Panoan Marriage Sections: A Comparative Perspective

Abstract

Almost 50 years ago, George Peter Murdock (1949:56) could assert that systems of marriage sections had never been reported outside of Australia and a limited area in Melanesia. Since then, ethnographers of several Pano-speaking peoples of southeastern Peru have given us reason to revise Murdock's assessment. The Cashinahua, Sharanahua, and Mayoruna have apparently maintained systems of social classification which strongly resemble the classical, four-section system of the Australian Kariera (Kensinger 1984b; Siskind 1973:55,62; Fields and Merrifield 1980). This article discusses these Panoan systems in light of comparative Amazonian ethnography, and shows how they can be understood as a way of alleviating cognitive inconsistencies generated in the articulation of emergent, exogamous descent categories and cross-cousin marriage in small, endogamous groups. Viewed as one of several possible strategies for avoiding such inconsistencies, Panoan marriage classes may shed some light on the emergence of section systems in general. During the past two decades, Dumont's (1953) two-line interpretation of Dravidian terminologies from southern India has been influential in shaping our understanding of social classification in lowland South America (Riviere 1973; Overing Kaplan 1973; Kensinger 1979, 1984a; Schwerin 1982). The defining feature of a Dravidian system is that the terms for cross-collaterals are the same as those for affines (cf. Buchler and Selby 1968:233; Shapiro 1984:2). Two-line terminologies have repeatedly been dissociated from the phenomenon of unilineal descent, with which they need not but apparently may co-exist. The kin-affine division of Dravidian systems is founded on principles distinctly different from those underlying sibs or moieties. Although they are equivalent in terms of the zero-generation opposition of parallel-cousins and cross-cousins, in the first ascending generation the two modes of classification will disagree (Shapiro 1970; Trautmann 1981:176-177). Dumont's (1953) model of the Dravidian system is implicitly cognatic in that it groups both F and M in supra-category of whereas both MB and FZ are terminological affines. A unilineal principle, of course, would oppose F and FZ, on the one hand, against M and MB, on the other. In other words, a Dravidian terminology appears to classify FZ as affine, while in a patrilineal descent system she is unquestionably a kinswoman. My question is: how do societies which allegedly feature both a Dravidian terminology and patrilineal descent categories resolve this contradiction? Though a kin-affine dichotomy is fundamental to both unilineal descent categories and Dravidian terminologies, the contradiction between these two ways of delineating it seems difficult to reconcile. I suggest that the Panoan marriage class system, by restricting the recognition of kin-affine dichotomies to (1) alternate generations and (2) same-sex relatives only, represents a reconciliation, perhaps even a stepping stone, between the cross/parallel logic of Dravidian classification and unilineal descent. Shapiro (1970) demonstrates how two-section terminologies will exhibit different structures if associated with two named super-categories (such as the exogamous moieties of the Kariera and other Australian systems) than if it has no corresponding sociocentric basis, as among the Beaver Indians of British Columbia, who have a Dravidian, two-line terminology but no moieties (cf. Ridington 1969). Whereas the former include FZ in the lineal (i.e., kin) and M in the affinal category, the latter reverses their positions. Shapiro (1970:386) observes that a confusion of these two types pervades much of the literature on two-section systems. Keesing (1975:108) has subsequently illustrated this observation by including FZ = WM in the Dravidian category of kin, while classifying M as affine(!) In the same vein, Buchler and Selby (1968:135) erroneously conclude that Dravidian systems are patrilineal (cf. …