ETHNOTERRITORIAL RECONFIGURATIONS OFSOCIAL CONFLICTS: FROM CULTURAL DIFFERENCE TO POLITICAL FRAGMENTATION (BRAZILIAN AMAZONIA)

AS RECONFIGURAÇÕES ETNO-TERRITORIAIS DOS CONFLITOS SOCIAIS: DA DIFERENCIAÇÃO CULTURAL À FRAGMENTAÇÃO POLÍTICA (AMAZÔNIA BRASILEIRA)

Abstract

Despite the constitutional inscription of the territorial rights of indigenous and quilombola peoples, a new kind of conflict has emerged between differently classified populations. This article documents this conflict through the study of a disagreement between 'quilombola' and 'indigenous' populations in central Amazonia. The solutions of various actors reveal doubts concerning the real 'ethnic contrast' between the protagonists. The current organisation in ethnic territories is seen as the most recent instance of a series of socioterritorial reorganisations already adopted by the populations in interaction with diverse bodies. The article reconstructs the shifting alliances between opposed factions, highlighting the modalities involved in their formation. When the rights granted by the State to groups of populations due to their exceptionality come under threat, as at present, it is essential to recognise that the fluidity between legal categories reflects their shared struggle for survival.

Keywords
Indigenous; quilombola; conflict; Amazonia; rights

Resumo

Apesar da inscrição constitucional dos direitos territoriais de povos indígenas e quilombolas, surgiu entre populações classificadas de modo diferente um novo tipo de conflito que o artigo documenta a partir do estudo de desacordo entre "quilombolas" e "indígenas" na Amazônia central. As soluções dos variados atores desvelam dúvidas acerca do efetivo "contraste étnico" entre os protagonistas. O alinhamento da atual organização em territórios étnicos é compreendido como uma última atualização de reordenamentos socioterritoriais já adotados pelas populações na interação com diversas instâncias. O artigo busca restituir o jogo de alianças entre facções opostas realçando as modalidades de sua formação. Quando os direitos concedidos pelo Estado a grupos de populações em virtude de sua excepcionalidade estão ameaçados de retrocesso, é imperativo reafirmar que a fluidez entre as categorias legais remete a luta compartilhada para sua sobrevivência.

Palavras-chave
Índigenas; quilombolas; conflitos; Amazônia; direitos

As an extensive literature has already shown, in terms of its organisation, space can be subject to different and even conflicting appropriations simultaneously (for example, Lefebvre, 1974; Vanier, 2009; Miller, 2011), situations where the coexistence of multiple readings of a specific geographic area turns into competition when one of these claims greater legitimacy than the others - as in the case of the sectorizations promoted by the State. Amazonian ethnology provides some striking examples of such divergences. Dominique Tilkin Gallois (2004: 39) calls attention to the tensions between the Brazilian judicial concept of a Terra Indigena (Indigenous Land: IL) with clearly established boundaries and the contiguous extension for the surface area of the land (terra) and the local conceptions of territory forming the basis of a specific society in which spatial discontinuities become conceivable. Contact, though, does not just instil the idea of a closed territory. As Gallois stresses, the process of land regularization generates a "context also favourable to the emergence of an ethnic identity," since, as she illustrates, it was during the process of demarcating their IL that the Wajãpi moved from "a non-centralized self-representation (without ethnic connotation)" to an "ethnic self-representation" reflected in the formation of the category "we Wajãpi." Similarly, Pascale de Robert (2004: 80) observes that during the process of fabricating a territory that they themselves call a 'cut land,' the Kaiapó ended up placing more emphasis on the border between 'us' and 'them,' the 'whites,' than on the internal differentiations that structure the world 'with each other.' The anthropologist adds that the process of demarcating the IL entailed the need for the Kaiapó to overcome their internal disagreements and present a unified image to the world, thereby readjusting their previous relations with space as part of a new kind of belonging.

The reader is presented with two important considerations, therefore. First, the agents and populations involved in these negotiations find themselves in an unequal power relation, forced to adapt to the State's regulations and labels. Second, the need imposed to learn the language of the State leads to the redefinition not only of the territory but also of the group itself. In the Wajãpi case, Gallois (2004: 70) writes, there was a "transformation of ethnic identity into a territorial marker."

Discussing the Brazilian Northeast, Jan Hoffmann French points to an analogous correlation between the emergence of a closed representation of the territory and the emergence of a distinct ethnic consciousness. In a thought-provoking analysis of populations who were embedded in the same kinship networks but took divergent paths in terms of identity - some becoming Xocó Indians at the end of the 1970s, others recognizing themselves as quilombolas (residents of quilombos, communities originally formed by escaped slaves) in the 2000s - French emphasizes, like Gallois and Robert, the interdependence between territorialization processes and the dynamics of emergent ethnicity: although they were kin, "the demand for land [by the Xocó Indians and by the quilombolas] simultaneously awoke, and was driven by, new assertions of cultural specificity" (French, 2009: 123).

This co-construction of the boundaries of a territory and a group occurs today within a legal framework renewed by the 1988 Constitution, which generated numerous expectations and, subsequently, many frustrations, given that the State's adoption of a multipolar approach to processing territorial demands - with FUNAI responsible for indigenous peoples, INCRA for quilombolas, and ICMBio for traditional populations - had at least two additional unforeseen consequences: rivalries between state institutions vying for control over the delimitation of their areas of work (Castro, 2012) and conflicts between populations that reformulated pre-existing local feuds in ethnic terms.² In their comparative study of two situations of territorial overlapping in the Amazonian region (Arara do Rio Amônia IL and Alto Juruá Extractivist Reserve in Acre, Escrivão IL and Tapajós-Arapiuns Extractivist Reserve in the west of Pará), Roberto Sanches Rezende and Augusto Postigo (2013: 126-127) argue that "the division of representativist institutions, based on ethnic profiles, enabled actors involved in pre-existing disputes in the communities to find institutional spaces where they could reproduce their conflicts."³ Moreover, the authors call attention to the fact that these locally observable tensions, combined with the competition between federal agencies, also have repercussions at an intermediate level, namely the level of the region and the various 'movements': "This projection of community conflicts within a regional institutional setting allowed them [...] to be read solely in terms of ethnic opposition, enabling each community dispute to also become a strategic dispute for the regional indigenous movement and for their opponents" (127).

In a way, all these discussions can be situated within a broader debate, identified by André Dumans Guedes (2016: 24), concerning the transition between demands for 'lands' and demands for 'territory' that illustrate processes of 'ethnogenesis.' In a thought-provoking article in which he analyses the diverse range of definitions and articulations mobilized by anthropologists vis-à-vis the two terms - including the territory as a more effective form of resistance to agribusiness than the struggle for land, and relating to a more complex reality − (25-26), the author opportunely recalls the profound divergence between the 'radical instrumentalist' and 'primordialist' approaches. He then proceeds to argue convincingly that among some populations the identification of particularities corresponds above all to the concerns of researchers within a given historical period, rather than the presence or absence of these traits. In the 1950s and 1960s

[t]he very definition of these groups, populations or persons as camponeses [peasants] emerged from the articulation of political and academic questions that involved foregrounding certain themes and topics with the potential to encompass heterogenous realities and universes [...] [By contrast] the current struggle of traditional communities manifests a preeminent need to intellectually and politically privilege certain aspects and traits [...] previously seldom paid much attention by intellectuals and social movements (Guedes, 2016: 28-29).

It was in this context that the territory became an indispensable element for "encompassing and evincing particularities and specific identities" (Guedes, 2016: 29).

Assimilating the proposal to conceptualise the territory as a political form and project (Guedes, 2016: 32), I wish to contribute to the discussion of ethnoterritorial configurations through the analysis of a conflict in a particular region of central Amazonia involving geographically proximate localities that are nonetheless assigned different ethnolegal categories. The case strikes me as interesting for three reasons. First the mobilization of an ethnic language to enunciate older antagonisms is a recent phenomenon. In fact, until the 2000s the povoados (small settlements) that today categorically insist on their 'differences' all identify themselves as moradores comuns, 'common residents,' an expression preferred to the term caboclo, which urban inhabitants used to refer to them. Second, their study has been helping to fill various ethnographic lacunas. Indeed, while the conflicts between indigenous and traditional populations have now begun to be documented,⁴ the case examined here is a rare occurrence in which collectives that affirm their identity as indigenous clash with others proclaiming themselves as quilombola. Finally this ethnographic example especially allows us to apprehend the sociopolitical and relational dimension through which territories are produced, including the fact that the insertion in networks of kinship and/or activism initially counts more than the discourse on the ethnic origins and the 'ancestral' occupation of land. Over the course of my field research, what appeared initially as a one-off and limited disagreement between two villages was shown to be just the latest development in a series of conflicts, all expressed in terms of ethnic opposition, encompassing various other localities in this geographic area. In this wide-ranging and complex interplay of forces, including the intervention of diverse outside actors, institutional and otherwise, and operating at various scales, the nearby settlements, including those yet to declare any specific ethnic identity, are invited to (re)position themselves. The size and boundaries of the lands whose official recognition is demanded from the State depend on this restructuring of the local political field.

