Atkinson on phoneme inventories in Science

[Ian Maddieson]

I owe an apology to Quentin Atkinson and readers of this list for suggesting
that critical details concerning method and sources were missing from the
article. They weren't; they were available in supplementary on-line materials.
Put it down to reading the article late a night while worrying about whether
I was ready for the next day's class. Apologies to all.

I still think that the most important shortcoming of the Atkinson article is
the equation of the number of contrasts with diversity. As I understand it, the
map of human genetic diversity shows an overall trend for populations
further from Africa to show less genetic diversity than populations within
Africa (with lots of local blips and complications, of course). The more
distant populations don't have fewer genes as individuals; they show
less diversity in genetic make-up across the set of individuals.
This is a somewhat separate point from the fact that details of
the individual genetic markers carried allow a phylogenetic tree to be
constructed that is rooted in Africa. Part of the explanation for this
pattern is due to founder effects (parts also to drift and other factors).
The appropriate test for a parallel founder effect in phonological patterns
therefore would be a measure of the differences in the content
of phoneme inventories not simply of their size.

Ian
-- 


Ian Maddieson
Department of Linguistics
MSC 03-2130
Humanities Bldg. 526
University of New Mexico
Albuquerque, NM  87131-1196