[Note: this transcription was produced by an automatic OCR engine]

Foreword

This study is an historical and structural investigation of the difliision of taxonomic
elements of the section system as known among dialectal groups of the Aust:ralian Western
Desert. What a section system is and where the Western Desert is located will be developed
in some detail in the following chapters. Let me explain why and how the present study came
about.

The initial aim was to develop a methodology for investigating the variation, internal
coherence and pattern of diffusion of kinship terminologies in the Western Desert. These
terminologies are characteristic for both their homogeneity and their regional variational
pattern, and they appear to be ideal-types for large-scale geographical comparisons, as well
as for reconstructions of “proto-systems” or “proto-terminologies”. Social categories, in
this case sections, naturally presented themselves as the initial test-corpus for such a study.
Lexical variations, as well as the overall complexity of relations expressed through the
semantic positions of these variations, are far less a feature of section systems than they are
of kinship nomenclatures. Studying local and regional terminological altemafives of the four
basic categories in the system is a far less complex enterprise than undertaking the same for
20 or more kinship classes, not to mention their variations, which are not solely of a lexical
but also of a structural and semantic nature.

The complexities of variational kinship tenninologies are well illustrated in the example
of way'irra, a word that throughout the Western Desert cultural bloc has some meaning
associated with “cross-cousin”, that is, a real or classiflcatory mother’s brother’s child or
father’s sister’s child However, while this term may be used by some dialectal groups, others
may only understand it; it may refer to a close cross—cousin among some groups and to a
distant and potentially marriageable cross-cousin among others; and it may be applicable
to both females and males among some groups, but only to males among others Dousset
20023. It is evident that the analysis of the variations—and of their underlying principles—
among 40 or more dialects, each with 20 kin classes or more, would be a complex enterprise.
Although closely linked to the structure of the kinship system, sections distinguish only one
gender, REflECT only a limited number of semantic variations, can be analysed, in principle at
least, without taking into account their pragmatic application, and are of limited number with
only four basic categories in each arrangement. The feasibility of a linguistic, historical and
anthropological study of the origins and diffusion of the section system in the Western Desert
is greatly enhanced by this apparent simplicity.

The results of this preliminary study, investigating pan-regional relationships of taxonomic
usages in relation to the section system, revealed it to be far more interesting than expected.
One reason is that some results of the reconstruction of the section system teIminclogy’s routes
of diffusion in the Western Desert corresponded surprisingly well with linguists’ independent
conclusions on similar questions, in particular those of Patrick McConvell. The second reason
derives from the first. Indeed, if the general conclusions are consistent with those gained