This approach invites a relativization of the idea of the primacy of the territorial dimension stricto sensu (i.e. with pre-established limits that the State is content to make official) by paying more attention to the dynamics of negotiation and the forming of intercommunity alliances, which, in turn, determine the strength of the factions present and their capacity to advance their land claims. In the first part, I present the conflict as I found it during my first stay in the region in 2011. We shall see how the diverse solutions (all unfavourable) introduced by a wide range of outside actors exposed doubts over the effective 'ethnic contrast' between the protagonists. Seeking next to place the current organisation in ethnic territories into perspective, I expand the historical focus of analysis to show that it amounts to just the latest in a series of formal organisational modes already adopted by the populations in their interaction with diverse state and religious bodies. This is also reflected in the diverse range of denominations given to the population clusters: community, residents association, indigenous association, quilombola association. In the third part, the article considers a geographically broader space, covering ten nearby settlements that claim different ethnic identities. Aiming to reconstruct the dynamic of alliance formation and the modalities informing the production of opposing factions, I present the sequences involved in the constitution of the two main identity constructs, highlighting their flexibility and the fact that the public identities assumed by the localities are neither the central motif nor an obstacle to inviting the participation of indigenous or quilombola mobilizations. In the conclusion, I return to the sudden transition from a situation characterized by the apparent organizational heterogeneity of the populations to another in which the latter seek to outwardly present an image of cohesive and uniform blocs.

Before proceeding further, I should cite a problem that emerged at while I was writing - based on the same ethnographic material - an earlier article on the construction of contrastive ethnic arguments and the 'cultural' dimension of social reconfigurations (Boyer, 2015): namely the issue of anonymity in an ethnographic situation, the need for such as a result of clear political motives and the difficulties that this approach generates for ethnographic description. In this specific Amazonian case, the local demands for recognition, one presented to FUNAI, the other to INCRA (its quilombola sector), were still being processed by the institutions, and the leaking of information threated to exacerbate an already latent conflict. Hence it was impossible to name places or people. Seeking to prevent any possibility of identification, I decided to invent names for the settlements and even change the ethnonym by which the indigenous population were making their claim, replacing it with another evoking a region in which it is public knowledge that no claim from black communities exists and thus no potential conflict between indigenous and quilombola populations. As far as I can tell, this strategy did not dilute the emphasis given to the common logics by which native discourses adapt to legal categories, despite their distinctivity. Nonetheless, in an article that, like the present, aims to contribute to the theme of 'territorial conflicts,' everything becomes a little more complicated a priori. Although the reconstruction of particular social processes does not exclude abstraction and inventing names, the discussion of a dispute over territorial boundaries involves explaining arguments based on references to the area's natural resources, the geographical elements that characterize it, and the neighbouring towns frequented by local residents for diverse reasons. Indeed, it is by indicating the particularities of a concrete space that social collectives are able to appropriate and transform it into a territory. As well as prioritizing one of the diverse possible readings of a space, any divergent positionings also, as we shall see, reflect the inclusion in networks of specific dialogue with diverse outside actors. Now, while the mention of the names of federal agencies or missionary orders working locally has little impact on anonymity, this is not the case of the names of entities working at regional or microregional level, like federations or NGOs. Despite the difficulties involved in transposing their localization in order to protect the case under examination, here I have chosen to maintain the fictional system already adopted, altering the name of people, places and entities that have worked or still currently work in the area, as well as omitting details that are not absolutely necessary to understanding the context. The dates mentioned and the sequencing of events, however, are as faithful as possible.

THE CONFLICT IN THE PRESENT (2010-2015)

In this portion of the Amazonian space in which 'quilombola' and 'indigenous' populations confront each other, domestic groups are related by extensive networks of kinship and godparenting, and share various forms of leisure (like football tournaments and dancing festivals) and political activities (joint mobilization to press the local council for road maintenance), buy hammocks, chairs, pans or towels from the same traveling vendors, and most receive the government family allowance (bolsa família). Whether used for domestic consumption or sold to wholesalers, many extract assai and other products and some men practice artisanal fishing and/or hunting. Notably, while each locality has its own school, association, bar and artesian wells (some installed by the local council, others by kin groups), some of them are covered by the same healthcare agent and the same bus lines, irrespective of the legal identity adopted. Their profiles are also similar in the religious sphere: Catholicism predominates, despite some people joining Evangelical churches, and many knowing herbs and prayers for curing.

In terms of autonomy and community life, there are, however, clear differences between the quilombola and indigenous populations. Though people in the aldeais (indigenous villages) sometimes provide services to outsiders or to wealthier relatives, everyone has a swidden with crops of maize, beans, manioc and so on, and keep small numbers of chickens and pigs in their yard; the women usually work together in the flour mills (casas de farinha) built and used by a group of kin, where they process the manioc; just one resident has a car, used for freight transportation. In the quilombola communities, though, wage labour or day work is much more common and most men have a motorbike for everyday commuting to the town, about 40 kilometres away. In each of these communities there is just one flour mill - mechanized and financed by the 'Brasil quilombola' program - whose actual operation, however, had been delayed while waiting for the visit of a technical advisor for some years already. In any case, the number of users would not be particularly high since, as the community's residents argue, they lack the space and time to plant swiddens. To some extent, the indigenous and quilombola situations can be said to contrast with each other in terms of inclusion in the labour market.

It was a few months prior to my first stay, in 2011, that the residents of the locality which I call São José da Ponte informed the competent institutions that they were 'assuming' their identity as an indigenous people. They did not then explain their decision with reference to their own values or to a specific identity, nor did they cite the need to return to their roots or to their ancestral culture - such arguments only surfaced at a later stage. At first, they asserted that they had decided to declare themselves indigenous after learning that the inhabitants of a nearby village, which I call Piratininga, were demanding recognition as quilombolas. I should immediately emphasize that, for them, the problem was not the declaration of quilombola ethnicity per se, considered as the legitimate exercise of a legal right by all the inhabitants of the area (Boyer, 2015), but in the very concrete harm that this official recognition of this identity would cause the São José da Ponte community: the interruption of access to natural resources indispensable to their reproduction.⁵

In this subregion of Amazonia, it is common for collectives to occupy a strip of land perpendicular to the river, providing them with access to different ecosystems: the floodland, or várzea, from which they extract assai, bacaba, andiroba, copaiba and other products, and where cattle are left to graze in the summer; and terra firme ('solid ground') where the swiddens are cultivated. In the particular case of these two localities, the geographical layout of the area meant that both populations ended up frequenting the same floodland zone. Now, since the territorial claim of the Piratininga quilombolas included this area, it directly threatened the future indigenous group of São José da Ponte: were the demarcation to be approved, Piratininga would have been entitled to block the latter population's entry into the zone.

As I mentioned in the introduction, the State's implementation of a different legal classification for the territories means that each has to be closed, stable and separate from the others. It also supposes that only one type of population has the right to residence and use. Indicating the boundaries of a terra (land) is thus indispensable to the work of the administrations - hence its supervision by their technical specialists − but this procedure can become highly sensitive when, as in the present case, some of the space is shared. For the inhabitants of São José da Ponte, the quilombola project struck them as a serious threat and it was thus in order to 'defend themselves,' they said, preventing their frente (front, fight for land) from being taken from them that they assumed their indianidade (Indianness, indigeneity). In so doing, they believed, they would be claiming a 'stronger' right than that of the quilombolas. As we shall see in the final part, this kind of discourse is not exclusive to the settlement.

In everyone's view, tensions have been growing between São José da Ponte and Piratininga, both of which lay claim to an area of approximately 3,200 hectares with overlapping borders. Certainly the inhabitants of the two places still take part in the same football tournaments, appear in the dance festivals that each group organizes, tell each other the latest news on the bus that runs to the town and regularly visit their relatives in the other communities. But they also admit that the subject of land needs to be avoided since they worry that the current disagreement could become an open conflict and even fear the possibility of deaths resulting.

According to data from the Comissão Pró-Índio and FUNAI, in January 2016 neither of their lands had been homologated. Despite the current clash, it is worth noting that there have been no lack of attempts to resolve the dispute by various actors (residents, mediators, institutions) working at various levels (local, regional, national) by finding an outcome satisfactory to everyone. The indigenous people of São José da Ponte say that the first initiative came from the quilombolas of Piratininga, who suggested that the former could claim an area higher up from their own swiddens, closer to terra firme, solid ground away from the floodplain, a region appropriated since the 1960s by large farmers holding more or less dubious land deeds. Aware of the risks of turning against such economically and politically powerful adversaries, who sometimes hire armed mercenaries, the residents vigorously refused: "they [the quilombolas] want us to get into a quarrel with the soybean growers and farmers. But we're better informed, we have a better idea" of the potential recourses. Rather than seeking to shift the conflict onto another actor, the second initiative, led by a lawyer from an NGO working to defend the territorial rights of vulnerable populations, sought to renegotiate the legal 'identities' involved: he recommended that the indigenous population declare themselves quilombolas, which, in his view, would have the advantage of allowing the inclusion of the São José da Ponte lands in the territory set to become a quilombo. This simplistic formulation implied that no clear distinction existed between the two settlements, allowing their residents to change legal category freely. Disrespecting the 'choice' of the local inhabitants, the idea was thus rejected vehemently by the cacique (indigenous leader), aware of their right to difference: "I said: there's nothing quilombola about us, an anthropologist isn't going to tell us [who we are]."

