Three-Way Contrast of Secondary Articulations in PIE

[Rich Alderson]

On 10 Mar 2001, Pat Ryan wrote:

[I had written:]

>> Later studies of the language have found that the proper analysis requires
>> that there be two vowels, /a/ and /I-/ ("barred i").  The major proponent of
>> this analysis is, as I recall, John Colarusso; see his grammar of Kabardian
>> for more details.

> But is not /I-/ a high/mid central vowel? opposed to /a/, a low central
> vowel.

High central unrounded.  And the operative word there is "opposed".

> I would be inclined to regard /I-/ as an allophone of /a/ in specified
> phonological environments.

Only if you don't look at Colarusso's evidence for the phonemic status of the
opposition, that is, that both occur in the *same* environments, and neither
can be motivated from the other.

> Without rehearsing my arguments, why do you not tell us a few IE words in
> which you believe /i/ and /u/ are primary?

Any word in which *i or *u can be reconstructed.  That they may alternate in
morphophonological rules with *y and *w is irrelevant to their phonemic status,
though the consonantal forms may have originated in fortitions of originally
vocalic segments.  (For the concept of fortition processes, see P. Donegan's
Ohio State dissertation, still available I believe as a working paper from the
linguistics department there.)

								Rich Alderson