Barry Garcia wrote: >I borrowed a book through interlibrary loan (from The University of Texas >at Arlington, BTW) on my grandfather's dialect of Visayan (A Philippine >Language) called "Aklanon" (Akeanon), and it says that it includes a >unique sound among the West Visayan Languages, which is the voiced velar >fricative. > By any chance, was it one of David Zorc's works on the language? As I recall from long-ago reading, it seemed to be some sort of palatalized (maybe velarized too) [j] sound-- but the description wasn't entirely clear to me either, and I've never heard the language. >It tells how to make the sound, but i just cant seem to do it. >What other languages have this sound (better yet, where can I hear this >sound on the web)?> >An interesting bit of info on it is, that in Akeanon, it is written as >"e", as in the native spelling for "Aklanon". It's also interesting to >note that this sound occurs where other Philippine languages would have >"l".> It's apparently the native Aklanon reflex of AN *l. Does Aklanon have a "normal" [l] with back vowels too? Probably borrowings from other Bisayan dials. or Tagalog etc. in that case. The book also says it usually occurs with a or o, and rarely with i. >The book makes sure to point out that l is usually with i, and not usually >the voiced velar fricative. Any reasons for this, and why would l and the >voiced velar fricative be linked?> Not hard to assume that /l/ developed backed or velarized allophones in the env. of back/round vowels-- the old "bright/dark" contrast-- cf. Engl. 'feeling' vs. 'falling'. Then if the tongue tip fails to maintain contact, lo and behold, you have a velar approximant or fricative. Related is the very common shift of final [l] > [w], also, the difficulty lots of American children have with both /l/ and /r/, pronouncing them as [w] in early stages of acquisition. At least one lang. of Indonesia has the shift *l > /j/-- Bare'e, a Toraja lang. of central Celebes (extensive grammar and dictionary by N. Adriani, in Dutch). Even Tagalog has a number of words with /y/ for *l, presumably loans from some neighboring lang. or dialect-- I'm not sure whether the source can be pinpointed. Interesting that *l tends to be one of the more stable sounds in AN languages-- unlike the other "fricatives", velar *R and palato-velar(?)*j (whatever that represents......). These seldom merge, but tend to show very similar developments: *R > r, g, h, l, y, zero (sometimes with fronting of surrounding vowels), z in Malegasy IIRC; *j > r, l, d, more rarely g, y, z; Polynesian s/h. I've always found this very perplexing and would suspect a methodological error somewhere along the way-- but both sounds are quite well attested all the way back. Not hard to see how *l could get mixed up in the same series of changes.