Another three proposals sought to address the crucial problem of defining the territorial boundaries between the settlements in more direct fashion. The first proposal, whose original source is uncertain but which was intellectually more reasonable, suggested granting the indigenous population the right to use the igapó (flooded forest) of the quilombolas. Creation of this innovative mechanism was unable to advance, however, due to the fact that the justice system lacks the legal instruments necessary to formally demarcate an area shared by lands with different property statuses. Furthermore, the indigenous population wanted the right to ownership and not just usage. The second proposal came from an anthropologist working for the Public Prosecutor's Office (MPF), who suggested measuring the distance between Piratininga and Pirapira, quilombolas located along the river shore and effectively surrounding São José da Ponte to the east and west. The riverside zone would then be divided equally between the three settlements. This project was also quickly discarded since the three allocated areas of floodland would be too small to meet the inhabitants' needs. The third proposal was based on the historical agreement reached between two other nearby villages, one a quilombola and the other indigenous, which, it is hoped, will serve as an example to be followed in other conflict situations, including the one analysed here. The adjustment of the boundaries between these two localities involved reducing the areas claimed by each. Only time will tell, though, whether, like their neighbours, São José da Ponte and Piratininga will accept relinquishing a portion of what they believe to be their legitimate territory.

Up to now, then, two localities linked by kinship networks and sharing many activities have found themselves in strong opposition and even competition for the same geographic space, and, as part of this conflict, have made use of different ethnolegal categories to strengthen their respective positions. This appropriation and elaboration of exogenous labels in their own terms (see Boyer 2015) within a localized dispute with practical finalities echoes the analyses that Stephen Nugent (1993: 102) developed in the 1990s of what he called an 'Amazonian peasantry.' For the author, comprehending this social formation created by colonial and postcolonial expansion meant observing two dimensions simultaneously: the internal dynamics and the relations with the exterior. He writes:

While it may be claimed that the concept of 'local society' has some integrity, it is hard to argue that the structure of that society is easily separable from external constraints. This is only to say that the fiction of 'local society' be recognized for what it is: an object of analysis whose internal structure is defined as much by external structures as by local ones (Nugent, 1993: 103).

This permeability to outside demands, discourses and expectations, however, does not mean that only dependency and alienation exists for the populations. In the case of the self-definition of 'identity,' whether indigenous or quilombola, the process - as we shall see - has undoubtedly been developing as a function of the connections with various external actors, themselves motivated by distinct concerns and introducing new norms. The multiple interventions made to mitigate the conflict, as listed above, already attests to the fact that these social formations are in constant dialogue with a significant range of actors.

The outcome of these discussions, however, is better understood as the latest version of a local pre-existing interplay of forces rather than as the adherence to an imposed model, since it remains, over and above everything else, linked to social relations and interests, such as defined in the micro context. Indeed, the very lack of success of these mediations demonstrates that the populations do not always accept decisions and discourses coming 'from above' or 'from outside.' It may be useful here to distinguish between two contexts of action that to some extent reflect the contrast formulated by Michel de Certeau (1980) between strategy and tactic. In the former case, oriented principally towards the outside, it can be observed that populations fighting to obtain 'rights' and protection from the State are strongly encouraged to resort to ethnolegal categories. It can also be noted that this approach depends, on one hand, on presenting their own image to the world in a particularizing way and, on the other, on indicating the borders of what is projected as their own ancestral 'territory.' From this viewpoint, it is difficult to deny the effects of the diffusion of these categorizations among the populations. It should be emphasized, however, that there also exists an appropriation and use of these legal notions deeply anchored in the local social fabric, where the meaning conferred results from their application in concrete situations - that is, vis-à-vis neighbours, compadres and possibly kin. The process of constructing 'their' difference - which, as we shall see in the third part, does not exclude identity reversals - may be intended to revert an unfavourable conjuncture into a group of allies, as Emilie Stoll (2014) has aptly showed in relation to the lower Tapajós river. Despite being intertwined, these two dimensions grow when we distinguish analytically between the formation of what I call the institutional chessboard, implying actions oriented towards 'outsiders,' and the micropolitical objectives between (almost)peers, according to rules that are sometimes not comprehended by external interlocutors.

This positioning eschews an a-historicism sometimes still prevalent in the study of these social formations. It invites us to take into account the recent history of the mobilizations that have taken place under the sign of identity, considering the unequal power relations between the outside actors and local populations and, at the same time, emphasizing the capacity of the latter to retake the initiative by adapting the series of rules to their own demands and expediencies.

SOCIOTERRITORIAL TRANSFORMATIONS

The range of external actors linked in some form to the ethnic claims found today in the area is impressive: representatives of the Catholic Church and urban ethnic militants who supported the identity 'awakenings'; representatives of NGOs working to provide the populations with the tools to campaign more effectively; State institutions that contact anthropologists to carry out the demarcations, many of them members of the Brazilian Anthropology Association (ABA). The populations position themselves in relation to the interventions of these actors and reformulate their modalities of political mobilization. This phenomenon is far from being unprecedented, however. On earlier occasions, as we shall see, they already had to deal with other proposals, which also led to reconfigurations of their sociospatial organization. In order to provide some historical background to the current transformations, I propose to reconstitute chronologically the appearance of the outside protagonists in the local and regional scenarios, as well as their respective forms of working with the populations. In the process, I turn primarily to the interviews I conducted during three research trips carried out between 2011 and 2014. It should be emphasized that here we are exploring a recent past, dating from the mid-twentieth century, which is the period to which my interlocutors refer,⁶ and an area limited to the settlements mentioned above and to their immediate neighbours.

Narratives concerning the history of the area systematically begin with the declaration that "there was nobody here when my father [my grandfather or my great-grandfather] arrived." This statement does not imply the absence of any prior inhabitants, though. People are aware that there were already a few scattered habitations. Indeed, they may at other times refer to those who had settled there previously and with whom they had sometimes entered into alliances. What they reveal with this assertion that 'nobody' was there is that, at the time when this period and the precise location were being recalled, none of them was recognized as 'kin.'

In the accounts of the residents of São José da Ponte concerning the arrival of their forebears in the area during the 1940s and 1950s, prominence is given to the considerable diversity of the geographic origins mentioned. Some of their kin, they say, come from this region of Amazonia, born a small distance away, on the neighbouring islands; others came from more distance places like the left shore of a large regional affluent of the Amazon; and some originated very far away, on the headwaters of the same river, around 250 kilometres from the region flying in a straight line. Another group came from the state capital, 700 kilometres as the bird flies; or further away still, from the Northeast and even foreign countries like Russia, the United States, Italy and so on. The new arrivals build their houses at a convenient distance from the domestic groups already present, which only visit on major occasions - holy festivals, marriages, burials and so on - or during emergencies like births, when people fall sick, or for community work rallies. In the 1960s, as well as cultivating their swiddens, many inhabitants grew jute, which was reaching the end of its economic cycle, and some extracted rubber until 1985. During this decade of the 1960s, land pressures grew with the arrival of farmers who presented themselves in the area as the 'legitimate' owners of lands and demanded the eviction of their occupants so that they could graze cattle there. At the time, various families decided to leave the space along the rivershore that they had occupied until then and moved to a higher area approximately three kilometres inland, towards terra firme, where they once again scattered. Even so they continued to extract assai and other plants from the flooded forest, a practice repeated by their descendants who still maintain this practice today - a fact of enormous importance from the viewpoint of the territorial rights currently in play, since it is these residents who today form the village of São José da Ponte, a short distance from the contemporary quilombolas of Piratininga.

The populations are beginning to emerge as members of specific territorial entities, that is, as administratively distinct collectives in the 1970s, insofar as the dispersed smaller settlements are being replaced by population nucleuses. The transformation of the forms of occupation stemmed in part from the action of the Christian Base Community movement (Movimento Eclesial de Base: MEB) that gave literary courses for adults via radio from 1965 onward, indirectly encouraging the households wishing to benefit from the broadcasts to remain within the reception area. The lists recorded in the diocesan yearbooks between 1970 and 1999 attest to this multiplication of povoados (small settlements). The sources also contain information that shed a new light on the conflict between quilombola and indigenous populations: there are two occurrences (in 1978 and 1983) of the name São José de Piratininga before the locality appeared in 1988 as São José and later, in 1996, as São José da Ponte. This suggests that the settlements today asserting their differences used to consider themselves a single collective, which, for unexplained reasons (family disagreements, political clashes, sorcery accusations or so on) ended up splitting, reiterating a dynamic common throughout Amazonia (Stoll & Folhes, 2014). Already separate, each constituted itself as a 'community,' building their first artesian wells in the mid-1970s.

