Re: [tied] Re: *-tro-/*-tlo-

On 2007-06-23 01:59, stlatos wrote:

> This is part of the reason I said the +/-syl didn't matter in Latin.
> Both full- and 0-grade forms of words in -tlo- are needed to explain
> forms without h so I don't see why all those with h should be treated
> as always showing full grade.

Perhaps. Olsen's work on the allomorphy of *-tlo- was pioneering, the
evidence is fragmentary and certainly much remains to be done. For the
time being I prefer to treat the more restrictive version as the working
hypothesis for methodological reasons, but I'm open to counterarguments.

>> I don't believe in this one. We also have <ukti-,
>> ukta-, vakti-,
>> vaktar-> etc. The ordinal suffix may have been
>> *-th2o- > IIr. -tHa-,
>> (Ved. caturtHa-, paktHa-*, s.as.t.Ha-), losing the
>> laryngeal elsewhere.
>
> You seem to be saying that a little analogy is impossible, but a lot
> is certain.

No, I'm only saying there is too little evidence to generalise from.
<uktHá-> is fine, and I could adduce e.g. <riktHá-> from *leikW-, but
unmotivated, synchronically unanalysable forms like <nákti-, nákta->
(*nekWt-), which are unlikely to have been affected by analogy, show no
/tH/, which makes me suspect that the aspirate must be explained
otherwise. The contrast between the adjectives <riktá-, uktá-> and the
nouns <riktHá-, uktHá-> is surely somehow significant.

Piotr