Language Families by Convergence (was: Genetic Studies and Aryan Mi

From: Richard Wordingham
Message: 46713
Date: 2006-12-22

--- In cybalist@yahoogroups.com, "mkelkar2003" <swatimkelkar@...> wrote:

> --- In cybalist@yahoogroups.com, "ehlsmith" <ehlsmith@> wrote:

> >Can you provide any attested case of a family
> > resulting from convergence?

> How about the "Hellenic" family. Please see below.
>
> http://ist-socrates.berkeley.edu/~garrett/BLS1999.pdf
>
> http://ist-socrates.berkeley.edu/%7Egarrett/IEConvergence.pdf

> "On the one hand because Mycenaen Greek
> shows innovations that are found only in some Greek dialects, it
> cannot be viewed as proto Greek; it is just an early dialect. On the
> other hand many innovations are found in every Greek dialect EXCEPT
> Mycenaean (Garett 1999, p. 3, emphasis in the original).
>
> These facts make the construction of a "proto-Greek" language
> logically impossible. A chance discovery of the Linear B script has
> lead to this realization. But what about the cases where such
> written evidence does not exist and is never likely to be found; for
> example "Indo-Iranian" According to Garrett what is known to be true
> of Greek has also happened in other cases.
>
> "If we apply what we learn from cases where there is evidence to the
> cases where there is none, it follows that the Indo-European family
> tree with a dozen independent, highly distinctive branches is nothing
> more than a historical mirage (Garrett 1999, p.9)."

What is being talking about here is the interaction of kindred
dialects. The South Slavic 'family' is another such example - if I
recall correctly, there is an ancient East-West split that has been
overlaid by more recent common developments.

The parallel for Indo-Iranian would not be that the Indo-Iranian
languages are genetically unrelated, but that they are related
languages or dialects that have undergone common developments. This
is not a new concept - remember the discussion of the third Slavic
palatalisation, which seems not to have reached all dialects. Compare
Romance palatalisation, which did not affect all Sardinian dialects.

This is not the same as a family being formed from unrelated languages.

I am not sure how we could find such an example. If such a family
were identified and it was then found that the members were actually
unrelated (relatively speaking), the family would be dismissed as a
mistaken concept or, at least, recategorised as a sprachbund. There
are cases where languages have been misascribed - Armenian has been
considered Iranian, the Austronesian language(s) Cham has been
classified as Mon-Khmer, and the Tai languages were long considered to
be related to Chinese, and I don't mean at the deep level where
Austronesian *may* be related to Sino-Tibetan.

However, Armenian was considered an *aberrant* Persian dialect, and
the other mistaken identifications have been uncovered.

So, perhaps the question is, 'What sprachbunds of unrelated languages
could usefully be treated in all respects as genetic families?'.

Richard.