Re: *gwistis

--- In cybalist@..., "Richard Wordingham" wrote:

> --- In cybalist@..., alexmoeller@... wrote:

>> But I know the Romanian "Deget"< lat. digitus is a new word in
>> the language.
>
> As a matter of curiosity, how do you know this?

He doesn't, he thinks he does. It isn't.

>> The old one , still in use, in fact very in use is "de$t".

/dedZet/ > /deZet/ (easier pronounciation) > /deZt/ (simplifying
an unstressed useless vowel) > /deSt/ (unvoicing [Z] by influence
of final unvoiced [t])
All these four forms are currently used (and understood) by
Romanian native speakers; the first is the literary standard,
while the last is the so-called "old one" which turns out to
be the newest.

>> So far I am informed, the Albanians have too "de$t"

Bad information. Albanian <gisht> is actually /giSt/.

>> Any idea how to explain that? ( ge >de)

No need to.

> Latin digitus seems to be exactly the sort of word that would
> syncopate, digitus > *digtus > *dictus. However, I am not sure
> that it did, for while French doigt - the g' has been silent since
> it was re-introduced - is consistent with vulgar Latin *dictus, I
> would expect Spanish *decho, not dedo as we actually have.

From Latin to Romance languages, intervocalic -g- would usually
disappear, already in CIL we have <trienta> instead of <triginta>,
<vinti> instead of <viginti>, <Austa> instead of <Augusta>, etc.;
Appendix Probi, 12: calcostegis non calcosteis. Palatalized -g-
was more resistent in peripheral Romance languages (as Romanian)
or in loanwords from Latin. Normally this would explain the fact
we have conserved (palatalized) -g- in Romanian <deget> (see also
Sardinian <dighitale> < digitale; but <didu> < digitus); also
<i^nger> (/1ndZer/) < angelus, <ager> (/adZer/) < *agilus (=agilis),
<lege> (/ledZe/) < legem, etc.. French "g" in spelling is due only
to Medieval Latin influence on "l'orthographe du Francais": it's a
matter of common knowledge in France that many spellings with no
phonetical relevance were introduced due to cultivated Latin (see
for instance <temps> with an useless final "s", never pronounced).

> The alternative evolutionary path is to drop the /g/ and merge
> the vowels.

Which is actually the case.

> Now, if there were also a plural *ge$ti in Romanian, the parallel
> of *ge$ti might have encouraged a 'slurred' pronunciation
> of 'degeti' as 'de$ti'. But the process probably needed no
> encouragement.

Any reasoning based on inaccurate input about Romanian words has
little meaning. :-)

Regards,
Marius Iacomi