On 2008-02-09 00:11, tgpedersen wrote:
> ?? In what sense convergent?
A _similar_ pattern of dissimilation, but not _the same_ historical
process. Grassmann's Law in Greek is not only younger than the devoicing
of aspirated stops in the language, but even younger than the lenition
of *s > h.
>> but independent change in each case, not a homology. Italic,
>> Germanic and Armenian preserve DH...DH roots and the contrast
>> between them and D...DH.
>
>
> I know all that, as I desperately tried to communicate in the posting.
> The idea was: suppose the <aspirated>V<aspirated> ->
> <unaspirated>V<aspirated> rule was active synchronically in all of
> PIE, Greek and Sanskrit. That it doesn't exist in the rest of IE would
> mean that it ceased operating there.
Ana a lost contrast was restored? Grassmann's Law means the falling
together of *D...DH and DH...DH. If a root like *dHegWH- had undergone
dissimilation already in PIE, yielding *degHW-, we would get Skt. dáhati
(correctly), but neither Lat. foveo:, nor even Gk. tépHra: come out as
expected. Greek would actually require a cancellation of Grassmann's
Law, *degWH- becoming *dHegWH- again, then aspirate devoicing
(*tHekWH-), and finally Grassmann's Law once again, yielding *tekWH- >
tepH-. And how would the "no-Grassmann" branches have managed to restore
*dHegWH-, leaving e.g. *g^embH- unaffected? How did they know which *D-
was original and which had resulted from aspirate dissimilation?
Piotr