Re: [tied] Mid IE Phonology

From: Glen Gordon
Message: 9996
Date: 2001-10-04

Marc V:
> >Thanks, Glen! Very interesting. This looks much more like a real
> >language than the traditional phonological reconstructions (eg,
> >Beekes).
>
>(I meant a/?vs o/e, and t/t:/d vs t/d/dh etc.)

Oh. Right. Well, it would seem that IE had *a as well, at least next to
laryngeals. As for t/d/dh, I'm not sure about how Piotr feels but, the odd
contrast may have been somewhat short lived (ending up t/th/d/dh in
Sanskrit, th/t/d in Germanic, etc).

The fortis stops *[t:, k:] for Mid IE are a way of acknowledging the
typological oddity of a missing **b (which would derive from **p: and
ultimately from **p? if it had existed) without accepting outright the
existence of ejectives in IE itself. Rather, the ejective stage of IE exists
but is pushed into the remote past where it belongs.

>Thanks again for the examples (very interesting - as always :-)). Sorry I
>was vague. I was thinking of examples comparing the traditional
>phonol.reconstructions (eg, p, t, k, kW) with yours (p, t, tW, k, kW).

Well, I reconstruct *tW for Steppe, IndoTyrrhenian, Old IE and Mid IE but
*not* for Late IE. For some reason, labiodentals don't seem to be treated as
single phonemes in Late IE even though I imagine they must have existed
once (Mid IE *sWekse "six", *t:Waxe "two" > *sweks, *dwo:u).

For instance, I reconstruct Steppe *kut:u for "five" (yes, "five", not
"four" as in IndoTyrrhenian). I realise now that Steppe *i and *u are
treated
differently in IndoTyrrhenian such that *i > *� and *u > *a (I used to think
that both *i and *u became *� but this new rule better explains the origins
of *e/*o ablaut in verb roots).

Hence, we obtain IndoT *kWat:Wa "four" (with residual labialisation and a
bit
of a semantic shift). This becomes Tyrrhenian *xotta and Old IE *kW�tWe-n
(with
the inanimate *n-suffix), later *kWetw�res with *-es plural and predictable
accent shift).

So from this, we can see that labiodentals are the result of Steppe
plain dentals in the environment of *u. However, by late IE instances of *tW
or *sW end up *tw and *sw respectively and only labiovelars are retained
as single phonemes. Does that make any sense? I'm not sure why this is but
I'm
not a real linguist anyway :)

>Are there extant languages with a comparable phonology?

NorthWest Caucasian languages... but I suspect that early IE was heavily
affected by them. Hmm, I can't think of any languages lacking palatal
consonants but having labial consonants off hand, if that's what you are
wanting. Although come to think of it, there is French /ma/ [ma] versus
/moi/ [mwa] without a *[mja] or /pas/ [pa] versus /poid/ [pwa] without any
*[pja] afaik. I'll have to think more on that.

>Perhaps a stupid question, but if there were forms like tW-/kW-, could
>there also have been forms like ty-/ky-?

If I correctly understand, Miguel thinks so. However, I don't think that
palatalisation was a factor in early stages of IE. I like Piotr's
explanation
of satemisation in terms of uvularisation very much (which was designed to
solve typological problems with the predominance of a marked and supposedly
palatal *k^) and it works well with a centralized vowel system of *[&, a]
where *a is the uvularizing factor. Hence *k^ is really plain *k (implying
neighbouring *e in Mid IE), and *k is uvular (implying a nearby *a in Mid
IE).

It makes sense then that we should find *k^ in consonant clusters... This
is because consonant clusters like *k^C- imply earlier *k^eC-. Due to my
new-found rule that only *e (schwa) can occur in unstressed positions, this
means that *k could never be uvularized here because this would require an
unstressed *a (which doesn't exist!!). Hence *k^wo:n < *kewane-se "dog",
*g^lo:us < *k:elaxwe-se "sis' hub" and more obscurely *ok^to:u < *ak^twa: <
*kWetWaxe "eight".

So in all, we can do away with palatalisation. I severely doubt that it
was significant in a language lacking front or high vowels anyways.



Now to some interesting stuff...
--------------------------------
But now this latter word *kWetWaxe "eight" gets me on a new tangent... What
if the development of this numeral can help solve our "bird" problem?

Is it possible that the euphonic prothetic *a- seen in the intermediate
form *ak^twa: was part of a larger phenomenon in Early Late IE? Could it
be that an earlier *xewei "bird" begat a new word *xewei�ne "egg" through
the use of a thematic genitive *-ane (> *-om). From there, after the
reduction of unstressed schwas, we would obtain *xwyan...

Now those are alotta consonants to pronounce in one syllable! Gee, it would
seem easier to pronounce them if we had a prothetic vowel at our disposal...
like *a perhaps... That would produce *axwy�n, later *oxwy�n in Early IE.

Of course, by this point we know where things are going since *oxwy�n
becomes
*�xwyon due to the new acrostatic accent of thematic nouns established
during
this time period, and the genitive *-on becomes *-om (affected by accusative
*-m as well as preceding labial *o).

Tada! So what do we get? Well, now we have *oxwyom (> *o:wyom... very
smooth!)
and we also end up with *TWO*, count 'em, *TWO* examples of a prothetic *a
used to intervene in cases of extreme initial clustering in Late Mid to
Early Late IE (c.5000 BCE). Clevah! Now, I gotta wonder how many other
examples there may be of this *a-... ??

Say, Piotr... Can you list those other examples of these supposed *Ho-
compounds? I'm agettin' suspicious.

-------------------------------------------------
Glen Gordon
Webdeveloper

home: http://glen_gordon.tripod.com
email: glengordon01@...
ph: (604)904.0320
-------------------------------------------------



_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp