Retroflex sounds

[Miguel Carrasquer Vidal]

Jacob Baltuch <jacob.baltuch at EURONET.BE> wrote:

>The southern Italian and Sardinian retroflexes seem to be
>allophones of /l/ since the posters were saying they only occur as
>geminates. I don't know about the Asturian example. There /ll/
>becomes a non-geminate retroflex so if a real [d] (non lenited,
>etc.) exists in that dialect then maybe, but we'll have to wait
>for confirmation from Miguel.

Confirmed.  Asturian /d/ behaves like Spanish /d/: it's the stop [d]
in initial position, and in the combination <nd>, a light fricative
or continuant [D] elsewhere.

The Asturian retroflex (found only two separate enclaves, the
villages of Sistierna and Aller [Felechosa, Casomera]) has nothing to
do with gemination per se.  It's a development of palatal /L/, which
in Asturian can come from Latin -ll- (as in Spanish) or from Latin
initial l- (as in Catalan).  For instance (Sistierna dialect):
/d.una/ "moon" (Spa. luna, Cat. lluna), gad.inas "chickens" (Spa.
gallinas).  In Aller, the result is not [d.] but a retroflex
affricate [t.s.] or [s.d.] (ts.una, gats.ina etc.).  The development
/L/ > /tS/ found in a wider area of Asturias, as well as in Gascon
and Pyrinean Aragonese is also thought to have gone through a stage
with retroflex [d.] or [t.] by some authorities.

But enough about Romance...

=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam