In most European languages the words for ‘brother’ and ‘sister’ are quite distinct, but in a few languages – notably Spanish (hermano, hermana), Portuguese (irmao, irma), and Greek (adelfos, adelfi) they share a common root and differ only in the gendered ending. The idea for this workshop arose from an email discussion between the organisers about whether the distinction between the two morphological patterns (hermano, hermana vs. brother, sister) had a social meaning. Partly on the basis of previous work (Kronenfeld 1991; Heady 2013) we thought there should be some connection with social practice, but if so, what was it?
An obvious suggestion, following Radcliffe-Brown, is that using (almost) the same term implies that relationships to one’s male and female siblings are similar, while the use of radically different terms implies that the relationships are very distinct. But this comes to grief over the fact that the ‘similar’ sibling terms are found in southern European countries where cultural ideas of ‘honour and shame’ traditionally led to particularly sharp distinctions between same-sex and cross-sex relationships. We tried a number of other hypothetical terminology-practice connections, and as we explored their implications the discussion broadened out to include the terms for other primary and secondary relatives and turned into a general review of patterns of gender-marking and kin-term morphology in European languages. Although the conclusions of the review are rather tentative, the material does suggests a number of associations – both with aspects of social practice and with the grammars of European languages.
The details of the argument are presented in the concluding dialog in the distribution packet for the workshop. The reason we think they matter is that the associations we tentatively identified cut across the criteria which define the classic set of terminological types. In one guise or another, these types (including Morgan’s distinction between Descriptive and Classificatory terminologies, and the main varieties of each kind: Eskimo, Sudanese, Hawaiian, Dravidian, Iroquois, Crow, Omaha, and Cheyenne) have provided a common framework for almost all ethnological (i.e., comparative) approaches to kinship terminology – starting with Morgan himself and including Murdock, Levi-Strauss, and Lounsbury – especially approaches that involve relating the kinship terminology to features of social organization. There is no doubt that the research carried out within this framework has been highly productive – generating robust empirical generalisations which will stand the test of time.
Even so, the trouble with frameworks is that they can easily become cages: by directing attention towards some aspects of a phenomenon they inevitably downplay others which may be equally important. Typologies have a further disadvantage: refining the typological categories and extending them to allow for special cases can become an end in itself – and this focus on “butterfly collecting” can distract from the central task of identifying the underlying causal principles that structure the field as a whole.
We are aware of at least three important aspects of terminology, which are reflected in our example, but which mainstream typological analyses have tended to overlook:
- The morphological form of kinship terms themselves. For instance, are they stand-alone lexemes, lexemic roots with grammatical inflexions, or composites of two or more equally meaningful elements? To the extent that the terms themselves are structured, they suggest the possibility of systematic meanings that go beyond the simple distinctions and equivalences required for the Morganian typologies.
- The ways in which kin terms or their usage encode information that is not strictly genealogical, particularly the relative age and gender of Ego and Alter.
- The relationships referred to. The standard Morganian typologies are mainly defined by their treatment of various kinds of cousins, aunts and uncles—or more recently by formal algebraic systems which are able to generate these distinctions. Sibling terms are not specifically treated, despite forming several clearly defined structures of their own. Other relationships which are drawn on to define particular types, but not comparatively analysed in their own right, include affinal, grand-parental, and parent-child links.
Source and further information: Max Planck Institute for Social Anthropology, Link (8 April 2019)