Other factors driving this trend towards residential clustering were the discourse of the priests, which encouraged the populations to present themselves as community collectives structured on a spatial basis, and the leadership training courses, also run by the priests.⁷ The so-called base communities thus began to acquire a visibility unattainable by dispersed families. It is likely, in fact, that the visits by State representatives, initially by INCRA employees to recognize land ownership claims, had been planned in the mid-1970s precisely due to the efficiency of the new organisation. The intensification of the State's presence in the 1990s - while, it is worth remembering, large migratory flows were still taking place, leading to the constitution of new localities - favoured the increasing institutionalization of 'communities.' In fact, the introduction of small farmer loans by the Constitutional Fund for Financing the North (Fundo Constitucional de Financiamento do Norte: FNO), instituted by the 1988 Constitution, was conditional on the setting up of 'resident associations,' in principle registered in a notary office, the only institution able to negotiate projects like extensions to the electricity grid or the installation of microsystems with local councils or state bodies. These associations were structured in two ways: some were responsible for just one locality, while others functioned as a kind of consortium made up of three or four small settlements that would nonetheless, almost 30 years later, have different ethnic destinies, as I indicated earlier.⁸

Understanding the process of identity bifurcation that subsequently occurred requires taking into account the theological transformation undergone by the Catholic Church, which led to the reformulation of its approach to local populations. In the microregion's most important urban centre, a middle-sized town, the base of the municipality and the prelacy, this was reflected in the pastoral guidelines adopted by the US Franciscans. In 1980, the latter founded the Black and Indigenous Friars Cell (Célula dos Frades Negros e Indígenas: CFNI), urging their Brazilian brothers not only to discover their minority 'roots' but also to choose to identity with just one of them. The CFNI lasted for 17 years until the friars who defined themselves as black decided to create their own association, Black Solidarity in Amazonia (Solidariedade Negra na Amazônia: SNA), while the indigenous religious leaders formed the Indigenous Ecclesiastical Group (Agrupamento dos Religiosos Indígenas: ARI).⁹

These new pastoral approaches may well have reflected changes that were affecting the whole of Brazilian society in the 1980s. Across the entire country, the black movement campaigned to include what would become Article 68 of the new constitution (Véran, 1999). In the capital of this Amazonian state, meanwhile, the mobilization of activists and university students - sometimes the same people - led to the formation of the Black Research and Resistance Nucleus (Núcleo de Pesquisa e Resistência do Negro: NPRN) whose work would have a significant impact: strengthened by the support of the Ford Foundation, the NPRN organised a series of political-cultural events in the middle of the decade, rotating among the various communities in the region that identified as black. The implantation of the black movement in the rural zone of this microregion was in large part due to these two urban initiatives (the CNFI and NPRN), which collided with the activities of the Pastoral Land Commission (Comissão Pastoral da Terra) at the time of the Church's campaign of brotherhood with the black population. Enthused by the former's dynamism and galvanized by their commitment to assume a black identity, activists disappointed with liberation theology opted to channel their energies into the ethnoreligious currents still dominant today. One such activist, Gabriel, recalls that after taking part in the Second Black Encounter in 1988 (the year when the new constitution was promulgated and the one hundred year anniversary of the abolition of slavery in Brazil), he embraced the black cause and, in the following decade, frequently visited the communities along the rivershore, or beiradão, passing on the message of quilombola rights: just like Pedro, in fact, the Franciscan who declared himself black and a friar to the black population. Arduous work, he recalls, since the inhabitants did not understand the meaning of the word quilombola and vehemently rejected the possibility of their ancestors having once been slaves. This, however, did not stop the federal university, based in the state capital, from conducting a survey of quilombola communities in 1995.

The shift in attitude among the area's populations occurred at the start of the 2000s, when, under the advice of the NPRN, Gabriel invited two representatives from each community to participate in a seminar in the state capital, which would also be attended by the governor and various university students. One local inhabitant who took part in the event noted that they spoke "a lot about colour" and "little about the past," and that they were given talks "on what a quilombola is, what [demarcated] lands are, what projects can be organized." After this experience the communities agreed to the organisation of multiple workshops in their community halls. In 2005, the local council manifested its interest in the movement by creating the Amazonia Black School (Escola Negra Amazônia) municipal program, which reinforced the Brazil-Quilombola national program at the regional level. The following year, the SNA was closed down, and the function of representing and coordinating the quilombola communities transferred to the recently created Quilombola Federation.

Equally decisive for the indigenous movement was the initiative of a Franciscan resident from the head office of the prelacy, who declared himself indigenous and a friar to the indigenous population at the same time as his colleague was declaring his black identity. In this case, though, we do not observe the same movement of people and information between the local and national levels, nor the formation of such a broad network of relations from the outset of its implantation in this part of Amazonia. In other words, as an activist from the black movement pointed out, there was no attempt by anyone from the federal university to survey all the indigenous communities in the area, nor any NGO project to take community leaders to the state capital.

The growth of the indigenous movement in this subregion was primarily due to the action of a religious leader whose objective was initially highly localized: to raise the awareness of those of his kin living in a Conservation Unit operated according to rules established by the ICMBio. It seems, therefore, that what had prevailed was the attempt, by a spatially localized network of kin, to approach a federal institution in order to rid itself of the tutelage of another. The same friar's creation of the Indian Land Institute (Instituto Terra de Índio: ITI) at the end of the 1970s increased the visibility of the cause. In 1997, almost 25 years therefore after the first indigenous assembly to be held in Brazil, sponsored by CIMI,¹⁰ the indigenous struggles irrupted in the political field of this region of Amazonia. These campaigns became consolidated over the 2000s with one landmark being the organisation of the first FUNAI Technical Group (Grupo Técnico: GT) in the Conservation Unit in 2008. As in the quilombola case, we can observe, after the movement became sedimented, the transfer of power from the religious leaders to the activists with the foundation of the Indigenous Conference in 2005.

In 2010 - receiving support from a missionary from CIMI who had defended the indigenous population since 1988, but had later worked more specifically, as he himself said, on raising awareness among povos ressurgidos (re-emergent peoples) - localities close to São José da Ponte saw a chance to launch a campaign against the neighbouring quilombolas and decided to embrace the cause. The following year, for precisely the same reasons, São José da Ponte submitted a claim to FUNAI.

This brief historical overview reveals how, over the space of a few decades, the populations adopted various new forms of organisation, attesting to their capacity to adapt - at least formally - their institutions to external demands, combining into larger communities to meet the Christian ideal and forming resident associations to facilitate their dialogue with INCRA or ethnic associations compatible with whichever legal identity was being declared to the appropriate government body. Other evidence also emerges: in all these situations the principle actors are two collectives that nonetheless should not be considered monolithic, given that the internal relations between their members range from cooperation to rivalry. This applies as much to the Church (with activists from CEBS, the Franciscans, CIMI, etc.) as to the State (with the Federal Public Prosector's Office, FUNAI, ICMBio, and so on).

It is worth emphasizing here that the clustering in either communities, resident associations or ethnic associations makes no difference in terms of their ultimate aims. Whatever the principal language used (religious, spatial or ethnic), the objective is always to ensure access to basic social rights. From this perspective, it cannot and should not be presumed that each organisational type refers precisely to successive 'phases' or corresponds to essentially different populations, since, despite the growing number of settlements that today consider themselves to be 'indigenous villages' or 'quilombola communities,' we can note in some geographic spaces the coexistence of ethnic associations, 'communities' and 'residents associations,' as in the ethnographic situation reported here.

Substituting an ideal-typical approach for the evolutionist, I suggest that these three models refer to intermediary configurations between a pole in which land is taken as 'free,' 'given by God' or a 'wasteland,' enabling temporary appropriations and disjointed tenures, and another in which land is considered a totality, the exclusive property of a person or group. In the idea of 'community,' the reference to an abstract entity tends to mask the negotiations over land between domestic groups in order to project a larger collective occupying a shared space (Araújo, 1993). The 'residents association' asserts further still this intention of transcending local belonging and federating particular interests, creating a territorial base founded on the fact of residing and remaining in a definitive place. Finally, in the case of ethnic associations, the territory appears as a stable and continuous totality, constructed by mapping its external borders, and where the subsequent right to live in the locality in question ends up competing with another right, based on the adherence of people to a legal identity. To a certain extent, the primacy that the idea of territory has acquired in the definition of a collective implies a reformulation of social relations and the way of expressing adherence to a collective.

