Re: Root Structures

Eric Christopherson wrote:

> > There are a few possible root structures for Proto-Indo-European roots...
> What's more, the theory seems to be that PIE root shad only one real vowel,
> written <e>. This vowel became <o> and 0 in different types of derived
> words. <I> and <u> are thought of as simply vocalic versions of <y> and
<w>.
> I've also heard that some people think Proto-Semitic roots had a vowel in
> addition to their three consonants, e.g. CVCC.
Yep.  I guess that a monovocalic system is pretty objectionable on
typological grounds, but at the very least it's tempting to analyze
long vowels away as reflexes of vowel+H(x), and many a's and o's away
as reflexes of (vowel)+H2 and H3 respectively, and i's and u's away as
vowelized versions of y's and w's (just like the vocalic R and L and M
and N).

But apparently even after all that it's hard to get rid of an older
e/o distinction, though some have tried to do so, invoking an old
stress pattern as the conditioner of e/o.

-----------------------------------------------
Boxcars are pulling an Ed of sorts out of town.
             edheil@postmark.net
-----------------------------------------------