A preliminary version of this article was presented to the Family History Seminar of the Shelby Cullom Davis Center for Historical Studies at Princeton University in the fall of 1977. The author wishes to acknowledge her appreciation to members for their comments. Joyce Riegelhaupt and Z. I. Giraldo deserve special thanks for their suggestions for the revised version.

¹ Freyre, Gilberto, Casa-grande & senzala: Formaçāo da familia brasileira sob o regime de economia patriarchal, 3rd ed. (Rio de Janeiro, 1938), pp. xiv-xvii.Google Scholar

² Junior, Caio Prado, The Colonial Background of Modern Brazil, Macedo, Suzette, trans. (Berkeley, 1967), pp. 410–11;Google ScholarFrancisco José de Oliveira Vianna, Instituićões politicas brasileiras (Rio de Janeiro, 1955), I, 259–60.Google Scholar

³ Duarte, Nestor, A ordem privada e a organização politico national, Brasiliana 172, 2nd ed. (São Paulo, 1966), pp. 70–71;Google ScholarCarone, Edgard, A República Velha (São Paulo, 1970), pp. 249–84;Google ScholarWillems, Emílio, ‘The Structure of the Brazilian Family,’ Social Forces, 31 (1953), 345;CrossRefGoogle ScholarSmith, T. Lynn, Brazil: People and Institutions, 2nd ed. rev. (Baton Rouge, 1963), pp. 459–61;Google ScholarWagley, Charles, An Introduction to Brazil, pp. 192–203.Google Scholar

⁴ Prado, Caio, p. 400.Google Scholar

⁵ Lewin, Linda, ‘Politics and Parentela in Paraiba: A Case Study of Oligarchy in Brazil's Old Republic,’ Diss. Columbia University 1975.Google Scholar

⁶ Political correspondence was drawn from the Arquivo Epitácio Pessoa at the Instituto Histórico e Geográfico Brasileiro in Rio de Janeiro and the Arquivos Col. Antonio Pessoa and João Pessoa at the Instituto Historico e Geográfico Paraibano in João Pessoa, Paraíba.

⁷ The officeholding population of 303 men was defined to include everyone who had been a national cabinet member, senator, federal deputy or a governor, gubernatorial cabinet member or state assemblyman. At any given time they numbered about fifty. Genealogy was elicited in oral interviews and drawn from published sources: Bittencourt, Liberato, Parahyba, vol. 2 of Homens do Brasil (Rio de Janeiro, 1914);Google Scholarde Azevêdo Bastos, Sebastiāo, No roteiro dos Azevêdo e outras familias do nordeste (João Pessoa, Paraíba, 1954);Google Scholarda Nóbrega, Trajano Pires, A familia Nóbrega (São Paulo, 1956).Google Scholar

⁸ Camilo de Holanda to Epitácio Pessoa, Capital, 13 May 1918; Arquivo Epitácio Pessoa (henceforth AEP), 12/117. The implications of family-based politics for the escalation of local conflict during the Old Republic are discussed in Lewin, Linda, ‘The Oligarchical Limitations of Social Banditry in Brazil: The Case of the "Good" Thief Antônio Silvino,’ Past & Present, 82 (02 1979).CrossRefGoogle Scholar

⁹ Willems, , pp. 339–43;Google Scholarde Queiroz, Maria Isaura Pereira, O mandonismo localna vidapolitico brasileira (São Paulo, 1969), pp. 16–18;Google ScholarCandido, Antonio, ‘The Brazilian Family,’ in Smith, T. Lynn and Marchant, Alexander, eds., Brazil: Portrait of Half a Continent (New York, 1951), pp. 297–301;Google ScholarKottak, Conrad Philip, ‘Kinship and Class in Brazil,’ Ethnology, 5:2 (1967), 428 and 442n.Google Scholar

¹⁰ Freeman, J. D., ‘On the Concept of the Kindred,’ Journal of the Royal Anthropological Society, 91 (1961), 195–99.Google Scholar Wagley has equated parentela with kindred, contrary to Freeman and Firth. See Wagley, Charles, ‘Luso-Brazilian Kinship Patterns: The Persistence of a Cultural Tradition,’ in Maier, Joseph and Weatherhead, Richard W., eds., Politics of Change in Latin America (New York, 1964), p. 175.Google Scholar

¹¹ Firth, Raymond, ‘Bilateral Descent Groups: An Operational Viewpoint,’ in Schapera, I., ed., Studies in Kinship and Marriage, Dedicated to Brenda Z. Seligman on her Eightieth Birthday, (London, 1963), pp. 30–31.Google Scholar On individual recognition in Brazil see Kottak, , pp. 432–33;Google ScholarWagley, , ‘Luso-Brazilian Kinship,’ pp. 174–77;Google Scholar and Miller, Charlotte I., ‘Middle-Class Kinship Networks in Belo Horizonte, Minas Gerais,’ Diss. University of Florida, 1976, pp. 11–12.Google Scholar

