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

has been left unused for three years, it is barely distinguishable from the surrounding
forest, except for the presence of banana trees that are still bearing fruit, long after the last
sweet potato has been harvested. After several decades, however, the forest becomes
similar in appearance to the primary forest (Paigmans 1976:80-1), even if the species
encountered are evidence of past human intervention (Cryptocarya, Euodia, Sterculia,
etc.). Only in places where villages have been regularly established over generations has
it disappeared. In these places, the incessant trampling hardly allows the installation of an
anthropic savannah (with /mperata cylindrica and Miscanthus essentially).

In the end, people spend half or even two-thirds of their time outside the hamlet in
which they have their main residence, even if they are sometimes only half an hour’s walk
away. In addition to this, there is an undeniable penchant for isolation among Ankave
families, which is itself not unrelated to the fear of attacks by witches or cannibalistic
spirits.l* Nevertheless, there are several occasions for gathering: collective consumption
of red pandanus or Pangium edule, ‘marketing’ of pork, death of a relative, initiation
ceremony or closure of a mourning.

In addition to this annual mobility, there is a second mobility that corresponds to the
garden cycle and extends over several years. When a family opens a new clearing, every
two or three years, it often clears a portion of its territory that is opposite the previous
site. It even happens that an Ankave opens a garden only to assert his right to use a portion
of forest: poorly enclosed, the garden is de facto abandoned, and ruined by pigs, as soon
as it is cleared.

Finally, it should be noted that these forms of residence, which could be described as
semi-nomadic, are superimposed by a slower but more geographically extensive
phenomenon: the age-old (even millennial) migration of the Anga from Menyamya to the
lowlands. The Ankave are one of the Anga tribes who were pushed westwards by stronger
enemy groups as the Menyamya Valley, where the Anga claim to have originated, was
populated. At the end of these migrations, the weaker groups dissolved into the lowlands
that form the hinterland of the Gulf of Papua. Unlike some of these extinct groups, of
which only the name and memory remain, the Ankave are alive and well, but the process
by which part of the population descends to Papua and never returns can still be observed.
Year after year, a few people leave the hamlet area because of dissension and move to the
far western area of Ankave territory, where there are usually about 50 people scattered
around New Year Creek. Here they are usually bound to enter a cycle of accusations of
witchcraft and armed revenge which, together with the high incidence of malaria, results
in an annual mortality rate of over 10%.

The originality of this mode of land use is spectacularly apparent in the landscape: a
valley which, among other Anga, such as the Baruya (Godelier 1982) or the Sambia
(Herdt 1981), would be covered by a tangle of garden and savannah, with only the ridges
covered by forest, appears to the Ankave as an almost entirely wooded expanse in which
an untrained eye can barely discern the areas where cultivation has recently ceased.
Similarly, instead of large permanent villages being easily spotted from several



19 Cannibal spirits whose attacks are signalled by symptoms that are in every way similar to those of
malaria. As malaria primarily affects families who move back and forth between the hamlet area and the
lowlands infested with Plasmodium falciparum, the tendency to isolate themselves in the lowest part of the
territory is reinforced.

29