Mate:
> It is just an example, there are not many clic languages in the world
> and if we assumed that all of them died out before we got to them,
> nobody would believe that such sounds could be phonemic. You can count
> on that. That *should* tell us something...
Not really. It's not a parallel example and all it shows is that since
clicks sounds are _rare_, they are not found often in protolanguage
reconstructions unless absolutely necessary. Occam's Razor strikes again.
>> I simply said that a monovocalic vowel system is completely
>> unattested.
>
> Not really. As has been metioned, there are some systems which have
> been analyzed so.
Mate, read Brian's post a while back (I think it was Brian's). He
clarified best the situation in this debate. Jens is analysing Sanskrit
or IE on the most _abstract_ level. It's different from saying that
a language truly has a monovocalic system in a phonetic or phonemic
sense, which is what I'm objecting to above.
Now, in _Jens'_ sense of the word, yes, we could analyse any language
really hard to yield whatever pattern we want, including so-called
'monovocalism', but this isn't real phonetic or phonemic monovocalism.
In the end it is a highly abstractified monovocalism. Basically, an
observed pattern in the vowel system that appears to mimick monovocalism,
just as tv mimicks real life.
If we're talking about true monovocalism, phonemic or phonetic, it
doesn't exist. However when you analyse Sanskrit such that, say, /e/
really 'represents' a _surface_ /ai/, you're talking about something
that is neither phonetic nor phonemic. It is completely abstract. In
a very very abstract sense, we might claim that monovocalism exists,
and I can't argue much against these crazy analyses.
The whole point, Mate, which you're missing is that while I can't
object to _ABSTRACT monovocalism_ as opposed to real monovocalism
(the latter is indeed completely unattested), I don't see what the
purpose is of contorting the facts of a language to suit one's whims.
What does it serve to warp IE or even Sanskrit into a monovocalic
system? What exactly? It's a fun game, but I don't get the point.
> But there are a number of reasons in PIE itself which point to
> monovocalism or something like that,
... Or something like that. As I've been trying to explain, it's not
a true monovocalism. IE had *a, *e, *i, *o and *u with long counterparts
making for a vowel system of at least 10 distinct vocalic phonemes,
eleven if you include any hidden schwas in words like *dHgHo:m or
*pxte:r, for example. The point that *i can be analysed as vocalic *y
doesn't change the fact that in words like *kWis, the segment is most
definitely a _vowel_ *i, not consonant *y. Likewise in Sanskrit, "e"
is just that, despite the possibility that we can pretend it's "ai"
instead for the purposes of the surface analysis to eke out a monovocalic
language. We know that Sanskrit is not _actually_ a monovocalic language
nor does the abstract analysis that Jens is describing show that the
ancestral stage of the language had true monovocalism (clearly not if IE
had 10 or more vowels).
I specifically object to _phonemic_ or _phonetic_ monovocalism. Please
understand. Abstract monovocalism is alright although not completely
without its problems, but I still find it completely pointless until
somebody explains what the point really is. So far, everybody is
quick to jump on my case to prove monovocalism to spite me, and the
wrong sense of the word 'monovocalism' at that.
= gLeN