Re: [tied] Re: IE & linguistic complexity

Piotr wrote:

>Of course, PNG has a completely different and much more variegated topography than Australia, with a lot of rough terrain, highlands, rivers, and tropical forests. There is agriculture and animal husbandry there. Additionally, a substantial part of PNG was colonised by Austronesians pretty recently. All these factors are potential equilibrium-busters and would account for numerous (if small-scale) "phylogenetic explosions" in the history of the region.

You're right, I forgot to mention Austronesian. A different boatload
altogether.

>What Australia and PNG have in common is the genetic diversity of their languages. Languages are more numerous and families tend to be slightly bigger in PNG (Austronesian, of course, is a case apart); there is probably more creolisation as well (as opposed to mere diffusion). Anyway, whatever the local conditions, old and relatively undisturbed linguistic areas tend to be occupied by many tiny families and linguistic isolates, and the construction of large-scale family trees for them proves difficult or impossible.

Well, yes, that's only to be expected. What I'm skeptical about are
Dixon's claims that prolonged "equilibrium" will yield a situation
like the Australian one, where family trees are difficult or
impossible to reconstruct because the languages are, in a way, *too
similar*, instead of them being *too diverse* (as in the Papuan, or
for that matter American, cases). It may well be that Dixon is right
about Australia (I don't know anything about the earlier reasons for
claiming Pama-Nyungan to be a genetic grouping, nor do I know anything
specific about Dixon's claims to the contrary), but in that case
Australia must be a special case.

=======================
Miguel Carrasquer Vidal
mcv@...