--- In
cybalist@yahoogroups.com, mkapovic@... wrote:
> How does Kortland's theory explain Croat. dial. dat. pl. grado``m <
> *gordom'U. According to Kortlandt, it should be *grado~m (like
go~r),
> right?
Yes it should. Note that the long reflex of the *o is actually
attested in Czech.
The short reflex in Slovak, Slovene and Croatian is obviously due to
analogical carry-over from the (b) paradigm: *stolòmU reflecting
earlier *stòlomU by Dybo's law. It goes without saying that in a
system where you have -òmU (with brevity) in one part of the nouns
and -ómU (with length) in another, generalization of one of those is
a trivial development.
In my view one of the attractive features of Kortlandt's theory is
the fact that it generates different quantities depending on the
accentual type or other features in lots of endings that are actually
attested both long and short, like the o-stem Dpl. Most rival
theories have little to say about such cases. Another attractive
feature is that (perhaps with one or two exceptions) all of the
analogies it operates with vary between the banal and the trivial and
do not presuppose suspension of disbelief.
All of which doesn't mean it's perfect and cannot be improved upon.
Willem