¹² Pinto, Luis de Aguiar Costa, Lutasdefamilia no Brasil (São Paulo, 1949), p. 90;Google ScholarPubMedAzêvedo, , p. 21;Google ScholarDuarte, , p. 68.Google Scholar

¹³ Firth proposed the appropriate term ‘ambilineal’ (p. 32). See also Kottak, , p. 429.Google Scholar The naming pattern originated in Willems, Portugal. E., ‘A familia portuguesa contemporanea,’ Sociologia (Sâo Paulo), 17: (1955), 24.Google Scholar For examples of differing elite sibling names, see Candido, , p. 298n,Google Scholar and Marcilio, Maria Luiza, A cidadede Sāo Paulo: Povoamento e população, 1750–1850 (São Paulo, 1974), pp. 72–73.Google Scholar The absence of a firm naming rule makes Brazilian family reconstruction very difficult. Marcilio noted it is a problem which still awaits solution or adaptation of technique. Cidade São Paulo, p. 74;Google ScholarNóbrega, , pp. 6–8.Google Scholar She has suggested that intergenerational patterns in given names of a descent group might be more useful for family reconstruction than family names: ‘Variations des nomes et prenoms du Bresil,’ Annales (1972), esp. 347, 353.Google Scholar

¹⁴ Marcilio, , Cidade São Paulo, p. 70;Google ScholarDecretos do Governo Provisorio … Brazil (Rio de Janeiro, 1890),Google Scholar Decree No. 181 of 24 Jan. 1890, Art. 56, Par. 4; Codigo Civil da Republicados Estados Unidos do Brasil… annotado pelo Dr. João Luiz Alves (Rio de Janeiro, 1917), Art. 240.Google Scholar

¹⁵ Wolf, Eric R., ‘Kinship, Friendship, and Patron-Client Relations in Complex Societies,’ in Banton, Michael, ed., The Social Anthropology of Complex Societies (London, 1968), p. 9.Google Scholar

¹⁶ Except for Governor Venancio Magalhāes Neiva, whose brother married a sister of the Baron of Lucena, four of the five gained political influence originally through connections with maternal uncles: Epitácio Pessoa, Alvaro Machado, Solon de Lucena, and Joāo Pessoa Cavalcanti de Albuquerque. See Lisboa, Joáo Goncalves, Oligarquias, séccas do none e clericalismo (Rio de Janeiro, 1909), pp. 11–12.Google Scholar

¹⁷ Firth, , p. 213.Google Scholar On the vulnerability ofbilateral descent systems to collateral fragmentation and recombination, see Davenport, William, ‘Nonunilinear Descent and Descent Groups,’ in Graburn, Nelson, ed., Readings in Kinship and Social Structure (London, 1971), esp. p. 209.Google Scholar

¹⁸ Isaura, Maria, p. 17;Google ScholarCandido, , p. 294;Google ScholarBevilaqua, Clovis, Direito da familia (Rio de Janeiro, 1896), pp. 213–14.Google Scholar Nineteenth-century prenuptial agreements regulated a couple's property as either community (Comunhāo dos bens, segundo o costume do Reino or da carta de metade), simple separation (separaçāo dos bens),Google Scholar a mixture of the two and/or dowral. In the absence of a prenuptial agreement the community property regulation prevailed. Pereira, Lafayette Rodrigues, Direito de familia, 4th ed. (Rio de Janeiro, 1945), pp. 158–69 (Arts. 56–83, 92, 96).Google Scholar

¹⁹ Brazilian civil law has always reckoned degrees of consanguinity according to Roman, not canon, law. Lineal degrees are counted by generation. Collateral degrees are reckoned by counting the generations back to a common ancestor and then down to the individual. Thus, reckoning one's uncle or aunt is two generations to a common ancestor (grandparent) and one more down in direct descent, a third-degree collateral. Bevilaqua, Clovis, Direito da familia (Rio de Janeiro, 1896), pp. 60–62.Google Scholar Except where defined as a term in civil law, all classificatory kinship terminology in this paper is drawn from family genealogies or the letters of the Paraiba Coronéis.

²⁰ Verbal usage confirms the historical weakening of cousin bonds in mid-twentieth-century Brazilian family structure. Today ‘primos filhos de irmāos’ has disintegrated to merely ‘primos irmaõs’ and is indiscriminately applied to all first cousins. ‘Coirmāos’ is already archaic.

