Another initial "vl"

Mark Mandel thnidu at GMAIL.COM
Fri Dec 28 17:11:52 UTC 2007

When in grad school in Berkeley I sang with a men's chorus called
"Slavyanka" (an SF offshoot of the Yale Russian Chorus), singing Russian and
other Slavic and other USSR music. I provided some advice on pronunciation,
but for Georgian I wrote to my aunt, Dee Ann Holisky, who was one of the
handful of Georganists in the US. Georgian has a thousand-year-old tradition
of 3-part men's liturgical music, very beautiful and quite different from
the more familiar Western European forms.

One of our pieces was "Shen Xar" ("x" = voiceless velar fricative, IPA [x]),
a hymn to the Virgin. I still remember most of it (tenor? baritone? part),
especially the last line; in ASCII,

    mze xar ga brtsq'in vebuli

In the fourth word, "ts" is an alveolar affricate and "q'" is a uvular
ejective. Dee wrote (paraphrasing) "You can drop the /r/ if you need to, but
DO NOT insert a vowel in the cluster."

Oh, and musically you can't attach the "b" to the preceding word. The "ga"
is no way proclitic: it's a long melisma, about 3 seconds, and then
"brtsq'in" has the downbeat.  :-)

m a m

On Dec 28, 2007 7:41 AM, Dennis Preston <preston at> wrote:

> As I suspected, many apparent clusters in Georgian (which has a rich
> inventory of clusters and does not need to be exaggerated) are not
> clusters at all. All sonorants (and /v/ is classified as one) ,
> including also  /m/, /n/, /r/, and /l/, can be syllable peaks. In
> fact, in certain environments an allophone of /v/ is a considerably
> reduced /w/-like rounding of a preceding consonant. Vowel epenthesis
> also breaks up many apparent clusters. Most of this rich consonant
> clustering comes about a a result of a rich affixing morphology and
> is not present in root morphemes.
> Go here
> for a thorough treatment.
> dInIs

The American Dialect Society -

More information about the Ads-l mailing list