In order to acquire a better understanding of this increasing territorialization of space, that is, its delimitation for the purposes of juridical classification, I return to the analysis of the conflict between quilombola and indigenous populations. Reconstructing the emergence of ethnic associations at local level, I approach it now from the angle of 'choice,' an idea recurrent in the conversations of my interlocutors.

ETHNOTERRITORIAL REFORMULATIONS

As well as the clustering of settlements organised in a variety of ways, it is not rare to find among the extensive networks of relations some kin rooted in quilombola and/or indigenous associations while others describe themselves as 'residents,' 'communitarian' or 'small.' This phenomenon is mostly considered 'normal,' almost a characteristic of the times. Diverging identifications only become problematic when they occur on the perimeter of a locality that has opted for a particular ethnicity. The explanation seems to be related to another difference between the religious, spatial and ethnic paradigms: while the former two models are expressed in a single form (the 'community' or the 'residents association'), the latter has revealed a double modality from the outset insofar as the 'associations' can be indigenous or quilombola. In this latter case, there is, therefore, direct competition between proposals each of which demand exclusivity and loyalty in terms of their declarations (making it impossible indeed to be quilombola and indigenous at the same time).¹¹

What is specific here is the importance acquired by the territorial dimension, and no longer just the spatial dimension, when it comes to defining the group's boundaries: the claimed land becomes the symbol of their existence. Through a feedback effect, since the territory should be occupied by people with the same legal status (a person cannot be a quilombola in an indigenous land, or vice-versa), the question of the 'choice' made becomes of extreme importance. Unity of the territory and uniformization of belonging are, therefore, at the basis of the legitimacy of the ethnic associations. It is worth noting that, at least until their official recognition by the State, the boundaries of the territory are susceptible to modification if one or more families resolve to change their legal identity. Such a decision does not occur without generating friction, given that it is perceived as an attack on the 'territorial' integrity of the group from which they intend to extricate themselves. Since deviation and internal difference threaten the ethnic cohesion and the integrity of the territory, the circulation between the legal categories can only be individual - as, for example, in the case of marriage or a change in the place of residence.

In this microregion of Amazonia, the propagation of what can ultimately be seen as multiple versions of the idea of 'community' (a collective in which individual interest is subsumed to the common good) occurs in rapid succession, each group of residents wishing to possess its own organisation to represent the group vis-à-vis the authorities. In the case of ethnic associations, though, the mimetism has involved the assertion of an absolute difference between collectives that previously considered themselves similar. Paradoxically, the ethnic contagion also therefore brought about 'cultural' differentiation.

Figure 1
Spatial layout of the different communities¹²

The first declarations of quilombola identity appeared in the region in 2002 in two rivershore communities that founded corresponding structures in the same year. The case of one of them, Iraruana (Q),¹³ is interesting, since the friar Pedro was already talking about quilombolas in 1994 and yet it was only 18 years later that the community took its 'decision.' Soon after it invited two contiguous localities with which it shared a residents association − Bênção de Deus (I) and Morubixaba (I)¹⁴ - to take the same route. The initiative suggests that, at the time in question, this 'declaration' was considered in terms of the possibility of renewing alliances between communities that, nonetheless, would later encounter different ethnic expressions (see the figure above). With each receiving the benefits that its status allowed, the quilombola associations and residents association operated in parallel until 2005, when the latter was disbanded - i.e. when the conflict with the indigenous population irrupted.¹⁵

The procedure of the second community, Pirapira (Q), was different, based on a search for people whose phenotype was compatible with the idea of a quilombola. This aim in mind, the community's leaders visited diverse rivershore localities: Bom Sossego (Q), considered to be black, and Piratininga (Q), where they had kin, encouraging both to join the movement. As well as these 'collectives,' they also invited individuals in other settlements: in São José da Ponte (I), a man and his son who were dark skinned; and in JM, its founder, who they knew to be originally from another locality already then openly quilombola. In fact, this man recounts that he had been invited another time, in 2010, by the anthropologist responsible for the technical report. The leaders from Pirapira (Q) were unsuccessful, however: the resident from São José (I) would later become its tuxaua, indigenous leader, and the founder of JM, after an attempt to take part in the quilombola association, with which he had become disillusioned, decided to eschew any ethnic definition, content with his status as a community founder,¹⁶ recognized by the local council in 2009 and which received the initials of his name.

In 2003, the arrival in Bom Sossego (Q) of an anthropologist from INCRA, testifying to the attention given to quilombolas by the State, encouraged the village and its neighbour, Piratininga (Q), to request official recognition too. In the next two years, the localities received numerous visits from outsiders: anthropologists to produce technical reports, historians for university research, biomedics to learn more about the health of black populations, TV crews to film reports, and so on. The year 2007 was an important landmark, since two NGOs (one anthropological-juridical in kind, the other environmentalist) organised workshops to design maps based on the indications of the residents concerning areas used for hunting, fishing and swiddens, as well as the conflict zones. To strengthen the quilombola demand, it was therefore judged expedient to join together the lands that each community was demanding: they made a claim for a strip of land along the rivershore from Bom Sossego (Q) to Iraruana (Q), passing through Piratininga (Q) and Pirapira (Q). In 2010, with the technical reports completed, the wait began for the contestations before, finally, the definitive land deed was released. Bom Sossego (Q) was the first to receive official recognition, in 2011.

The indigenous mobilizations began a short time after the first quilombola declarations. Having withstood pressure from farmers for many years, Morubixaba (I) in 2003 experienced a double threat: from further inland with the installation of soybean farms using pesticides that contaminate their water supply; and closer to the river, with the claim made by the quilombolas of Iraruana (Q) over the floodland area that they traditionally explored. The reasons given for rejecting the quilombola invitations range from the critique of how they were formulated, considered somewhat unclear ("the mistake of the quilombolas was failing to communicate"), to the affirmation of a sharp cultural difference ("our characteristics didn't match theirs"), which some residents say, in fact, was endorsed by the opinion of the friar to the black population himself ("Pedro said to us that we have nothing quilombola about us"). Whichever was the most persuasive argument, the fact is that Morubixaba (I), following the line of defence suggested by another friar, declared its 'indigenousness' to FUNAI and in 2004 founded the appropriate indigenous association. In 2005 it promoted an encounter in a small town in the region with their Ticuna 'kin,' since they had 'chosen' to adopt this name despite the distance between them. Also in 2005 the association was officially recognised. In 2008, proposals surfaced to demarcate an Indigenous Land and the first workshops were held on medicinal herbs and indigenous languages by the missionary from CIMI.

In 2009, the year when Morubixaba (I) and Iraruana (Q) managed to reach agreement concerning their respective limits, new problems emerged on the boundaries between indigenous and quilombola. Aratú (I), a neighbouring locality that also took part in the residents association in Iraruana (Q), contested the territorial claims of the latter. São José da Ponte (I), for its part, separated from this zone of tensions by two other settlements, also opposed another quilombola community, Piratininga (Q), which decided to publicly affirm its indigenous status at the end of 2010, thus abandoning its organisation as a residents association, an institution founded in 1999.

On the first of April 2011, the three settlements met in Morubixaba (I) to celebrate the Day of the Indian. According to local accounts, the event even included indigenous people from outside the region, and, amid the festivities, they decided to combine forces to make a claim for a joint Indigenous Land. On the occasion, invitations to take part in the movement were made to Bênção de Deus (I), whose geographic position between Morubixaba (I) and São José da Ponte (I) was strategic to the project. The reaction of the inhabitants from this locality was uneven: many considered it a good option to defend themselves from a farmer, avenging his attacks on them (including destruction of the school and the slaughter of their pigs), but some families that recognized themselves as quilombola, "because of their colour"¹⁷ or their kinship ties (recalling that the settlement had already received an invitation from Iraruana (Q)), refused to back the indigenous cause.

The same year, Morubixaba (I) and São José da Ponte (I) asked for their lands to be mapped by one of the NGOs that had produced the cartographic survey for the quilombolas. The request was important since the publication of the quilombola map had shown the indigenous settlements without any indication of their names, as conflict zones, equivalent in effect to the areas occupied by farmers and soybean cultivators. Responding to their request, 'participatory' meetings began to make another map that would help rebalance the forces between the quilombola and indigenous settlements. Providing the latter with cartographic proof of their existence and their 'ancestral' use of the flooded forest areas (just as the map of the quilombolas verified their equally 'ancestral' use of the várzea floodland), allowed one right to be asserted against another. It was in this upsurge of mobilization that Bênção de Deus (I) finally decided 'to assume' its status, thus strengthening the indigenous position. It is worth noting that in the publication of the indigenous map, the quilombola areas are not shown as 'conflict zones'; in addition and despite being called the Ticuna do Planalto Indigenous Territory, the document does not contain an indication of the territorial borders.