²¹ Pinto, Costa, pp. 90, 152;Google ScholarVianna, Oliveira, I, 275;Google ScholarSchwartz, Stuart, Sovereignty and Society in Colonial Brazil (Berkeley, 1973), p. 343;Google ScholarChandler, Billy Jaynes, The Feitosas and the Sertdo dos Inhamuns (Gainesville, Florida, 1972), p. 78.Google Scholar

²² Codigo Civil 1916/Alves, Art. 183, IV and p. 158.Google Scholar

²³ This interpretation of cousin marriage and the concept of ‘kin and affines’ is drawn from Dumont, L., ‘The Marriage Alliance,’ in Goody, Jack, ed., Kinship: Selected Readings (Baltimore, 1971), pp. 183–85.Google Scholar Subsequent discussion draws on Fox, Robin, Kinship and Marriage (Baltimore, 1967) esp. pp. 175–207.Google Scholar

²⁴ Similar aspects of cousin marriage among Boston Brahmins are discussed in Hall, Peter Dobkin, ‘Family Structure and Economic Organization: Massachusetts Merchants, 1700–1850,’ in Herevin, Tamara, ed., Family and Kin in Urban Communities, 1700–1850 (New York, 1977), pp. 38–61.Google Scholar

²⁵ Cabral, Nelson Lustosa, Paisagensdo nordeste (São Paulo, 1962), pp. 129–33;Google Scholar cf. Alves, Marcio Moreira, A Grain of Mustard Seed (New York, 1973), pp. 26–40.Google Scholar

²⁶ In Portuguese law surviving spouses first acquired inheritance rights from King Pedro (1357–67) only in the absence of lineal descendants and ascendants or collaterals to the tenth degree. Ordenações do Senhor Rey D. Affonso V, (Coimbra, 1787),Google Scholar Livro III, Titulo LXXXXV. The one-third (one-half after 1907) customarily assigned to wives was not theirs by legal right as a surviving spouse, as is commonly assumed. It was bequeathed by the husband as that part of his estate over which he had testamentary freedom, the other two-thirds (a legitima) being assigned inalienably by law to his lineal ‘necessary heirs.’ Ord. IV, 94;Google ScholarBevilaqua, , p. 213; Decree No. 1839 of 31 Dec. 1907, Art. 1. Codigo Civil 1916/Alves, Art. 1576. Filial challenges were eliminated by Art. 1725.Google Scholar

²⁷ Brazil. Decretos do Governo Provisorio … (jan.-abril de 1890) (Rio de Janeiro, 1890). Decree No. 181 of 24 Jan. 1890, Art. 58, Par. 1–3.

²⁸ One hundred members of the officeholding elite had identifiable wives. Fifteen percent married cousins and 14 per cent did not. The remaining 71 percent could not be identified conclusively in terms of kinship. Twenty-nine percent of the officeholding elite had identifiable sets of parents. Thirty, or about one-third, married cousins or nieces, but thirty-two, or slightly more than one-third, did not. The kinship of the remaining third could not be verified or excluded. Darrell E. Levi discovered a coeval decline in endogamous marriage among the Prados of São Paulo; A familia Prado (São Paulo, 1977). pp. 103–05.Google Scholar

²⁹ The ratio of Nóbregas who did not marry jumped from 7.5 percent in the fourth generation to 24 percent in the fifth generation (or 19 percent if those with unknown spouses are included), suggesting that both patrimony and land were much less available by the close of the last century. Fifteen hundred and seven Nobregas lived to adulthood: 13 in the second generation, 118 in the third, 381 in the fourth, and 995 in the fifth. Those without identifiable spouses and those who did not marry (a combined total of 387) were excluded from Tables 1–3. (The sex distribution of the unmarried was nearly even.)

³⁰ The probable effect of the distribution of consanguineous marriages shown in Table 3 corresponds to an average value of shared genes in the population of the third (0.0270) and fourth (0.0260) generations close to that of a first cousin once removed (0.03125). The fifth generation average inbreeding coefficient (0.0178) fell closer to that of a second cousin (0.0156), i.e., a sixth-degree collateral. Coefficients were derived from Salzano, Francisco M. and Maia, Newton Freire, Problems in Human Biology (Detroit, 1970), pp. 82–83.Google Scholar See ch. 4 for an extensive treatment of consanguineous marriage in Brazil from a geneticist's perspective.

