Re: [tied] *h3 (This odd feeling of deja-vu)

From: Piotr Gasiorowski
Message: 15517
Date: 2002-09-16

----- Original Message -----
From: P&G
Sent: Sunday, September 15, 2002 11:12 AM
Subject: Re: [tied] *h3 (This odd feeling of deja-vu)

> I guess we can always sugggest -oh1- or -oh2- instead of -eh3-!   But that leaves me with two questions:
> (a) What about h2o?   There is an arguable case that the reflex of h2o is /a/ in Greek and  Latin, not /o/.  See Sihler p45.
There might be a case for Latin (though the evidence is anything but strong), but it's precisely in Greek that we find different reflexes of *h2e- (> a-) and *h2o- (> o-), the latter e.g. in <okris> (*h2ok^-), <ou> (*h2oju) and <ogmos> (*h2og^-), not to mention <au-> and <ou-> variants in the 'ear' word. Sihler's best counterexample is perhaps <├ínemos>, but other explanations are available for that word.
> (b) As I said in a previous post, isn't it better to keep a regular -e- grade in some of these verbs, expecially in forms such as the reduplicated presents?  The pattern that is beginning to emerge of -e- in some forms and -o- in others is broken if we have o-grade in all the places that require h3.
The above is largely true, which is why I follow the majority in reconstructing *h3 in such cases, while sympathising with reductionist tendencies -- if only as a possibility for which better justification may be found in the future.