of kanaimà is therefore an integral part of its ritual and symbolic force, for practitioners are both able to travel vast distances in a short space of time and are able to expand their life-time and life-space by the means of the ritual practices that are involved in being kanaimà - that is they have access to the plane of shamanic encounter and exploration (karawali), as well as being directly nourished by the production and consumption of the ‘ honey’of necrosis (maba), derived from the bodies of their victims.

28In the early nineteenth century colonial sources assimilated kanaimà to notions of revenge more generally. As a result any instance of revenge became one of kanaimà. But this assimilation of kanaimà to all revenge killings meant that a vast array of different kinds of death were attributed to kanaimà when they may not have been ritual killings at all. Thus, as more systematic ethnographic observation begun, and as the native discourse of kanaimà certainly would have raised the possibility of action by kanaimàs in all manner of deaths and injuries, it could appear that a kanaimà killing was no more than a vague attribution of sickness and injury to an all encompassing evil force. Divested of its corporeal reality in native cultural practice, the symbolic force of kanaimà was then mimetically assimilated to the power of the colonial state itself.

29In the external meanings and representations of kanaimà it will be apparent that this latter idea, of kanaimà as a native «Satan» or evil spirit, was central to the ethnological and literary production of kanaimà. At the same time there was a symbolic convergence and mimetic elaboration of the category «kanaimà» in both native and colonial discourses (Whitehead, 1997b). This entailed that the wider cultural perception of kanaimà was as an example of the savagery of the native population. Nonetheless, kanaimà eventually transmutes in the ethnological literature into a functionalist mechanism for sustaining social order. The maiming and the killing become a kind of ‘jungle justice’ which, in the absence of the Hobbesian state and its police, regulates and punishes the «criminal». As a result the indigenous symbolic and ritual meanings of kanaimà were effectively overwritten or ignored, being understood as no more than an unfortunate welling up of latent Amerindian superstition and savagery.

30The colonial gaze that produced this ethnological and literary view of kanaimà was certainly entranced by its violence, but kanaimà practice and its meaning was not properly understood. This was as much a limitation of the kinds of interaction that missionaries and others might have had with the interior peoples, as it was a conscious attempt to misrepresent what otherwise was evident. Even if the cultural importance of kanaimà violence was vividly represented as the «oppressive nightmare» of the highland peoples, the cosmological significance of kanaimà as a shamanic practice was persistently overlooked.

31Ethnological representation always emphasized apparent juridical features to kanaimà, picturing it as a codified system of revenge - a lex talionsis. This resulted principally from a failure to discriminate the various forms of shamanism and magical action. Although the piya (curing shaman) was seen as the key figure standing in the way of rapid and complete missionary conversion, in fact only a limited understanding of such «weedy entanglements of evil» was required for its suppression. The highly physical nature of kanaimà killing and its connection to notions of revenge, or more exactly the rebounding violence of a challenge to social norms, therefore obscured its occult meanings for the colonial commentator. In this way kanaimà served as an answer to the puzzle of how native society might regulate itself, in the absence of formal institutions of law or criminal justice. As Thomas (1994: 127) has pointed out with regard to British missionary efforts elsewhere, each mission «field» was apt to dramatize certain key cultural practices - such as head-hunting, cannibalism, or widow-burning - in order to provide an index of growing evangelical success. As a result these kinds of representation tend also to present the cultural practice in question as on the verge of extinction. This was very much how shamanism generally, including both piya and kanaimà, were presented.

32Nonetheless kanaimà proved sufficiently elusive that it could not become a basis for the campaigns of cultural extirpation that are associated with other dramatized cultural practices, such as cannibalism. This elusiveness, coupled with pervasive native reference to kanaimà, meant that kanaimà still required some codification and so the legalistic interpretation was pursued by the major missionary and ethnological writers - William Brett (1868, 1851), Everard Im Thurn (1883), Walter Roth (1915) and John Gillin (1936).

33The ferocity and horrific mutilations of a kanaimà attack nonetheless were impossible to ignore but they were seen as an almost laudable aspect of the rigor of native «justice», provoking some authors to favorable comparison with «tribal» Saxon and Judaic ideas of retaliation and recompense. The profound spirituality of kanaimà violence was in turn erased through its presentation as a folkloric belief, of uncertain corporeal reality, akin to the vampire or werewolf.

