On Wed, 05 Jul 2006 14:13:07 +0200, Mate Kapovic
<
mkapovic@... hr>
wrote:
>How do you explain possible Hirt's Law in bo´´lto? Shouldn't
it be the same as in <golv'a?
<I'm glad you
asked.
<*gal@... may have /a/. I would put Hirt's law
before
<Winter's law, and Winter's law still distinguishes
between
</a/ and /o/, so Hirt's law may work differently for *oRH
vs.
<*arH.
<The rule also doesn't seem to apply to *a and *o
from *e
<after *h2 and *h3: Latv. ar^t < *h2ar@3téi and ar^kls
<
<*h2ar@3tlóm, aûst < *h2au@... (I don't have any
actual
<examples for *h3e-). Whether it applies in the case
of
<"unmotivated" /a/ (not in the neighbourhood of *h2, *k,
*g,
<*gh) is unclear: dê"verI (*daiHwé:r) and its Baltic
cognates
<would seem to suggest it does, but Latv. sal^t
(*k^al@-téi)
<suggests it doesn't.