2009/12/21 Maxime Papillon <[log in to unmask]>

>
> >
> > If you answered any of those questions with "no", then the very idea of a
> > causation falls apart, at least as a synchronic explanation. And even as
> a
> > diachronic explanation it looks suspect to me.
> >
>
>
>
> I don't think I ever thought of it as a synchronic explanation, as it would
> obviously make no sense.


Well, that's how you described it.


> I also agree that it is suspect diachronically. I merely observed a
> correlation which aroused my interrested curiosity. No need to attack it as
> a threatening heresy.
>
>
Where am I attacking it? I'm just saying it is not sensical. In what way is
that attacking it "as a threatening heresy"?


>
> You missed the part when I said in free variation (e.g. "ces, des, les,
> mes, tes" or "événement/évènement) _or_ strictly one or the other. Notice
> that I did not say "I could indifferently pronounce [elEv], [elev], [Elev],
> or [ElEv]", but only that it would not create a new word (change the
> meaning).
>
>
I think we have a different definition of "creating a new word". To me,
pronouncing [ɛlev], [ɛlɛv] or [elev] would *definitely* create a new word.
It might not be an existing word, but it's still a different word from
[elɛv]. In other words, those pronunciations definitely *change the meaning*
of the word: they transform a word with a meaning into words with none.

And note that in my idiolect, all those various pronunciations are all valid
words according to phonotactics. That only 1 variant can be found in
dictionaries is immaterial here. All four variants *could* be valid French
words contrasting with the other variants.


>
>
>
> They can only contrast in one very limited morphophonological context,
> while being freely variable or in complementary distribution everywhere
> else. It seems to me a good candidate for being a border case if such a
> thing exists.
>
>
You are not listening to me. I am saying that they are not "freely variable
or in complementary distribution everywhere else" in my dialect. That's my
point: they might not meaningfully contrast in many places, but they are
still not in free variation in my dialect. Without the free variation, I
just can't consider them allophones of a single phoneme, nor can I consider
them a border case.

>
>
> I don't think they have _ever_ been described as separate phonemes... what
> makes you think such a thing?
>
>
I was taught them that way, and they are definitely phonemic in my dialect.
Once again, there are no meaningful minimal pairs, but I can produce
meaningless words that contrast them without a problem (and could before
even learning other languages). If they weren't phonemic at least in my
dialect, I *shouldn't* be able to do such a thing!


>
> Indeed, the merger affected these two even in regions of France that have
> not merged [e] and [E] completely. The two are still contrasted on this side
> of the Atlantic. I'm sure I said something along the lines of "obviously all
> this only applies to the dialects that still do these contrasts". I am
> perfectly aware of the high variability among French accents.
>
>
I know. My point here was not about the merger, but about the fact that
after the merger it was the [ɛ] version that remained in my dialect, rather
than the [e] version.


> > >
> > All in all, I still end up with the same problem: I hardly make any of
> the
> > distinctions you list, yet in my idiolect /e/ and /ɛ/ are definitely
> > separate phonemes, and not even border cases (and that's the case of
> > everyone who comes from Haute-Normandie like me, as far as I can tell).
> > There's not even a single sign that they may collapse into one phoneme,
> like
> > /a/ and /ɑ/ did in my idiolect 15 years ago. That whole business with
> > minimal pairs just doesn't seem to work that well here.
>
>
> I don't know what makes you so sure, given that 1) they did merge in many
> accents (so it is certainly possible), and 2) the number of contrasts is
> perhaps too low to be worth contrasting two phonemes of such articulatory
> and acoustic proximity (i.e., large cost, little benefice).
>
>
I'm sure of it because that's how I speak, and that's how my family speaks,
and everyone I know who comes from the same region as I come from. Heck, a
colleague of mine comes from Algeria (with French as her native language),
and her pronunciation is closer to mine than to what you are describing!

So yes, they did merge in many accents, but that's irrelevant. Spanish
doesn't have [β] as a separate phoneme from [b], but that doesn't prevent
other languages from distinguishing those two. And your second point is the
one I feel is being overused here. You say that [e] and [ɛ] *ought not to*
be separate phonemes because they don't create enough meaningful contrasts.
I'm saying that as logical as it may seem, it just *isn't that way* in my
dialect. I just can't treat /e/ and /ɛ/ as allophones of the same phoneme
when I can natively create as many minimal pairs as I want (whether those
minimal pairs actually have a meaning is irrelevant. What's relevant is that
they would be recognised as *different words*).

I get the impression that you can't accept what I'm saying because it
doesn't jive well with your theory. But that's not my problem. I can't
change the way I speak my native language just because it *ought to* be
different. And I can't change the way I feel about which sounds are phonemic
or not in my native language. After all, I'm the one who speaks it!
-- 
Christophe Grandsire-Koevoets.

http://christophoronomicon.blogspot.com/
http://www.christophoronomicon.nl/