that this is exactly what happens to an innovative idea or concept that
begins as a provocative twist on existing theory and ends up as a form of
outworn and over-used dogma in itself, simply because too many followers
and epigones have imported it holus-bolus into their accounts for want of
seriously rethinking matters from their ownfield data in the way that
Fredrik Barth did in every new study that he undertook.
The ethnography of the Papua New Guinea (PNG) Highlands is
replete with examples of trying out versions of theory on areas not pre-
viously investigated in depth by professional anthropologists (although
often highly informative and scholarly accounts were available, but in
German or French rather than English). We will take here for considera-
tion two topics that we have examined in previous publications, to illus-
trate our present theme of breaking the frames.
First, we will look at the constitution of groups. By any account, groups
and networks are salient and important aspects of the lives of people in the
PNG Highlands, falling within the general purview of kinship, marriage,
and associative partnerships built on or modeled on kinship and affinity.
Of this, there is no doubt, and it is probably as true today as it was in the
1950s and 1960s when Anglophone research workers began to undertake
fieldwork in every available corner of the region. The major analytical
question at this time, however, was: What were the principles on which
these groups were established? Was there a rule of descent that primarily
determined group membership? Separately, if there was a rule of descent,
did this have to be unilineal descent or could it be cognatic descent of a
type found elsewhere in the Pacific? Anthropologists had found that
unilineal descent was important in some African cases, or was said to be
so, notably among the Tallensi studied by Meyer Fortes in Ghana (Fortes
1945 ) and the Nuer studied by Evans-Pritchard in the Sudan (Evans-
Pritchard 1940 ).‘Descent’, whether unilineal or cognatic in character, is
a typological concept, not a concept of process. It suggests that rules and
structures are in play, and this in turn suggests that these might not always
be followed, if circumstances so determine. The concept of a rule is also
variable in its import. How rigid or important is the rule, how central to
social values? For the Tallensi, in Fortes’s account, unilineal descent was of
prime importance at all social levels and formed the basis of authority,
leadership, and ancestral worship. Tallensi were heavily populated agricul-
turalists, for whom land near to their compounds was especially important.
For the Nuer, a different pattern emerges. The Nuer were pastoralists,
depending on their cattle and moving around with them. They belonged
38 BREAKING THE FRAMES