On Sun, 11 Mar 2001 12:11:22 +0100, "Piotr Gasiorowski"
<
gpiotr@...> wrote:
>In <haras> we very likely have *hara: < *H(V)ro:n plus an analogical -s, which is why I didn't include it among my examples.
There are very few examples of Hittite non-neuter n-stems (Friedrich
only gives MUNUS-za, MUNUS-nas "woman" and the handful of -as/-anas
words like haras "eagle" [under "mixed n-/a-stems"]), so I don't have
a good idea of how to interpret them. Van den Hout (U. of Amsterdam)
doubts between Friedrich's explanation (suppletion between a-stem
nom.sg. haras, and oblique n-stem haran-) and a phonetic explanation
(nom.sg. haran-s > hara(:)s). He mentions an "a-stem" acc.pl.
arkammus "tributes" (arkammas, arkammanas) for expected *arkammanus,
and irregular nom.sg. ishiminas, acc.sg. ishimenan besides ishima:s,
ishimanan "knot". The word memiya(n)- ~ memin- "word" is highly
irregular, ocuurring both as neuter and as common gender noun (nom.
INIM-as (memias), acc. INIM-an (memian), n. memian, obl. memiyan- ~
memin-).
>I have doubts about the derivation of sius (<DINGIR-us>) from *siuns, since the Nom.pl. siwannes would imply Nom.sg. *siwa(s), IMO. I prefer to assume that <sius> had a by-form *siw-an-/siu-n- (with PIE *-on- or *-h1on-), formally corresponding to Germanic weak nouns, and that the attested pattern is suppletive, with the simpler variant surviving in the Nom./Voc. I can't check it at the moment, but I recall that the root ablative <É-ir-za> 'from home' and the root locatives <É-ir/É-ri> 'at home' (*per, *peri) are also attested beside <parnaza> and <parni>.
My sources give dat-loc sg. (as well as nom.pl.) É-ir, but not the
abl. form you mention.
[on Egyptian <r>:]
>Thanks for the references, but it seems that the whole interpretation is really a matter of opinion. The reconstruction of Middle Kingdom pronunciation via PAA smacks of obscurum per obscurius, seeing, in particular, how insecure PAA reconstructions are and what untidy Egyptian reflexes are being derived from PAA *l. Couldn't the <r>/<3> contrast have been between /r/ (tapped): /rr/ (strongly trilled, or /R/, or anything "emphatic")? Supposing MK Egyptian lacked distinctively lateral phonemes, the substitution of <r> for foreign /l/ (as in Japanese or Korean) would have been natural, while Egyptian words with <r> would have been borrowed into other languages with a rhotic.
See the quote from Kammerzell in my reply to another message. The
reconstruction of Middle and New Kingdom pronunciation does not depend
on PAA at all, although it does depend to a certain degree on the
reconstruction of Proto-Semitic. The reconstruction of Old Kingdom
Egyptian is more insecure (in all respects), and PAA etymologies play
a larger role. However, most of it is still based, as far as
possible, on the spelling of foreign names in OK Egyptian, OK Egyptian
names in foreign parts, and most of all, spelling variants within OK
Egyptian itself.
A good reference is James E. Hoch "Semitic words in Egyptian Texts of
the New Kingdom and Third Intermediate Period", 1994. After the
development of /r/ to some kind of laryngeal glide, probably by way of
uvular /R/, what Middle Egyptian lacked was a distinctively rhotic
phoneme. According to the data in Hoch, we have:
Semitic transcribed in Egyptian as:
/d/ <d> 79.8%
<t> 13.7%
<r> 6.6%
/l/ <r> 76.6%
<nr> 11.0%
<n> 9.2%
<3> 3.3%
/n/ <n> 97.2%
<r> 2.8%
/r/ <r> 98.2%
<nr> 1.8%
Egyptian transcribed in Semitic as:
<3> <?> 92.9%
<?l> 7.1%
<n> <n> 85.6%
<l> 14.4%
<nr> <l> 85.7%
<r> 14.3%
<r> <r> 60.7%
<l> 37.8%
<n> 0.8%
<d> 0.8%
=======================
Miguel Carrasquer Vidal
mcv@...