34The beneficial effects of kanaimà in producing social order were identified in both its proto-legal functions and its supposed consequences for sustaining social distance and the particularity of ‘tribal’ identity. This analysis is given persistent emphasis, to the point that kanaimà becomes a particular instance of legal systems founded on retribution and punishment; the only distinction with western society being that it is the retributive force of government and society, rather than that of the clan or family, which is being expressed. In the twentieth century imagining of kanaimà this notion has been taken to its conclusion and, as the State supplants Religion, so kanaimà becomes a particular instance of the way in which the «capriciousness» of the cosmos may be represented (Simpson, 1964: 4-7).

35Unlike sixteenth century Spanish attempts to delineate American «cannibalism» or nineteenth and twentieth century British attempts to determine the «reality» of African witchcraft, kanaimà was not the subject of «official investigation». However, as in Evans-Pritchard’s (1937) classic work on Azande sorcery, the ethnology of kanaimà shamanism ‘reveals’ itself as a system of justice and legality. In this way the cultural force of kanaimà, that so manifestly entrances and troubles the colonial imagination, is shown to be a matter of the ineffable and irrational nature of colonial subjects. That primitive mystery is then made less threatening through the «science» of ethnological description which sanitizes sorcery through its selective re-presentation in familiar, intelligible and comforting terms - as a form of law and order.

36This literary and observational process is underwritten also by a high degree of inter-textuality among the colonial authors. Certain key accounts, especially those of the Schomburgks, are constantly recycled thereby progressively constricting the interpretive space for subsequent new descriptions or information. As a result, by the end of the nineteenth century, the colonial imagining of kanaimà becomes embalmed in the textual record and later accounts have largely failed to escape this rendition.

37In sum, kanaimà was clearly misrepresented in the colonial sources and that was done in ways that enabled the progress of colonial administration, especially through missionary evangelism but also through the imposition of colonial legal codes. It is notable that a number of the authors depict kanaimà as a institution of primitive law thereby also laying the ground-work for a later appeal to colonial justice as an advance on this primitive, if somewhat admirable, law of blood-revenge.

38In indigenous counter-representation, tales of whites killed by kanaimà and the possibility that whites might be susceptible to the physical mutilations, though not the occult meanings, of kanaimà attack are to the fore. But such threatening meanings are averted in colonial mentality by the reassuring suggestion that the «white man» might act as a final court of appeal in the jungle justice of the blood feud. This ethnological representation also allows the «naturalization» of retributive law and capital punishment, so that kanaimà aids the cultural project of government in the metropolitan as much as the colonial setting. At the same time, this very focus on piya and kanaimà shamanism signaled the significance of spiritual practice for the colonial regime and so encouraged the expression of resistance or opposition in these modes. This attitude in itself gave an external importance to kanaimà that has fed into its current resurgence in native communities.

39Kanaimà practice itself has become closely attuned to the violence of the development frontier and thus symbolically, as well as materially, directly engages with this external discourse of savagery. Modernity, development and kanaimà violence are all related in current Patamuna thinking. On the one hand they are keenly aware of the way in which development, both now and in the past, seems to have passed them by, eluded them and made necessary efforts, to go and «fetch it up» from Georgetown. On the other hand, and conscious of the power of anthropology as the arbiter of «culture», the Patamuna are also keenly aware of the potential for a loss of tradition, a distinct way of «being Amerindian», which, paradoxically perhaps, the ritual skills of the kanaimà or piya - death or resurrection - best express. In the face of modernity kanaimà becomes a potent symbol of continuity with the past.

40For us too, «modernity» is a much debated concept and, by way of definitions it needs to be said that the idea of «modernity» employed here is one that emerges only in opposition to the notion of «tradition». In other words I do not think that ‘modernity’ can be understood apart from those social and cultural processes which, in any particular situation, also give meaning and content to the idea of «tradition». In this sense «modernity» is both ancient and plural, an aspect of the continuous construction of tradition. The meaning of these categories is therefore closely interrelated and historically contingent, the mutual condition of their possibility.

41As a result local ideas of the modern and traditional may not be heavily inflected by the original Western content of such globalized notions, and there has not been that «convergence» of modernities in the manner suggested by both Marx and Durkheim, to produce a global social and cultural uniformity, but rather an explosion of alternative modernities.

