On Fri, 17 Dec 2004 00:13:29 +0100 (CET),
mkapovic@...
wrote:
[me:]
>> - Stang's law. The accent is retracted from weak (usually
>> final, a.p. c) yers, and from medial long circumflex
>> (non-acute) vowels (e.g. iterative -i~-, lengthened thematic
>> vowel -é~- [what caused this?]).
>
>Nothing caused it. There was no lengthening of thematic -e-. It's just a
>myth. The -e- is still short in Croatian a. p. c stems, in two archaic a.
>p. b verbs (mogu and hoc'u) and it's short in Czech as well. Slovak -e- is
>short in a. p. a and b.
If I understand Dybo et al.'s brief treatment in OSA p. 18
correctly, there are traces of length (in SCr, Svk, OCze,
OPol) in all three a.p.'s, but not everywhere at once. In
SCr. there is only length where the acute (a.p. a) or
neoacute (a.p. b) has been maintained, but the -e- is short
where the (neo-)acute has become a neocircumflex.
In any case, even if mogu and hoc'u have short -e-, the
retraction of the accent (mo'z^e, ho'c'e) is not
understandable unless the -e- was once long, and triggered
Stang's law.
Willem Vermeer (in Lehfeldt 2001) refers to Gustavsson '69
for a description of the phenomena, and to Kortlandt '75
(pp. 29-32) for an explanation. I haven't been able to
consult Gustavsson yet. The passage in Kortlandt refers to
van Wijk's law (gybnes^I > gyne:s^I), which would mean that
the lengthening would only be regular in a subset of the
né-verbs (perhaps also the jé-verbs?). I also recall a
suggestion, but I forget where, that the length was
originally confined to the 3pl. (-oN-, indeed long in e.g.
Polish -a,) and spread analogically from there.
=======================
Miguel Carrasquer Vidal
mcv@...