201centuries in two distinct guises [ 216 ]. Using contemporary European society and
cultural ‘development’ as a benchmark for the human species, proponents of the
former phase assigned other cultures to three developmental stages—savagery,
barbarism and civilised—and theorised that this was the trajectory taken by all
societies. In the second, more modern instance of Cultural Evolution, the focus
shifted from the crude generalisations of the old system to what was thought to be a
set of empirically sound, scalable traits. Here, too, a developmental sequence was
proposed: band, tribe, chiefdom, state. In his critique of both schools of thought,
Dunnell [201] points out how the former struggled to come to terms with the con-
tinued existence of traits categorised as primitive in many societies around the
world, and with the lack of any apparent ‘development’ (comparable to Europe) in
the archeological record of Australia and North America. Supporters of the second
phase, who attempted to order societies using scalable empirical traits, were guilty
of choosing traits that were, “ by and large, traditional ad hoc descriptive catego-
ries ” (p. 178). Dunnell concludes that for such academics, “ increasing complex-
ity or progress is not discovered by scaling societies, but rather it is built in a
priori by trait selection ” (p. 178).
Berlin admits that some anthropologists might object to his use of the word
‘ evolution ’ (p. 272), but the above discussion makes it clear to see that at least some
prominent evolutionary biologists share the anthropologists’ misgivings. The exis-
tence of subgeneric folk taxa in the classifi catory systems of some languages might
indicate a higher level of ‘morphological hierarchical complexity ’—just one of
McShea’s four metrics—over languages that lack this level sensu Berlin [ 9 ].
However, to conclude that the former type of system is more complex than the latter
is, in my opinion, premature—the use of a metric other than the presence or absence
of binominal subgeneric taxa might reveal a very different picture. To hypothesise
that the taking up of agriculture by previously hunter-gatherer populations is usually
accompanied by a proliferation of subgeneric taxa in their language (i.e. a historical
correlation) seems unproblematic, especially if a mere ‘passive trend’ is being pos-
ited. It is quite another matter, however, to posit a causal relationship, and claim that
agriculturalists take a “more careful” look at nature, and that this high level of scru-
tiny allows them to detect hitherto unobserved patterns within the “undifferentiated
generic gestalten” that they had previously taken for granted. The comprehensive
list of Anindilyakwa plant and animal names complied by Waddy [ 26 ] is meant to
be in keeping with Berlin’s views on the ethnoclassifi cation systems of hunter-
gatherer communities—it completely lacks any subgeneric taxa, and so must be of
a lower complexity than the classifi catory systems of agricultural communities.
However, as Baker [ 13 ] has cogently argued, Australian languages are, in general,
able to function just as well as the languages of agriculturalists in the absence of
binomial subgeneric taxa, with their speakers making fi ne-grained distinctions
between hundreds of closely-related biological species.
The Solega cannot be satisfactorily placed within the neat agriculturalist / hunter-
gatherer dichotomy, as they not only cultivate food crops such as fi nger millet and
corn, but also forage for an extensive range or forest products for daily use. As
mentioned earlier, they do not keep honeybees, but instead collect wild honey in an
7.1 On the Manipulation of Resources