On Sun, 11 Mar 2001 16:23:08 -0000, "Sergejus Tarasovas"
<
S.Tarasovas@...> wrote:
>--- In cybalist@..., Miguel Carrasquer Vidal <mcv@...> wrote:
>> <`> *d, *dz^, *dz, *dL; also *3 [?]
>
>What phonetic implementation of <`> would you suggest for Ancient
>Egyptian? Traditionally postulated `ayin-like sound could hardly
>develop from the alveolar and palatal sounds you enumerate.
Loprieno says: "in the apical and interdental series, voiced *d, *z
and *ð [i.e. what I call *d, *dz and *dz^] develop into the pharyngeal
phoneme /3/, probably going through an intermediate stage with a
pharyngealized lateral: *d, *z, *ð (> *l~ >) > /3/". A footnote
compares the well-known Aramaic development of PSem. *d. (i.e. *tL.,
an emphatic lateral affricate) to <q>, later <`> (`ayn). A further
footnote refers to the existence of doublets such as Demotic <ms^d>
"to wander", vs. <ms^`j>, with PAA *d preserved.
There is no question that Middle Eg. <`> was indeed the same sound as
Semitic `ayn (the figures from Hoch are Sem /`/ -> Eg. <`> 100%, Eg.
<`> -> Sem. /`/ 92% (with two isolated cases of /h./ and /G/).
I'm not sure I buy Loprieno's explanation, but I have no better one to
offer. From the systemic point of view, it seems clear that in
Ancient Egyptian there was a tendency to eliminate the opposition
voiced ~ unvoiced in favour of aspirated ~ ejective (the sounds
traditionally transcribed as <d>, <d_> and <g> were probably /t./,
/c^./ and /k./). In the coronal series (dental/alveolar stops,
dental/alveolar affricate, palato-alveolar affricates, lateral
affricates), this resulted in <t> (or <s>) = /th/, <d> = /t./ and /3/
(`ayn), with complete loss of the obstruent element in teh original
voiced phoneme. It could be said that in a system like the PAA one,
with a three-way distinction between T, T., D (etc.) and with the
phonemes /h/, /?/, /3/, the former could be interpreted as C + /h/
[voiceless aspirate], C + /?/ [voiceless ejective], C + /3/ [voiced],
or, alternatively, the latter as the "zero consonant" in aspirated,
ejective and voiced versions.
=======================
Miguel Carrasquer Vidal
mcv@...