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

49

became patrilineal filiation, and vice versa. Tjarurru replaced Paljeri, but the child of
a Tjarurru woman substituted for Paljeri is now Purungu, marrying Karimarra.“

4 Mardu, Koara—>Ngaatjatjai1'a, Ngaanyatiarra: Here we find a diffusion of a
combination of the Mardu and the Koara system: Purungu marrying Milangka, and
Purungu marrying Karimarra. The combination results in an assimilation of Karimarra
and Milangka. We shall see later that Karimarra has probably progressed into the
desert only from the south-west, by way of the Koara, in common with Tjarurru and
Yiparrka.

Again, the Yulbaridya case presents a twofold problem‘ First, this is the only group
where Burgulu and Panaka are jointly present. Second, it is not Milangka but Burgulu that
replaces Paljeri among the Yulbaridya cf. McKelson 1980.” To understand this complexity,
it is necessary to extend the field of combinations of categories to analyse not only the joint
appearance of two categories, but of four. A flrst solution is, however, already available: two
logics of combinations of section names have been progressing following two principal routes.
To conclude the discussion of this hypothesis III.c, I represent them in space Map 9.

1: Paljai --> h/fllangka : Tjanmu
Panaka —-> Bnrgulu
2: Paljni -A> Tjanlml
Burgulll --> Yiparrka
3: Paljeri --> Burgulll
Panaka ——> Panaka
4: Panaka ~> Panaka

1 and2: First route
3: Second rmlle



Map 9: Routes of diffusion and replacement of Paljeri and Panaka.”

23 Lee Sackett pointed out to me that the Kuwarra Koara system was reported by others as follows:

Tindale l939:9l 6 Liberman 1977
Taroro = Burukulu Tjarurru = Yiparka
Karimara = Punmgu Karimara = Purungu

According to these data, the transfomtation only implies an equation of Yiparrka and Burgulu over time.

19 There is another astonishing feature reported for the Yulbaridya Yulparitja. Indeed, Tindale 1953:20 reports
the following system:

Panaka/Purukula = Karirnarra/Purungu

Karimba = Paritj arri

Provided Paritjnrri is a cognate of Paldjari or Paljeri, this system seems to reflect an amalgamation of the Yulparitja
system as reported by McKelson and the systems reported for the Broorne area and environments, among groups such
as the Djaberdjabera, Jawuru, Ngombal. Njul Njul and the Karadjeri, where Garimba is reponed to espouse Paldjeri.
Tindale is reporting an older system, so it seems that the synonyms Pamika/Purukula and Karimarra/Pumngu were
divided each into father-children sections, replacing the former Karimba and Paritjnrri couple.

3° Some additional explanations are necessary. In step 1 Paljeri --> Milangka : Tjarurru, Paljeri has already been