W. Shapiro Secondary Unions and Kinship Terminology: the Case of Avuncular Marriage

W. Shapiro Secondary unions and kinship terminology: The case of avuncular marriage

In: Bijdragen tot de Taal-, Land- en Volkenkunde 122 (1966), no: 1, Leiden, 82-89

This PDF-file was downloaded from http://www.kitlv-journals.nl

Downloaded from Brill.com09/29/2021 01:57:01AM via free access SECONDARY UNIONS AND KINSHIP TERMINOLOGY: THE CASE OF AVUNCULAR MARRIAGE •

his paper is divided into two parts. The first is a purely ethnographic statement of the indications and influences of Tavuncular marriage (marriage of a man with his sister's daughter, a woman with her mother's brother) in the kinship terminologies of various societies. The second part is a theoretical discussion of certain issues concerning the relation between marriage patterns and kinship terminology.

A terminological equation of two (or more) kin-types can be said to be indicative of, or "due to", a certain genealogically defined type of marriage if two conditions are satisfied: (1) the kin-types occupy a single position in an ideal marriage network defined solely by the marriage-type in question; (2) no other marriage-type yields an ideal network in which these kin-types occupy a single position.2 Thus the equation FZ = MBW indicates sister-exchange marriage, since father's

1 This paper has grown out of a larger comparative study of avuncular marriage. I should Hke to express my thanks to those who have read and commented upon this study in its various manuscript forms: Ann Chowning, Meivin Ember, Robert F. Murphy, and W. E. H. Stanner. 2 These propositions are put forward merely as guiding principles for what follows. I am well aware that there are many societies with positive marriage rules in which considerations of genealogical connection are entirely outweighed by those of descent group and/or kinship category, and that the uneven character of sociocultural change often makes for situations in which kinship equations reveal nothing about, and cannot be ascribed to, current marriage patterns. Further, there is probably no marriage-type which is always in a simple relationship with a particular kinship equation; hence the attribution of an equation to a marriage-type should be supported by additional evidence, preferably direct information that the marriage-type exists in the society or societies under consideration.

Downloaded from Brill.com09/29/2021 01:57:01AM via free access SECONDARY UNIONS AND KINSHIP TERMINOLOGY. 83 sister would always be the wife of mother's brother if every man married his sister's husband's sister, and since this would not hold if any other marriage-type were exclusively practiced. With these points in mind, we may consider the terminological reflections of avuncular marriage in several societies. (Fig. 1 should be consulted throughout this section). Probably the most common reflection of avuncular marriage on kinship terminology is the equation ZC = MBC; this is found in the kinship systems of at least the following societies: the Barama River Caribs of British Guiana (Gillin 1936), the Bhumia and certain Gond groups in central India (Fuchs 1960, Karve 1953:240-47), the Gururumba of the New Guinea highlands (Newman 1965), the Kanarese of South India (Beals 1962, Karve 1953:231-35, McCor- mack 1958), the Mundurucü of central Brazil (Murphy 1960), the Nama of Southwest Africa (Hoernlé 1925, Schapera 1930), the EGO Sirionó of eastern Bolivia (Holmberg 1950), the Tukuna of northwestern Brazil (Nimu- endajü 1952), and the now-extinct Tupi- namba of Brazil's Atlantic Coast (Fernandes 3 1963). It is reported also for the Ngombe Fig. 1: An avuncular mar- of the northwestern Congo (Wolfe 1961), riage system. The positional identity of ZS and MBS is though here the term used for ZS and MBC shown; the reader is asked is applied to several other kin-types which to tracé the other identities would not occupy the same position in a mentioned in thé text for himself. system in which every man married his sister's daughter; moreover, avuncular marriage is reported to be prohibited, and there is no reason to believe that it was ever allowed.4 In all of these systems except Mundurucü, Ngombe, and Sirionó, the MBC-ZC term is also applied to FZC and to no other consanguineal

3 Actually, the equation holds for all four kin-types concerned only in the systems of the Gururumba, Kanarese, Mundurucü, Nama, and Tukuna. The Barama River, Bhumia, Gond, and Tupinamba systems have the equation ZD (m.s.) = MBD, but have separate terms for ZS (m.s.) and MBS; the Sirionó system has the equation ZS (m.s.) = MBS, but has separate terms for ZD (m.s.) and MBD. 4 Much the same can be said for the Sirionó. Here the term for ZS (m.s.) and MBS is applied to several other consanguineal kin-types and avuncular marriage is reported to be prohibited; in this case, however, there ir reason to believe that it is actually practiced, or at least once was (cf. below).

