Retroflexion

[George Thompson]

In response to the post of Lars Martin Fosse:
>
>Just as part of the general argument, I would like to supply the information
>that both East Norwegian and West Swedish have produced retroflexion. So
>have some South Italian dialects. All quite independent of Skt.
>

Yes, this sort of retroflexion is frequently cited by Hock as well, both in
the recently cited article and in his valuable book, _Principles of
Historical Linguistics_.  But to my knowledge all of the examples that are
cited show these retroflexes as allophonic variants of either dentals or
alveolars, or even palatals.

It seems to me that reference to such phenomena in other languages would be
more cogent if one could produce examples of *phonemic* distinctions, not
just allophonic ones. In Indic languages one can point to minimal pairs
like

Skt:    pAta ['flight'] vs. pATa ['portion']

        [this is one of Hock's preferred examples]

Hock has not made it clear that such minimal pairs can be found in
non-Indic languages. Can anyone on the list supply such minimal pairs?
There is little dispute about the suggestion that retroflexion is an areal
feature of the SA subcontinent. The question remains, however: how special
a feature is it?

Thanks in advance

George Thompson