We know how, in stories of ‘attraction and pacification’ of indigenous groups, references to the existence of ‘angry indians’ (índios bravos) in the vicinity are recurrent. These can be contacted later on, but their presence while isolated, that is, as present in the form of an absence – an existence suggested in the negative by traces, footprints, noises, fleeting shadows – only seems to disappear under the condition of reappearing further ahead. It all happens as if every ‘exit’ made by an isolated people were to induce the hidden presence of another isolated people. Frequently the isolated group is a recalcitrant or lost part of a group which gave in to attempts at contact or which actively sought it. Given the morphological characteristics of amazonian sociality – and their historical courses within a continent under occupation – there is always a part apart in every collective. As much as the adjective ‘isolated’, the noun ‘people’ must be understood as relative and relational – at least until the statal nomos arrives to fix identities and enclose territories. An indigenous community whose contact was made in a given indigenist ‘attraction’ operation rarely coincides with the entire population of the people it belongs to (in whatever scale or perspective this term is taken), except where there have been significant demographic losses. The spatial dispersion, temporal instability and political fractionality of indigenous collectives means that local groups have a lability capable of inhibiting unanimous movements and ‘tribal’ type solidarity. On the other hand, the internal heterogeneity of a cross-community network, from the point of view of intensity of interaction with Whites, is frequently an adaptive functional characteristic of Amazonian collectives. Even when an entire people has been ‘pacified’ and its territory completely surrounded by the white ocean, leaving little or no possibility of the existence of uncontacted groups in the surroundings, the isolated peoples remain the object of a sort of presumption of existence, of a variable epistemological status, from the contacted peoples. Material indices of a non-familiar presence in the surroundings of the village; perceptual illusions and hallucinations; dreamlike experiences; shamanic testimonies; anecdotes; expeditions in remote areas… all kinds of materials support, cumulatively or alternatively, the discursive persistence of the figure of the isolated people.