Downloaded from Brill.com09/29/2021 01:57:01AM via free access 84 WARREN SHAPIRO. kin-type, indicating the existence of a "two-section" marriage system in which the sister's daughter and both cross-cousins are possible mates but unions between parallel relatives (at least close ones) are prohibited. Of the other equations reflecting avuncular marriage, the only ones which — to my knowledge — occur in at least two kinship systems are MB = FZS — reported for the Kanarese, Mundurucu, and Sirionó, and FZ = MM — reported for the last two as well as for the Nambi- kwara of central Brazil (Lévi-Strauss 1948). The remaining equations of present relevance are listed below: the first is found in the kinship terminology of the Orokaiva of Papua (Williams 1930); the others are reported for the Kanarese by McCormack (1958):

WMB = ZDH = MBDH FF = MMF FM = MMM ZDH = B (m.s.) ZS = WB ZD = WZ

Our final terminological indication of sister's daughter marriage — reported for the Cayua of Paraguay and the adjacent parts of Brazil (J. Watson 1952, V. Watson 1944) — does not take the form of a kinship equation. The Cayua kinship system formerly distinguished between ZD (m.s.) and ZS (m.s.), but not between BD (f.s.) and BS (f.s.); moreover, the terms for S (= BS) (m.s.), D (= BD) (m.s.), and ZS (m.s.) were phonetically very similar but differed sharply from the term for ZD (m.s.). Even today, when avuncular marriage is no longer allowed, ZD (m.s.) is accorded a special term, though a Hawaiian pattern prevails otherwise in the first descending generation: "Why, then, a special term for sister's daughter, if not probably to express a special relationship ?" (J. Watson 1952:36). Before concluding this section, it will be necessary to consider the direct evidence for avuncular marriage in the societies mentioned. Such evidence is unequivocal for the Barama River Caribs, Cayua (in the past), Gond (but not all groups), Kanarese, Mundurucu (in the past - cf. Horton 1948:277), Nama, Nambikwara, Orokaiva, and Tupinamba. Marriage with the sister's daughter is neither reported nor denied for the Bhumia, Gururumba, and Tukuna; for the Sirionó it is expressly denied, albeit not in as precise a fashion as one might wish: Holmberg (1950:64) states that sexual relations between "uncle and niece" are prohibited. However, there is good reason to believe that the institution in question is either practiced today by the Sirionó or else once was: the terminological equations noted above for this group are too sug- gestive to ignore; moreover, avuncular marriage is reported for several

Downloaded from Brill.com09/29/2021 01:57:01AM via free access SECONDARY UNIONS AND KINSHIP TERMINOLOGY. 85 other Tupi-Guarani peoples (e.g., the Cayua, Mundurucu, and Tupi- namba, all noted above).5

"A primary marriage is typically the first union which an individual contracts; a secondary marriage is any subsequent union" (Murdock 1949:28-29). We may also regard a genealogically defined type of marriage as "secondary" in a given society if it occurs rarely, and/or less frequently, than another marriage-type. Thus, marriage with the wife's brother's daughter is, in a given society, "secondary" in both senses if it occurs less commonly than marriage with the mother's brother's daughter. I suspect that in at least most societies with positive marriage rules, a marriage-type which is secondary from the standpoint of the life-cycle of the individual is also statistically secondary, though the reverse is certainly not true. For most of the peoples mentioned in the previous sectdon for which the appropriate information is available, it can be shown that avuncular marriage is secondary in at least one of these senses. Thus in those societies in which the members of both sexes generally marry at approximately the same age, it is unlikely that many men will be in a position to take^their sister's daughter as a first wife. Such a lack of age differential between the sexes at marriage is reported for the Gond, Bhumia, and Tukuna.6 This, however, is definitely not the case (i.e., such a differential is reported) for the Kanarese, Nambikwara, and Tupinamba, men being considerably older than their wives in all three of these societies.7 Nevertheless, for one Kanarese village at any

