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

45

not enter the desert, where Milangka replaces it, and below 27° of latitude Tjarurru replaces
it. The extension of “Paljeri --> Milangka : Tjarurru” is what I have been calling a valeur afa
section, that of Paljeri in this case.

What happens outside this Paljeri zone? Two hypotheses may be formulated. The first
one would be to state that the valeur of Paljeri is, simply, absent there. This hypothesis easily
resolves the problem, but contradicts the basic hypothesis according to which the diffusion
of section names is accompanied by an underlying logic the general diasystemic map that
organises the relations between them. The second hypothesis is, consequently, that the valeur
of Paljeri is present, but that the term has been replaced by another category: the valeur
changed content.

With the exception of Ooldea 1, all groups outside the zone “Paljeri --> Milangka :
Tjarurru” possess the three categories that may potentially replace Paljeri: Milangka, Tjarurru
and Yiparrka. To begin with Ooldea 1, the Berndts 1942-45 claim that this system moved to
Ooldea from the west, probably with the Waljen. The Waljen are part of the sub-zone “Paljeri
--> Tjarurru”. Ooldea 1 is therefore an outgrowth, related to migrations, of the zone “Paljeri
--> Milangka : Tjarurru”. In summary, we have, for the diffusion and replacement of Paljeri,
a first route alongside the coast fi'om north to south, and a second route from west to east that
borders the desert along its southern limit. We may, therefore, add to the linguistic routes of
diffusion of Paljeri proposed by McConve1l l985a:11 a “logical” or statistical route, or a
route of the “va1eur of Paljeri”, which equally leads to central Australia, but goes through the
southern part of the region considered here Map 8.

:-

McConvell’s linguistic
thesis and the present
“logical” thesis

Paljeri
2 McConve||’s linguistic
thesis

3 Present “logical”
thesis





us m ma :2: in us m In us: is: us my in

Map 8: Schematic “linguistic” and “logical” routes of the diffusion of Paljeri

III.c Third h_wathesis: Burgulu may be replaced by Milangku or Yiparrka

Bates 1985: 102 suggests that in the Upper Murchison, that is, among the Ngarlawonga,
Burgulu would replace Panaka. If this assertion is correct for the Ngarlawonga, the Ngaiawonga
and the Koara, two examples call this rule into question. The first is the reappearance of
Panaka south-east of the Koara, among the Waljen; the second is the joint presence ofBurgu1u
and Panaka among the Yulbaridya.