To complete this depiction of the situation in the area in 2015, I mention the case of another community, Igarapé Arara (T), which, for its part, 'chose' a non-ethnic legal identity: that of a 'traditional population.' After creating their association in 2000 to fight against a farmer, the residents accepted INCRA's proposal to form an agroextractivist settlement (assentamento), a legal category that afforded them various benefits - microsystems, home loans, electricity, schools and so on - although it has proven to have its limitations: the demographic growth and, according to the accusation of one of the settlers (assentados), the monopolization of the best plots of land by long-standing leaders, as well as the continued occupation by a farmer, mean that two thirds of the domestic groups find themselves landless today. This 'choice' of ethnic indeterminacy left doubts hovering over both the quilombola and indigenous populations, since both believe that the "people call themselves traditional, but that's due to a lack of knowledge." Not even this, however, persuaded them to send an 'invitation,' probably because the lands of Igarapé Arara (T) were outside the claimed zone.

The description of the constitution of the quilombola and indigenous associations makes evident the flexibility of the ethnic destiny of povoados whose identity is defined, above all, through 'choice' in response to situations influenced by the intervention of different external actors. The sequence by which settlements joined the two parallel movements also highlights the local constitution of networks of alliances and their modalities, and, at the same time, the importance of territorial factors. Sometimes this territorial imperative is clearly evident. Hence, despite never being extended to large landowners - Morubixaba (I) did not allow the soybean cultivators to join its association, nor did São José da Ponte (I) accept the association of farmers located inland from its own lands − the 'invitation' was sometimes made to neighbours with better economic conditions. The leaders of São José da Ponte (I), for example, tried to convince some people to include their names in the IL request: three small proprietors whose lands were located on the boundary with Piratininga (Q) and Pirapira (Q) (two of them already approached by the quilombolas) and a gaúcho (southern Brazilian) farmer who owned lands containing a waterfall that he exploited as a spa resort.

Most of the time, though, invitations were made directly to individual and those of their dependents considered to be 'kin' (whether through filiation, marriage or godparenting) or closer to what is expected from a quilombola or indigenous person (whether due to physical appearance or place of birth). The founder of JM, a native of a quilombola community, and the tuxaua of São José da Ponte (I), father of the president of the Piratininga (Q) quilombola association and himself dark skinned, meet at least one of these criteria. Something else at stake in these 'invitations' was a redefinition of territorial boundaries, since acceptance of these men would entail the breakup of their respective 'communities': the separation of the lands occupied by them and their children in order to combine them with the quilombola territories. As I indicated earlier, the rejection probably arose from the fact that the invitation involved more costs than benefits: the former man would risk losing his status as a founder and his autonomy; the latter would be exposed to a serious fight with his neighbours and potential isolation. At the time, in fact, it was deemed unnecessary to even report the fact to the president of the former residents association, believing that "it was nothing important."

In other localities, when the person 'called' has a stronger position in the residential group or in the 'community' and is thus more able to mobilize and convince their kin and neighbours to pass on the invitation, the action can be presented as collective. Such was the case of Bênção de Deus (I), whose president spent more than a year organizing meetings, inviting sympathetic local leaders and friars, until everyone, including the former quilombolas, had heard the message and indianidade (Indianness, indigeneity) had become a locally shared and agreed 'definition.'

The example is illustrative of the importance, in the interplay of forces between quilombola and indigenous groups, of those populations that are still undetermined but whose 'definition' is capable of leading to a radical transformation of the ethnic and thus territorial configuration. These declarative changes do not always imply ruptures in political power (the father of the cacique (leader) of Morubixaba (I) was the community president), but their success clearly depends on respect for pre-existing political equilibria (for example, in São José da Ponte (I), the cacique belongs to the part called 'above,' the tuxaua to the 'middle,' and the pajé (shaman) to the 'below,' in a tripartite organisation already present in the residents association with the distribution of the functions of president, secretary and treasurer).

Unsurprisingly the 'choices' made by the collectives can be inscribed without difficulty in networks of concrete relations that radiate within a particular geographic space. As we know, relations, in the case of these and other populations,¹⁸ are valued as a resource that enables access to goods and services. Obviously this does not mean that spatial limits are unknown or unrecognised: everyone knows where the neighbour's swidden is located and where the zone of influence ends, but a request to hunt, plant crops or the like will seldom be denied. In this context, 'ownership' of land means not so much exclusivity of use and associated rights than public assumption of the role of the principal administrator capable of granting licenses to third parties. This idea still prevails within the community and even in the residents association, but tends to disappear among the ethnic associations.

Two remarks illustrate complementary aspects of the transformations under way. The first is the advice received by the cacique of São José da Ponte (I) from an outside interlocutor − "try to make a territory and you'll be stronger" - signalling that occupation of a space, knowledge of its characteristics and the association of its inhabitants are not sufficient to obtain the attention and interest of the State. For this the space needs to be enunciated as a territory, that is, as a delimited zone in which a determined group of people reside, called 'kin' and possessing 'rights.' In another conversation, in which the relationship between the definition of territorial boundaries and kinship was evoked, the coordinator of São José (I) exclaimed: "were we to go in search of kin, it would be no use: we would get all of Piratininga (Q)." In other words, the reticular character of the kinship networks means that, from their point of view, the indigenous population could lay claim to quilombola lands. The hypothesis is absurd, comic even, since, as is well known, the 'choice' made must be respected. To conform to this shared value, vis-à-vis outside agents, it is thus better to keep quiet about the breadth of the kinship mesh constructed in the local vicinity in everyday life. Instead, people emphasize relations with those who, geographically close or distant, present the same legal situation. In this sense, kinship and territory are capable of sustaining each other. The potential weakening of intercommunity solidarity, as in the case analysed here, is the price to pay for these new political opportunities.

In summary, the different territorialities elaborated as a result of the interlocutions with outside actors can be said to impact both intra and inter-settlement relations. Nonetheless, it should also be stressed that while the forms of institutionalization of collectives may be borrowed, these populations appropriate them in order pursue their own agendas and strengthen the position of a set of kin and their allies within the local political field.

CONCLUSION

As I mentioned in the introduction, in the case of indigenous peoples the literature points to processes involving the co-production of the borders of a territory and a collective. This ethnographic example contributes, in turn, to documenting analogous processes in populations considered - and until recently considered by themselves - to be 'mixed.' The interesting aspect specific to the situation presented here is that the legal framework enabled their identification not only with an ethnolegal category but with two of these at least (leaving aside 'traditional populations'): 'indigenous' and 'quilombola.' This possibility led to the coexistence of diverse organisational models in a relatively small geographic space. Alongside those povoados that perceive themselves as a 'community' or a 'residents association,' others come to see themselves as an 'indigenous village' or a 'quilombola community.' The plasticity of formal organisations, in parallel with their diversity, is the best confirmation of the idea that the populations really do exercise a power of 'choice,' all the more so since the adherences are open to modification, for example, when localities decide to publicly change their 'identity.'

However, this does not impede the identification of a number of rules governing the 'choices' made and some constraints. The example analysed here show that the positioning of the localities in favour of one or other ethnic label took place as a result of their embedding in supralocal relational networks and their appropriation of information brought by diverse external agents. In this process, the elaboration of the language of ethnicity in conceptions concerning the access to 'rights' - i.e. to citizenship - was fundamental. However, the adherences were also linked to a local political interplay of forces between settlements with similar economic conditions, making use of competing legal devices to obtain control of a zone of common interest. From this viewpoint, it is as though the idea of citizenship, present in all the claims, had various facets, some affording more rights ('indigenous'), some less ('quilombola') and some even none ('common residents').

On the other hand, however much these publicly asserted ethnic 'identities' were later deployed in the consolidation of alliances, it can be observed that, initially, the 'invitation' is made not on the basis of previously observable differences or the 'origin' of people, but rather of the quality of active relations: competing with São José da Ponte (I) for access to the floodland, Piratininga (Q) communicated the existence of the project to just one resident, who became kin of the president of the quilombola association; dissatisfied with governance of the association by Iraruana (Q), the leader of Bênção de Deus (I) convinced the settlement to join the indigenous group; no resident was excluded for being Northeastern, a descendent of Italians or failing to match the expected physical type.

There is no doubt that, until then and without presuming the kinds of reversals arising from shifts in federal government policy, the populations found the legal framework - in which diverse legal categories are made accessible to them - to be a source of further opportunities to demand protection from the authorities. Given the challenges posed to social mobilizations by the conflict-ridden territorial configurations, however, it is also essential to stress that the initial fluidity between legal categories reflects a shared struggle of the populations to ensure their own survival.

