--- In
cybalist@yahoogroups.com, "mkelkar2003" <swatimkelkar@...> wrote:
> "Thus a language family can be the product of divergence, convergence
> or a combination of the two (with emphasis on either). There are
> virtually no criteria that would indicate unambiguously to which of
> the two modes of development a family owes its existence. When we are
> dealing with languages so closely related that almost all the elements
> of vocabulary and morphology of each are present in all or most of the
> other members (allowing for sound correspondences), it is more natural
> to assume convergence than divergence."
There is a glottochronological test I can think of. If one compares
the correspondences in the Swadesh list, for the case of divergence
one should get higher match rates amongst the more conservative
meanings. One problem with this approach is that convergence followed
by long divergence will be hard to distinguish non-parametrically from
mere divergence. This test should work when 'dealing with languages
so closely related that almost all the elements of vocabulary and
morphology of each are present in all or most of the other members',
but some of that description is largely hyperbole.
I'm having difficulty of thinking of cases where convergence should be
picked up. I expect an analysis of Germanic languages would show the
divergence of English and North Germanic rather than the slight
convergence due to the massive incorporation of North Germanic vocabulary.
Is there a respectable 'family' that is due to convergence? The best
I can think of is the 'Indian family', but I suspect that that will be
an unfair example. Austric is mostly too weak on vocabulary
correspondences to test, though one might just be able to do something
with the (Malayo-Polynesian)-(Tai-Kadai) or
Austroasiatic-(Malayo-Polynesian) correspondences. (The
Austroasiatic-PAN correspondences are too few for statistical
analysis, and there are indications that Tai-Kadai is related in some
fashion to an Austronesian language related to Malayo-Polynesian,
specifically the 'East Coast Linkage'.) Nostratic is likewise too
weak on vocabulary correspondences to hope to quantify convergence
statistically.
> "The position I (Lincoln) urge is the following. First, we accept as
> established the existence of a language family that included
> Tocharian, [...] and Celtic. Second, we
> acknowledge that the relations among these languages can be described
> in several fashions. Of the available hypotheses, the Stammbaum model
> is the most popular, but by no means the only one. It ought not to be
> accepted as long as others exists, [...]
> Still less the
> existence of a protopeople, protomyths, protoideology, or
> protohomeland [...]"
Not altogether unlike Romance, though that does have a proto-language.
Bear in mind that a non-Roman religion was adopted after the
establishment of the Roman Empire, and that local pagan cults survived
in conquered territories.
Richard.