[Note: this trAnscription wAs produced by An AutomAtic OCR engine]

TABLE or CONTENTS

Foreword......... . . . . . . . . . . . . . .

I.Introduction................................................10

I.A The four-section system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

I.b Section systems And history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22

Le Homogeneity And heterogeneity in the Western Desert . . . . . . . . . . . . . . . . . . . . . . . . 30

I.d TheoreticAl considerAtions on the question of diffusion . . . . . . . . . . . . . . . . . . . . . . .32

I.e MethodologicAl preliminAries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35

II. The spAtiAl distribution ofindividuAl terms . . . . . . . . . . . . . . . . . . . . . 38

III.ThespAtiAldistributionofpAirsofterms.......................42



III.A First hypothesis: the identity PAnAkA - YipArrkA . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43

III.b Second hypothesis: the zone of extension And substitution of PAljeri . . . . .

IH.c Third hypothesis: Burgulu mAy be replAced by MilAngkA or YipArrkA . . . . . . . . . . .45

IH.d CombinAtions of four cAtegories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
IV. The spAtiAl distribution of logicAl systems . . . . . . . . . . . . . . . . . . . . . . 52

NA Impossible mArriAges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52

IV.b The Pintupi, or the trAnsition fl-om four sections to eight subsections . . . . . . . . . . . .53

IV.c NyAngumArtA: the inlAnd And the coAstAl system . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

IV.d The cAse of LAverton: first encounter between WAljen And MAndjmdjA . . . . . . . . . . 60

We The cAse of KAlgoorlie: second encounter between WAljen And MAndjindjA . . . . . . . 62

V. The diffusion of section nAmes: An overview . . . . . . . . . . . . . . . . . . . . . 65

VI. Origin of section nAmes And lAnguAge or tribAl nAmes. . . . . . . . . . . . 67

VH. DistributionAl pAttern of AlternAte generAtionAl moieties. . . . . . . . . 71

VIII. Why And how did sections diffuse? A network ApproAch . . . . . . . . 78
VIII.A How did sections difliise; how were they Adopted? . . . . . . . . . . . . . . . . . . . . . . . . 78
VIII.b Why were sections Adopted; why hAve they difliised? . . . . . . . . . . . . . . . . . . . . . . 82

IX.Conclusion................................................90



A StAtisticAl tools used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
b Stories explAining the AnivAl of sections in the RAwlinson And WArburton RAnges . . 95
c Corpus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97

Referencescited..............................................101