42Consistent with this view experience of modernity in Guyana has largely been in terms of the spread of «governmentality» - the apparatus of State by which we are all rendered «citizens» (Foucault, 1994). However, despite being repeatedly surveyed, classified and converted, the systems of law, education, sanitation and economy which were supposed to bring the Patamuna the fruits of development and progress have failed to materialize. As in other colonial and post-colonial contexts, industrial capital and western democracy have appeared wearing the differing costumes and masks of evangelical redemption, medical services, schooling, economic development, democratic rights - the full regalia of modernity, as it were. In the case of the Patamuna, and other peoples of the highlands, the experiences of such successive modernities has been highly episodic and fleeting, a series of one-night shows, short runs and rapidly folding productions. Nonetheless, the fact that the theater of modernity has been trying to establish itself in the highlands over the last two centuries means that Patamuna conceptions of the traditional and the modern are more complex and sophisticated than a simple opposition of, say, feathers and loincloths to trousers and shirts.

43In response to this repeated failure of modernity to establish itself a «hypertraditionality» has emerged that is inimical to both external and indigenous notions of modernity. This persistence of a notion of modernity is manifested in a cultural consensus, shared by most Amerindians and other Guyanese alike, as to the relevance of these convergent, globalized ideas of tradition and modernity, those «costumes and masks» referred to above.

44This by no means entails that there is any consensus over how those shared conceptions might be realized or enacted, but it does underline their coevalness. The cultural and political trajectory of kanaimà in the face of these multiple modernities in all their scene and costume changes has been violent resistance.

45This needs to be understood as a persistent enactment of a notion of ‘tradition’ rather than some active counter program of «modernity». This resistance has become more potent as the idea of «tradition», especially «authentic», «pre-modern» tradition as often derived from Amazonia itself, has become widespread in modernist society. Given the anti-modernist rhetoric of much of the «new age» interest in shamanic tradition, perhaps we should after all consider kanaimà a truly post-modern development; an alternative traditionality in response to an indigenous modernity. Either way designer tribalism and ethnic chic are well established cultural tropes for us, and the production of tradition among people like the Patamuna, is likewise an issue of cultural persistence and affirmation.

46The external projection of kanaimà as a cultural tradition and its meaning for practitioners and their victims have therefore become increasingly entwined. As a result there has been a notable literary use of the idea of kanaimà, from the missionary William Brett’s (1880: 152-54) poetical work, Kànaima (see Appendix I), to Rómulo Gallegos’s (1935) novel Canaima, as well as Wilson Harris’s (1995) short story, but the symbolic force of kanaimà, potent enough within an indigenous poetic of violent death and already reflected in these mimetic cultural productions of outsiders, continues to haunt the modernist cultural imagination.

47In this sense the kanaimà takes his place alongside the «thuggee», «bogeyman», «cannibal», «head-hunter», «zombie», «vampire», and now «terrorist» as another ghoul in the colonial and modern nightmare of irrational, cultic violence that springs from enigmatic and atavistic cultural proclivities. This usage occurs in more sensationalist presentations, such as the pulp-novel Skull of Kanaima, by Victor Norwood and other recent materials such as Brazilian role-playing games, that feature «Kanaimà» as a game-player identity, or the Ukrainian Death Metal band, Tessaract, that features a song Black Kanaimà (see Appendix II). However, our reactions are not restricted to the cultural lexicon of colonialism or modernism and the anti-modernist search for other kinds of meaning can represent the kanaimà as a figure that is inspiring in virtue of its very cultural opacity, as in the case of the poetry of Pascale Petit (1998).

48In these ways scientific tropes of evolution, literary metaphors of wildness and the anthropological categories of culture and society have eased the assaults of development by providing an intellectual framework for the comprehension and control of the native. This program is then made manifest in the polices of development which become a material realization of this imagery of indigeneity. However, the practice of kanaimà constricts and chokes off the space for material and spiritual redemption through progress, for it is a counter discourse on modernity in which the promise of development is violently rejected in favor of a new kind of tradition. The value of this tradition is precisely the way in which it eludes external attempts to classify and analyze native society - for it is beyond the visible and material, as well as of them.

49Such material relations of representation in turn recreate native and colonial or national realities, such that these representational practices achieve an ontological status as ethnological «fact», evidenced in the cultural practice of development and the counter-performance of a kanaimà killing. To be brutally short, representation is not just «about» violence but is also part of violence.