In the case in question, instead of a broad coalition between nearby settlements to demand an extensive area, there was the gradual constitution of two 'groups-territories' (Boyer, 2014), both insisting on the contrast and anteriority of their ancestors' presence. Although the ethnic argument never adheres to a criterion of genealogical and absolute truth, but rather an authenticity demonstrable by the willingness to defend and fight for the cause, it becomes more imperative after the lands are demarcated, since while the boundaries of the 'territories' are designed in response to alliances and tensions, from the moment when they are made official, the boundaries established act as proof of ethnicity. Thereafter any redefinition of the relevant socioterritorial units as the outcome of cooperation or conflict, or the renegotiation of spatial divisions, are rendered impossible. The only thing that matters is the list of residents established for a determined area and transmitted to the institutions: what prevails, therefore, is the appeal to the law and writing. Ana Beatriz Vianna Mendes (2008: 2) rightly observes that while the demand for demarcation clearly has to do with the management of territories, it also relates to the "guarantee of special rights." Now, in this context defined by an "institutionalization of an arena of disputes" (Mendes, 2009: 5), the risk is simply losing the status of kin and resident, as well as those rights associated with it. As Eduardo Brondizio and Rodrigo Penna Firme (2007) and Wilian J. Santos de Arruda (2014) have already pointed out vis-à-vis the quilombola and indigenous mobilizations, respectively, although the use of ethnic language favours the access of marginalized populations to political and economic resources, it also contributes to creating socioterritorial barriers that give rights to some and not others, which exacerbates the internal social inequalities.

It is, therefore, the work of the administrations that favours the transformation of space into territory, what João Pacheco de Oliveira (1998) calls territorialization, or what I call territorial framing (Boyer, 2016). This sometimes leads to cuts in the networks of social relations that traverse the territorial space and give it density. The researcher's perplexity at the discovery of people circulating between territories considered hermetic partly results from this divergence between the logic of the State and the local dynamics. Of course, in this context of a geography of power in the process of being reconfigured (Oslender, 2002: 90), it remains to be seen whether the formality of the law prevails in all cases or whether informal agreements - or 'cooperative arrangements,' to use the expression of José Cândido Lopes Ferreira (2013) - persist once the lands are demarcated. Put otherwise, only a supplementary study can say whether a rule of access (the law establishing who possesses the formal right over natural resources) can be reconciled with a rule of use (who has effective access to the areas).

To conclude, I call attention to the fact that the research data suggests the coexistence - or rather the overlapping - of two narrative modes. The first relates to the memories of how this microregion was occupied, which I have attempted to reconstruct through the accounts of the residents: in them appear shared memories of exchanges, recent arrivals and older inhabitants, alliances and misunderstandings, transformations in living conditions and increasing contact with the city. Over the more recent period, during the 1980s, another discourse emerged, militant this time, centred on origins and, to a certain extent, mirroring the difference between the narrative of the nation and the narrative of the myth of racial democracy. The new reading, which came to the fore in the 1990s, focuses on the parallel histories specific to each of the ethnoracial components: the history of conquest for the 'indigenous' population, and the history of slavery for the 'quilombola' population, always in opposition to the domination of the 'whites.' To support the demarcation requests, these narratives have to be compatible with the declarations. They act, therefore, as founding myths for the groups-territories, leaving no room for 'mixture,' recompositions and mutual borrowings. Ceasing to be public, the latter thematics are relegated now exclusively to the personal sphere. However, they prove to be of huge importance when thinking about the modalities of ethnic differentiation. Although the impression may sometimes have been generated that the identification of the settlements was random in nature (due to the interlocutors and the proposals that they brought¹⁹), this certainly was not the case: the steps taken by the activists were undoubtedly guided by their own networks of relations, buoyed, in turn, by alliances and by conflicts.

In a recent article, Guedes (2016: 28) observes that, in the past, the definition of these populations as camponesas (peasants) emerged "from the articulation of political and academic questions that implied privileging certain themes and topics as potentially capable of encompassing heterogeneous realities and universes." The author then calls attention to the fact that during this process "aspects, traits and dynamics that escaped the boundaries delineated by this definition were 'non-visibilized'' [...] or ignored" (28). In a way, researchers did not see the 'quilombola' population (or 'indigenous' population) due to their intellectual project and disciplinary blindness. By contrast, Guedes continues, "in the current fight of traditional communities, there is a preeminent need to privilege, intellectually and politically, certain aspects and traits that [...] were not previously the subject of much investment by intellectuals and social movements" (28). In other words, since the discipline's gaze conditions its research objects, when anthropologists regard 'quilombola' populations (or 'indigenous' populations), they tend to overlook the circulation of people and what is shared by the majority. In order to avoid that the work of 'unfreezing' the definitions of a quilombola or indigenous collective leads to the 'freezing' of the social groups involved, it is essential for the discipline to meet the challenge of providing more knowledge about the society (or societies) involved. It would also be of major interest to continue with an ethnography of the other side of the mirror - of the diverse organisations and institutions, specialists and activists.

NOTES

• 1
  My thanks to Emilie Stoll for the careful and critical reading of an early version of the article. I also thank Peter Fry for his generous comments. The two helped me to improve the article significantly.
• 2
  For an inspiring ethnography of such conflicts on the lower Tapajós, see Emilie Stoll (2014).
• 3
  The advantages or frustrations of different territorial statuses can, indeed, lead populations to solicit ethnic reclassification. See Ana Beatriz Vianna Mendes (2009: 12) who observes that the restrictions imposed in the case of conservation units "can be taken as the causes of diverse social and political mobilizations and responses to the newly created sociojuridical reality." It is these new norms, resented as unfair constraints, that sometimes lead the populations to try to replace the label of 'traditional' with 'indigenous' (for Acre state, see, for instance, Pantoja, 2008).
• 4
  Among other examples in Amazonia: Abirached, Carlos Felipe de Andrade; Brasil, Daniel & Shiraishi, Juliana Costa (2010); Costa, Eliza Mara Lozano (2012); Eduardo Vieira Barnes (2006).
• 5
  Similar observations can be found in Arruda (2014: 90) apropos a conflict between "indigenous [and] traditional populations." There is no relation of illegitimacy or radical opposition to the Kuntanawa ethnic identity, but rather a conflict over the collective use of natural resources in a determined area by the two groups.
• 6
  Taking into account the written documentation (the sesmaria land deeds, the chronicles of the Jesuits, the records of slave sales, or the descriptions of travellers, for example) could, of course, shed a new light on the history recounted by local inhabitants, adding more information on the national and international migratory flows. However, it would be more difficult to maintain the anonymity of the ethnographic situation. I therefore decided to concentrate on the collected accounts. It is worth stressing, nonetheless, that the quilombola and indigenous demands both have a historical basis, with the region known to have contained a sugar plantation worked by slave and servile labour. For a general panorama of the forms of territorialization imposed on indigenous populations, see João Pacheco de Oliveira (2016).
• 7
  The work of Roberto Araújo (1993) is an essential landmark in the studies on the notion of 'community,' calling attention to its constitution as a religious ideal, a domestic model that encompasses relations of domination. The description proposed in the final part of this article concerning the sequencing of the ethnic 'invitations' is informed by the reading of this thesis.
• 8
  Significantly, the farmers, soybean growers, wholesalers and traders, who undoubtedly played an important role in this process of forming larger groups, remain absent from the narratives.
• 9
  Jan Hoffman French (2007) analysed a similar case in the Northeast in which the actions of members of the Catholic Church affected the indigenous and quilombola mobilizations. It is important to note the differences, though. First, the two land campaigns occurred there over an interval of 20 years, delineating a very different situation to the one present in the Amazonian case under study here where they are almost simultaneous. Second, the friar to the indigenous populations and the father to the black population, who had not passed through the seminary training together, did not consider themselves part of the 'people' who they were defending.
• 10
  For a sociohistory of the indigenous movement, see Jean-Philippe Belleau (2014).
• 11
  In passing, it is worth mentioning that the arrival of Evangelical churches in Catholic settlements, with their structuring into congregations competing with the community, is capable of provoking a similar double configuration. It can be observed, in fact, that the congregation is constituted on the existing community structure, seeking to replace its hierarchies and occupy its place in defining authority, reinforcing kinship with religious ties (Boyer, 2016). From this viewpoint, the Evangelical and Catholic versions seem to reflect the contrast between quilombola and indigenous associations. The similarity ends there, however, when we observe that the Evangelical message - affirming that salvation comes from Jesus and not from some institution - itself contains the possibility of a new duplication of the congregation. The churches need, therefore, to adapt to the fragmentation in the field of legitimacy, a principle intrinsic to its expansion, but strongly rejected by the community and the ethnic associations, which aim to represent everyone.
• 12
  The quilombola settlements are the most populous: Piratininga contains 70 families; Pirapira, 85; Bom Sossego, 80; and Iraruana, 86. The indigenous villages are smaller: 46 families live in São José da Ponte, 40 in Morubixaba and 22 in Aratú. As for the 'undefined' settlements, these vary between 15 (Bênção de Deus) and 40 (JM). And 90 families live in Igarapé-Arara.
• 13
  In order to facilitate the comprehension of the processes described here, I have chosen to indicate their classification after the name of each locality: (I) for indigenous, (Q) for quilombola, (T) for traditional, adding nothing when the inhabitants consider themselves 'common residents.'
• 14
  There was no mention of an invitation to Aratú (I), which also participated in the residents association and later assumed an indigenous identity.
• 15
  São José da Ponte was the last settlement from the area to declare itself indigenous, which it did at the end of 2010.
• 16
  In this case, the founder was not responsible for initiating the occupation but for making it official. The first person to arrive in the 1940s, as he readily acknowledges, was a woman fleeing from the large floods.
• 17
  The colour argument is indissociable from the quilombola land claim, since quilombolas are defined as descendants of African slaves. Three observations should be made. This argument does not to be applicable to everyone: Bom Sossego, where many residents are darker skinned than those from other settlements in the area, appears here as an exception. In most cases, the inhabitants are content with indicating a person who, due to their phenotype, embodies the black (or indigenous) genealogy. This does not mean, though, that people were indifferent to colour: for example, I was able to hear people from both the indigenous and quilombola populations say that the tuxaua of São José, whose skin is black, may well have 'chosen' to be quilombola. The reference to physical traits may also have a pejorative intent with the emergence of arguments based on 'non-compliance' with what is expected of an indigenous or quilombola person (Boyer, 2015). Finally, it should be observed that this does not necessarily imply the vanishing of discrimination - in one conversation, a director from one quilombola association referred to a persistent racism, "including among children."
• 18
  See Araújo (1993) on the TransAmazonian Highway and Hoffmann (2004) on Colombia.
• 19
  In this microregion of Amazonia, the propagation of what can ultimately be seen as multiple versions of the idea of 'community' (a collective in which individual to one member of the black movement, the fact that they had been unable to declare themselves indigenous in the survey conducted by the university contributed decisively to the advance of the quilombola movement, while the indigenous population struggled to organize themselves until the appearance of the 'territorial problems.'