5 Dunning (1959:167) has reached the same conclusion concerning the Sirionó, though on different grounds. I might add that our conclusion is at odds with Needham's analyses of the Sirionó kinship and marriage systems (Needhatn 1961,1964); thus — to take only one point of conflict—Needham believes the Sirionó to be organized into exogamous matrilineages, a situation which would make avuncular marriage impossible. I have presented orally a re-analysis of Holmberg's material (Shapiro 1965) in which these and other points are made; I find that the influence of avuncular marriage on Sirionó society and culture is much more marked than has been indicated here. Unfortunately, I have not had sufficient time as yet to préparé this study for publication; it is, however, available in mimeograph form to interested parties. 0 Only as an ideal, however. Regrettably, there is no statement of acutal tendencies. 7 Regarding the Sirionó, Holmberg (1950: 82) states: "The age requirements for marriage are very elastic." For the other groups, no information of present relevance is available.

Downloaded from Brill.com09/29/2021 01:57:01AM via free access 86 WARREN SHAPIRO.

rate, McCormack (1958:35-36) found that only 9.8 per cent of the marriages in one of his samples was avuncular; even this modest figure is larger than the actual proportion of unions with the sister's daughter, since the sample includes only marriages between cognates and the figure just given pertains to unions with females of the kin-type FBDD ("classificatory sister's daughter" in the Kanarese system) as well as to those with females of the kin-type ZD. For the Nambikwara, Lévi- Strauss (1943: 399) — although he does not present figures — makes it clear that avuncular unions are markedly infrequent. Finally, for the Barama River Caribs — for whom information on marriage-age is lacking — Gillin (1936:95) reports only two cases of sister's daughter marriage. Avuncular marriage is thus frequently — if not usually — a secondary type of marriage both ontogenically and statistically. Now Murdock has contended that marriage-types of this kind do not have important effects upon kinship terminology. To quote him: "Customs of preferential marriage, in the opinion of the present writer, are likely to influence kinship terminology when they apply to all or most marriages within a society, but not when they apply only to occasional unions or to a distinct minority of all that take place" (1949:123); "... preferred secondary marriages of whatsoever type are not to be included among the significant determinants of kinship terminology" (1949:177). Murdock's first remark is not accompanied by supporting evidence — it is purely an obiter dictum. On closer scrutiny it appears incredible, for simple demographic reasons. Let us take, e.g., that much-discussed kin-type, MBD: what proportion of the men in a society even have a representative of this kin-type, let alone a marriageable one? (I, for one, do not have one, since neither of my mother's brothers was kind enough to sire a daughter for me). The question may thus be raised as to whether there is a single society in which marriages between two particular kinds of relatives, genealogically defined, constitute "all or most" of the unions in that society. I think not, and I respectfully challenge Professor Murdock to name even one such society. At any rate, the material presented in the preceeding section, coupled with the statistics given above, directly invalidate his first position.8

8 To stipulate that unions with both "actual" and "classificatory" relatives of a particular kind be taken into account in defining marriage-types is, of course, to beg the question of kinship terminology; hence this strategy cannot be used to defend the Murdockian position. And it is only very reecntly that a technique has been developed which enables one to apply close kin-type distinctions to a wider sphere without taking kinship terminology into account (cf. Kay 1965).

Downloaded from Brill.com09/29/2021 01:57:01AM via free access SECONDARY UNIONS AND KINSHIP TERMINOLOGY. 87

The second of Murdock's quoted remarks is based upon a statistical study of the association between various features of kinship terminology on the one hand and the levirate, sororate, and marriage with the mother's brother's wife and wife's brother's daughter on the other; the generally low "Chi squared" values of the correlations obtained suggest that these marriage-types are not important determinants of kin-terminological patterns. However, Murdock is not justified in applying this conclusion to all forms of ontogenically secondary marriage, as the previous section shows for the sister's daughter type of union. There are, to be sure, societies in which avuncular marriage is practised without having repercussions on kinship terminology: the Kamaiura of central Brazil (Galvao 1953, Oberg 1953), the Tswana of Bechuanaland (Schapera 1950), and the Waiwai of British and Brazilian Guiana (Fock 1963) may be cited in this connection. Never- theless, unless it can be shown that the terminological features noted in section I stem from other factors — and I cannot see how this can be done — we must acknowledge the significance of at least one kind of ontogenically and statistically secondary marriage in influencing the classification of kin. WARREN SHAPIRO