³¹ A Paraiba historian has observed regarding nineteenth-century endogamy: ‘Consanguineous unions were frequent, above all, between uncles and nieces, not only because of prejudice about white skin but also because of the fear of allowing an outsider to enter the intimacy of the family group.’ de Almeida, Horácio, Brejo de Areia (Rio de Janeiro, 1958), p. 217.Google Scholar

³² Freyre, Gilberto, Sobrados e mucambos, 4th ed. (Rio de Janeiro, 1968), I, 129–33;Google Scholarde Almeida, E., pp. 139–46;Google Scholarde Almeida, H., pp. 216–25;Google ScholarNobrega, , pp. 578–80.Google Scholar

³³ Decree of 31 Oct. 1831, Colleção das leis do Imperio do Brazil de 1831, 1^a pie. (Rio de Janeiro, 1875).Google Scholar

³⁴ Decree No. 181 of 24 Jan. 1890, Arts. 17, 23–46; Codigo Civil 1916/AIves, Art. 213.Google Scholar

³⁵ Inácio Dantas to Epitacio Pessoa, Teixeira [Paraiba], 18 Jan. 1912; AEP 10/81.

³⁶ The literature reiterates the same set of dramatic examples: cf. Freyre, , Sobrados, pp. 69–70;Google ScholarPinto, Costa, pp. 56, 185–91;Google ScholarCandido, , p. 295;Google ScholarAraripe, Alencar, ‘Pater-familias nos tempos coloniais,’ Revista do Instituto Histórico e Geográfico Brasileiro, LV, Pte 2 (1893): 15–23.Google Scholar The legal basis for physically punishing a wife (Ord., L. 5, Tit. 36, Par. 1,36, and L. 95, Par. 4) was struck down by the Brazilian Criminal Code of 1831 that superseded the criminal provisions of the Philippine Code. Rodrigues Pereira, p. 133. The father was permitted to punish minor children moderately. Ibid., pp. 275–76.

³⁷ Candido, , p. 303;Google ScholarPinto, Costa, pp. 79–89.Google Scholar

³⁸ Nóbrega, , p. 146.Google Scholar See also, Freyre, , 1, 95–108 and 140.Google Scholar

³⁹ The ‘loophole’ in the law was the provision permitting distribution of portions of the legitima as gifts prior to the testator's death. The colação facilitated the legal return of such portions to the deceased's estate posthumously. Codigo Civil 1916/Alves, Art. 1785. The distribution of urban real estate to heirs by the testator in contemporary Brazil is discussed in Miller, , pp. 176–77;Google ScholarDuarte, , p. 69.Google Scholar

⁴⁰ Entailed heirs (morgados) were prohibited by Law No. 57 of 10 Oct. 1835. John Armitage suggested, however, that the entailed estate (morgadio) never took hold in Brazil. Historia do Brasil desde o período da chegada da familia de Bragança em 1808 até a abdicação de D. Pedro l^o em 1831, 3rd ed. (Rio de Janeiro, 1933), I, 50–51; II, 8, 16.Google Scholar

⁴¹ Freyre contends only sons of inferior intelligence and poor health returned to the patriarchal domain. Sobrados, I, 18.

⁴² Vianna, Oliveira, I, 263, 295–338.Google Scholar

⁴³ Lewin, , Politics, p. p.Google Scholar

⁴⁴ Decree No. 1839 of 31 Dec. 1906, Art. 1; Codigo Civil 1916/Alves, Art. 1603. The legal range of the kindred in Brazil has been progressively restricted so that it is now considerably narrower than in the United States. Decree-Law No. 9.461 of 15 July 1946 limited inheritance rights to collaterals within the fourth degree (first cousins), in whose absence the estate escheats to the federal government.

⁴⁵ Pinto, Luiz, Sintese historica da Paraiba, 2nd ed. (Rio de Janeiro, 1960), pp. 113–20;Google ScholarJuliāo, Francisco, Cambão—The Yoke (Middlesex, England, 1972), pp. 76–79, 132–35;Google Scholarde Carvalho, Jose Murillo, ‘Barbacena: A família, a politica e um hipotese,’ Revista Brasileira de Estudos Politicos, 20 (1966), 153–94;Google Scholarde Andrade, Manuel Correia, A terra e o homem no nordeste, 2nd ed. (Sao Paulo, 1964), p. 203;Google ScholarSoares, Glaucio Dillon, Sociedade e politica no Brasil (Sāo Paulo, 1973), esp. pp. 97–135.Google Scholar

⁴⁶ Miller, , pp. 211–17;Google ScholarWagley, , ‘Luso-Brazilian Kinship,’ p. 174;Google ScholarMaeyama, Takashi, Familialization of the Unfamiliar World: The Familia, Networks and Groups in a Brazilian City (Ithaca, New York, 1975), pp. 32–34, 138–40;Google ScholarBerlinck, Manuel Tosta, The Structure of the Brazilian Family in Sāo Paulo (Ithaca, 1969), pp. 137–38.Google Scholar