BIBLIOGRAPHY

• Abirached, Carlos Felipe de Andrade; Brasil, Daniel & Shiraishi, Juliana Costa. (2010). Áreas protegidas e populações tradicionais: Conflitos e soluções. Paper presented at the 5th ANPPAS National Meeting, Florianópolis.
• Araújo, Roberto. (1993). La cité domestique: stratégies familiales et imaginaire social sur un front de colonisation en Amazonie brésilienne PhD Thesis. Université Paris X Nanterre.
• Arruda, Wilian J. Santos de. (2014). Os dois lados do rio: etnicidade e disputas territoriais na reserva extrativista do Alto Juruá Final Course Assignment (Social Sciences). Centro de Filosofia e Ciências Humanas/Universidade Federal do Acre.
• Barnes, Eduardo Vieira. (2006). O conselho do Parque Nacional da Serra do Divisor: Rituais políticos, (sobre)posições e representações territoriais Master's Dissertation. Universidade de Brasília.
• Belleau, Jean-Philippe. (2014). Le mouvement indien au Brésil. Du village aux organisations Rennes: Presses Universitaires de Rennes.
• Boyer, Véronique. (2016). The demand for recognition and access to citizenship: ethnic labelling as a driver of territorial restructuring in Brazil. In: Lehmann, David (ed.). The crisis of multiculturalism in Latin America New York: Palgrave, p. 155-178.
• Boyer, Véronique. (2015). Énoncer une 'identité' pour sortir de l'invisibilité: La circulation des populations entre les catégories légales (Brésil). L'Homme, 214, p. 7-36.
• Boyer, Véronique. (2014). Regimes de diferenciação, registros de identificação: identidades, territórios, direitos e exclusão social. In: Birman, Patrícia et al. (eds.). Dispositivos urbanos e trama dos viventes: ordens e resistências Rio de Janeiro: Ed. FGV, p. 181-196.
• Brondizio, Eduardo & Penna-Firme, Rodrigo. (2007). The risks of commodifying poverty: rural communities, quilombola identity, and nature conservation in Brazil. Habitus, Goiânia, 5/2, p. 355-373.
• Castro, Fábio de. (2012). Multi-scale environmental citizenship: traditional populations and protected areas in Brazil. In: Latta, Alex & Wittman, Hannah (ed.). Environment and citizenship in Latin America: natures, subjects, and struggles New York: Berghahn, p. 39-57.
• Certeau, Michel de. (1980). L'invention du quotidian, I: Arts de faire Paris: Union Générale d'Éditions.
• Costa, Eliza Mara Lozano. (2012). Conflitos entre índios e não índios no rio Amônia Paper presented at the 36^th ANPOCS National Meeting, Águas de Lindóia.
• Ferreira, José Cândido Lopes. (2013). Organização social e regimes de propriedade numa comunidade quilombola paraense - Trombetas Master's Dissertation. Programa de Pós-Graduação em Antropologia/Universidade Federal de Minas Gerais.
• French, Jan Hoffman. (2009). The rights and wrongs of land restitution: 'Restoring what was wurs'. In: Fray, Derick & James, Deborah (eds.). The rights and wrongs of land restitution Abingdon: Routledge, p. 123-140.
• French, Jan Hoffman. (2007). A tale of two priests and two struggles: Liberation theology from dictatorship to democracy in the Brazilian Northeast. The Americas, 63/3, p. 409-443.
• Gaillois, Dominique Tilkin. (2004) Terras ocupadas? Territórios? Territorialidades?. In: Ricardo, Fany (ed.). Terras indígenas e unidades de conservação da natureza. O desafio das sobreposições territoriais São Paulo: Instituto Socioambiental.
• Guedes, André Dumans. (2016). Lutas por terra e território, desterritorialização e território como forma social. Revista Brasileira de Estudos Urbanos Regionais, 18/1, p. 23-39.
• Hoffmann, Odile. (2004). Communautés noires dans le Pacifique colombien: innovations et dynamiques ethniques. Paris: IRD/Karthala.
• Lefebvre, Henri. (1974). La production de l'espace, Paris: Anthropos. Available at <http://www.persee.fr/doc/homso_0018-4306_1974_num_31_1_1855.> Accessed 10/07/2017
  » http://www.persee.fr/doc/homso_0018-4306_1974_num_31_1_1855.
• Mendes, Ana Beatriz Vianna. (2009). Conservação ambiental e direitos multiculturais: reflexões sobre Justiça Campinas: Ed. Unicamp.
• Mendes, Ana Beatriz Vianna. (2008). Ambientalização de direitos étnicos e etnização das arenas ambientais: populações tradicionais e povos indígenas da Reserva de Desenvolvimento Sustentável de Mamirauá (AM). Paper presented at Reunião Brasileira de Antropologia, Porto Seguro, Bahia.
• Nugent, Stephen. (1993). Amazonian caboclo society. An essay on invisibility and peasant economy Providence/Oxford: Berg.
• Oliveira, João Pacheco de. (2015) Stratégies politiques des indigènes dans l'Amazonie brésilienne: agence, échelle et territoire. Cahiers des Amériques Latines, 78. Available at <http://cal.revues.org/3562.> Accessed 10/07/2017.
  » http://cal.revues.org/3562.
• Oliveira, João Pacheco de. (1998). Uma etnologia dos "índios misturados"? Situação colonial, territorialização e fluxos culturais. Mana, 4/1, p. 47-77.
• Oslender, Urich. (2002). The logic of the river: a spatial approach to ethnic-territorial mobilization in the Colombian Pacific region. The Journal of Latin American Anthropology, 7/2, p. 86-117.
• Pantoja, Mariana Ciavatta. (2008). Os Milton. Cem anos de história nos seringais Rio Branco: Edufac.
• Robert, Pascale de. (2004). 'Terre coupée'. Recomposition des territorialités indigènes dans une réserve d'Amazonie. Ethnologie française, XXXIV/1, p. 79-88.
• Miller, Byron. (2011). Spatialities of mobilization: building and breaking relationships. In: Nicholls, Walter; Miller, Byron & Beaumont, Justin (eds.). Spaces of contention: spatialities and social movements Aldershot: Ashgate, p. 326-342.
• Rezende, Roberto Sanches & Postigo, Augusto. (2013). Reconhecimentos territoriais e desconhecimentos institucionais. Ruris, 7/2, p. 119-142.
• Stoll, Emilie. (2014). Rivalités riveraines: territoires, stratégies familiales et sorcellerie en Amazonie brésilienne PhD Thesis. École Pratique des Hautes Études.
• Stoll, Emilie & Folhes, Ricardo. (2014). La (dés)illusion communautaire: notes sur l'ambivalence de la notion de 'communauté' en Amazonie brésilienne. Journal de la Société des Américanistes 100/2, p. 73-103.
• Vanier, Martin (dir). (2009). Territoires, territorialité, territorialisation. Controverses et perspectives Rennes: Presses Universitaires de Rennes.
• Veran, Jean-François. (1999) "Les découvertes du quilombo. La construction hétérogène d'une question nationale". Problèmes d'Amérique Latine, 32, p. 53-72.

Publication Dates

• Publication in this collection
  May-Aug 2017

History

• Received
  30 Jan 2017
• Accepted
  22 Mar 2017