The Australian National University Canberra, Australia

BIBLIOGRAPHY Beals, Alan R. 1962 Gopalpur; a South Indian Village. New York. Dunning, R. W. 1959 Social and Economie Change among the Northern Ojibwa. Toronto. Fernandes, Florestan 1963 Organizacjio Social dos Tupinamba. Second edition. Sao Paulo. Fock, Niels 1963 Waiwai: Religion and Society of an Amazonian Tribe. Copenhagen. Fuchs, Stephen 1960 The Gond and Bhumia of Eastern Mandia. Bombay.

I think it can be shown that most of those who have written about the relation between marriage patterns and kinship terminology have considered the former in terms of the close genealogical sphere of primary, secondary and tertiary relatives.

Downloaded from Brill.com09/29/2021 01:57:01AM via free access 88 WARREN SHAPIRO.

Galvao, Eduardo 1953 Culture e Sistema de Parentesco das Tribos Alto Rio Xingü. Boletim do Museu National (Rio de Janeiro), No. 14. Gillin, John 1936 The Barama River Caribs of British Guiana. Papers of the Peabody Museum, Vol. 14, No. 2. Hoernlé, A. Winifred 1925 The Social Organization of the Nama Hottentots of Southwest Africa. American Anthropologist 27: 1-25. Holmberg, Allan R. 1950 Nomads of the Long Bow: the Sirionó of Eastern Bolivia. Smithsonian Institution, Institute of Social Anthropology, Publication No. 10. Horton, Donald 1948 The Mundurucu. Bulletin of the Bureau of American Ethnology No. 143:3:271-82. Karve, Irawati 1953 Kinship Organisation in India. Poona. Kay, Paul 1965 A Generalization of the Cross/Parallel Distinction. American Anthropo- logist 67:30-43. Lévi-Strauss, Claude 1943 The Social Use of Kinship Terms among Brazilian Indians. American Anthropologist 45:398-409. 1948 La Vie Familiale et Sociale des Indiens Nambikwara. Journal de Société des Américanistes (Paris) 37:1-131. McCormack, William 1958 Sister's Daughter Marriage in a Mysore Village. Man in India 38: 34-48. Murdock, George P. 1949 Social Structure. New York. Murphy, Robert F. 1960 Headhunter's Heritage: Social and Economie Change among the Mundu- rucu Indians. Berkeley. Needham, Rodney 1961 An Analytical Note on the Structure of Sirionó Society. Southwestern Journal of Anthropology 17: 239-55. 1964 Descent, Category, and Alliance in Sirionó Society. Southwestern Journal of Anthropology 20: 229-40. Newman, Philip L. 1965 Knowing the Gururumba. New York. Nimuendaju, Curt 1952 The Tukuna. University of California Publications in American Archae- ology and Ethnology 45: 1-209. Oberg, Kalervo 1953 Indian Tribes of Northern Mato Grosso, Brazil. Smithsonian Institution, Institute of Social Anthropology, Publication No. 15.

Downloaded from Brill.com09/29/2021 01:57:01AM via free access SECONDARY UNIONS AND KINSHIP TERMINOLOGY. 89

Schapera, I. 1930 The Khoisan Peoples of South Africa. London. 1950 Kinship and Marriage among the Tswana. In A. R. Radcliffe-Brown and Daryll Forde, eds., African Systems of Kinship and Marriage. London. Shapiro, Warren 1965 A Re-examination of Sirionó Sociology. Paper read at the Department of Anthropology seminar. The Australian National University. Mimeo. Watson, James B. 1952 Cayua Culture Change. Memoirs of the American Anthropological Asso- ciation, No. 73. Watson, Virginia D. 1944 Notas sobre o Sistema de Parentesco dos indios Cayua. Sociologia (Sao Paulo) 6:31-48. Williams, F. E. 1930 Orokaiva Society. London. Wolfe, Alvin W. 1961 In the Ngombe Tradition. Evanston.

Downloaded from Brill.com09/29/2021 01:57:01AM via free access