Re[4]: [tied] Re: [pieml] Labiovelars versus Palatals + Labiovelar A

At 5:05:57 AM on Saturday, November 1, 2008, Arnaud Fournet
wrote:

[...]

> Right, there is a sound like [w] in French,
> now the issue is : what is its status ?
> 1. you say it's a consonant /w/
> 2. I say, wrong, this is just a vocalic component in
> diphthongues.

> There is a rule that French native words should start with
> only two consonants, and most clusters are C-l- or C-r-

> If you choose option 1 : /w/ exists,
> then this item has the oddest distribution possible :

> 1. it appears as a third consonant in groin [gr-w-ä~]

> 2. It never appears word-finally and in most regular
> contexts where consonants are expected

> 3. it never appears in initial clusters C-l- or C-r (when
> most other consonants do)

Your hyphenation is confusing; do you mean that [wl-] and
[wr-] don't occur? That's also true of modern English,
which certainly has /w/.

> Next, this item can always be vocalized as [u] in slow
> speech.

Irrelevant: that's a characteristic of [w].

/w/ is certainly marginal in French, but there are a few
minimal pairs, e.g., <loua> [lua] vs. <loi> [lwa].

> There is no phoneme /w/.
> There is just a sound [w] that is a form of the vowel /u/
> in pre-vocalic context.

Note that this is precisely the position to which you were
objecting in the case of PIE.

Brian