Languages besides Toda with /ɽ͡r/?

[2+3 clusivity]

This morning I was reading Patterns of Sounds by Ian Maddieson and noticed something funny in the 1984 edition--not sure if there is a more recent one.

On p. 87 he writes, The 7-liquid languages are 2 Australian languages (Kariera-Ngarluma . . . [and] . . . Arabana-Wanganura . . .) which contrast laterals at 4 different places of articulation, trills at 2 places and also have a retroflex approximant."

On pp. 330 and 332 he gives phoneme inventories for both. Kariera-Ngarluma has a dental/alveolar voiced trill contrasted with a voiced retroflex trill while Arabana-Wanganura has an alveolar voiced trill contrasted with a voiced retroflex trill.

So, as someone who likes funky trills, I was pretty floored to say the least.


Both of these languages do not appear to have much research done on them. Some quickly googling brought up the following for me.

Kariera-Ngarluma: http://en.wikipedia.org/wiki/Ngarluma_language (saying the two are mutually intelligible but considered seperate langauges), http://coombs.anu.edu.au/WWWVLPages/Abo ... /4_4_4.htm

Arabana-Wanganura: http://en.wikipedia.org/wiki/Arabana_language, http://en.wikipedia.org/wiki/Wangganguru_dialect (Both of which show phoneme inventories WITHOUT 2 trills, and says Wangganguru is an extinct dialect). http://books.google.com/books?id=Mp4PqB ... &q&f=false (On p. 135 showing the source for Wikipedia's phoneme inventory)


So, no real luck there. I go back to the UCLA Phonological Segment Inventory Database (UPSID) that Maddieson used. Maddieson p. 5; http://web.phonetik.uni-frankfurt.de/upsid_find.html. It contains neither Kariera-Ngarluma nor Arabana-Wanganura. I guess the database available online is *updated*? http://web.phonetik.uni-frankfurt.de/up ... uages.html. There a search for a pure retroflex trill only turns up "Kurukh"--a Dravidian Language. http://en.wikipedia.org/wiki/Kurukh_language.

If you go to the Dravidian Languages by Bhadriraju Krishnamurti in the Cambridge Series on Linguistics, Kurukh is, however, not shown with a retroflex trill! pp. 58, 75.
http://books.google.com/books?id=54fV7L ... &q&f=false.

Anyone else have better luck with google, &c.? Am I chasing ghosts?