166 HUMAN ORGANIZATION Human Organization, Vol. 68, No. 2, 2009 Copyright © 2009 by the Society for Applied Anthropology 0018-7259/09/020166-15$2.00/1 Introduction B razil is an exceptional case in Latin America for hav- ing a reported indigenous population that constitutes such a low proportion of the total population (0.4%). Only the indigenous population of Uruguay is lower than Brazil’s, being just 0.02 percent of the national population. Demography, Territory, and Identity of Indigenous Peoples in Brazil: The Xavante Indians and the 2000 Brazilian National Census Nilza de Oliveira Martins Pereira, Ricardo Ventura Santos, James R. Welch, Luciene G. Souza, and Carlos E.A. Coimbra Jr. Brazilian census data show a remarkable increase in the population self-reporting as “indigenous” between 1991 and 2000 but do not readily enable that increase to be analyzed in terms of the nearly 200 specific indigenous societies or ethnicities that exist in Brazil. In this article, we investigate some instances and implications of how the 2000 Brazilian National Census employed categories conceived for the national population to register one specific people—the Xavante of Mato Grosso, Central Brazil—with their own inherent social arrangements and morphologies. We do so by comparing census data corresponding to Xavante Indigenous Reserves with an independently collected set of demographic data for the same year. Although we found census data to adequately represent basic characteristics of the Xavante population (population size and age and sex distributions), we also found they reclassified and transformed Xavante households and thereby denatured Xavante sociality of its demographic and sociocultural complexity. The Xavante case is an example of how national demographic censuses not only capture data regarding indigenous peoples but also help shape those data by contributing to how indigenousness is perceived. Our findings suggest that the Brazilian National Census should seek to be more sensitive to indigenous realities and thereby to assess more accurately fundamental aspects of indigenous societies. Key words: Brazil, South American Indians, census methods, cultural anthropology Nilza de Oliveira Martins Pereira (Instituto Brasileiro de Geografia e Estatística / is with the Brazilian Institute of Geography and Statistics— IBGE, Rio de Janeiro) (nmartins@ibge.gov.br). Ricardo Ventura Santos (Escola Nacional de Saúde is with the Pública / National School of Public Health, Fundação Oswaldo Cruz and Department of Anthro- pology, Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro) (santos@ensp.fiocruz.br). James R. Welch (Ph.D. is a Candidate in Anthropology, Tulane University and Affiliate Researcher, Escola Nacional de Saúde Pública / National School of Public Health, Fundação Oswaldo Cruz, Rio de Janeiro) (jwelch@tulane.edu). Luciene G. Souza (Fundação i with the Nacional de Saúde / National Health Foundation—FUNASA, Brasília) (lucienegsouza@uol.com.br). Carlos E.A. Coimbra Jr. (Escola Nacional de Saúde is with Pública / National School of Public Health, Fundação Oswaldo Cruz, Rio de Janeiro). This research was made possible by funds provided by the Brazilian Research Council (CNPq), grant number 483.483/2007-9, and Fulbright Commission (Fulbright-Hays Doctoral Dissertation Research Abroad Fellowship), award number P022A040016. Although heterogeneous demographic histories may explain much of the variation in indigenous population proportions among Latin American countries, they are insufficient to explain the radical differences in indigenous population size between Brazil and its immediate neighboring countries. For example, Brazil’s indigenous population size is striking when compared with Guyana (6.0% indigenous), Colombia (2.0%), Venezuela (2.0%), and even Argentina (1.0%) (Meentzen 2001; Montenegro and Stephens 2006). In Brazil, as in many other South American contexts, indigenous ancestry is prevalent in the population even though perception of or identification with that ancestry may be uncommon in many segments of society (Ramos 1998; Santos and Maio 2004; Yashar 2005). Accord- ingly, it is that very perception, whether within the censused population or institutionalized in the census project, that may have enormous impact on reported national figures regarding indigenous population size. Furthermore, this and other percep- tions regarding indigenousness and non-indigenousness ines- capably permeate the entire census process, thereby affecting their results and tangibly influencing public discourse. One of the most surprising results of the most recent Brazilian National Census, done in 2000, compared to the previous census (1991), was the increase in the number of people self-reporting as indígena (“indigenous”) (IBGE 2005; Kennedy and Stephen 2000). The number of people