Re: [tied] W & V

From: Piotr Gasiorowski
Message: 8843
Date: 2001-08-29

There are some natural trajectories for the "reinforcement" of [w] in initial positions (the syllable- or word-initial slot is normally occupied by consonants, hence the tendency to make glides less vocalic and more consonantal there). High back vowels and the corresponding semivowels (glides) are almost universally rounded (one well-known exception I recall is Japanese, where both /u/ and /w/ are typically unrounded), and a syllabic [V] would be a doubly unusual vowel (not only unrounded but "labiodentalised" as well, something that vowels normally are not) -- pronounceable, no doubt, but not quite expected to occur in a natural language.
How can a [w] become more consonantal? Here are the most common trajectories:
(1) Unrounding, with the simultaneous weakening of the velar component. The result is a "lax" bilabial approximant, which may be strengthened into the bilabial fricative [B], and further into the bilabial stop [b].
(2) "Labiodentalisation" instead of rounding (a characteristically consonantal approximation mode replaces a characteristically vocalic one). The result is the labiodental approximant [V], which may increase its consonantal strength by becoming the labiodental fricative [v]. Velarisation is typically lost as well in the process.
(3) Devoicing and strengthening into a labial-velar or labialised velar fricative ([W] or [xW]). Further stages such as [f] or [x] are possible.
(4) Strengthening the velar component, often with the gradual loss of labiality. The result may be the velar fricative [G] (or [GW] if rounding _is_ retained). Further development may lead to [g] or [gW].
Some typical IE examples:
(1) *w- > b- in Spanish, Hindi, etc.
(2) *w- > V- in some varieties of Dutch and Danish
    *w- > v- in Russian, German, French, etc.
(3) *w- > f- in Irish
(4) *w- > G- in Parachi
    *w- > gW- in Welsh
    *w- > g- in Armenian
----- Original Message -----
From: liberty@...
Sent: Wednesday, August 29, 2001 8:04 PM
Subject: [tied] W & V

... Is this w/v alternation some sort of universal tendency?  Don't Hawaiian and Swahili also have this?  In any of the languages where [V] is an allophone of /u/ are there any cases of the syllabic allophone also having labio-dental approximation in place of rounding?  I've never heard nor read of such a pronunciation but I wonder why it shouldn't be.