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

22

If sections are institutions that are of importance for their external social functions, then it
can be expected that they would lend themselves well to diffusion from one group to another.
If sections are a linguafranca for kinship, then they must have spread among neighbouring
parties and must be known and understood by groups and families that, every now and then
at least, interact in one way or another. Their character as a “ready-reference index” Fry
l933:267 is the engine for its own difliisional capacity in present as well as in former times.
This is a point noted very early by Lang l9l6:l67, who explains that “when names were
given to the anonymous classes’ in any locality they would be likely to spread, owing to
the convenience in identifying the classes of strangers”. Elkin summa.rises this point when
discussing kinship systems in the Pilbara and the Kimberleys:

The meeting of tribes for ceremonial purposes, and nowadays, too, the mixing
of members of difierent tribes in white employ, facilitates and encourages the
spread of such systems of summarizing kinship. They are naturally of Very great
value at intertribal gatherings, enabling camping, social activities and marriages
to be readily ARRANGED, whereas the labour of comparing and adjusting the actual
relationships through kinship terms alone in different languages would be a very
difiicult process indeed Elkin 19322325.

I.b Section systems and history

The observation of the geographical distribution of section and subsection systems and
nomenclatures throughout the continent and the hypothesis that such systems have diffused
over large areas are not recent. Service’s writings, again, are a particularly revealing example,
especially when he argues that the descent theory type of explanation of social category
systems is inaccurate:

Some Australian tribes have no named classes, others two, four or eight. They
all have patrilineal descent and are remarkably similar in most other aspects of social
organization. Why do they not all use class nornenclatures? Why do the southern
Arunta have four named classes and the northern eight, when they are otherwise
identical? How can the diflusion of the system be explained, if its presence is
caused by some strong local social function or need? Service 19602422

Service makes it clear that, if sections were a device responding principally to local social
needs, then they would not have to difiuse over such vast areas. While the causal relationship
he describes between a strong cultural feature’s local function and its diffusional capacity is
not necessarily demonstrated, one would have to agree with Service that a strong external
fimction would be conducive to adoption by neighbouring groups. Indeed, possible internal

9 Lang l9l6:l6S defines these anonymous classes as “the divisions which are in fact made in a tribe by the

operation of the terms of relationship”; one would call these kin categories today.