h, s, s-hacek in Iranian

[Michael Witzel]

On Mon, 29 Jun 1998, Vidhyanath Rao wrote:

> In Iranian, PIIr *s becomes h, PIE *k' becomes s and the
> ruki-conditioned s becomes s with hacek.

(1) as is well known,

IE * k'mtom >  Latin centum : Skt. s'ata- (zata-)   Avestan sata- /
(satem)  e = schwa
Hence, "Satem languages".

and ruki  is`', us`' etc.

but there are amny other cases.

>  Can these three be stuffed into just two phonemes? In particular, can h
> and s-hacek be considered allophones of a single phoneme in Iranian?

Certainly not. s`' (hacek) has several origins and is carefully
distinguished from both s and h in Avestan.

(2) s`' (hacek )  <  also from *k's/tk' (Avest. kas`'a/ Ved. kakSa);
and from ps > wz`';  etc. etc.
and of course from the ruki cases.

The whole is made more difficult be several sub-dialects of Avestan
How to write Avestan in a 7-bit scheme?!!

(3)
In addition there is also palatal s`' '  (hacek plus accent aigu on top)
which developed from the equivalent of Vedic cy
i.e. the famous Nirukta case of Kamboja (= East Iranian) s'avati  for
Young Avestan s`' ' auua(i)ti. ~ Ved. cyavate

In good MSS this is distinguished from simple s`'.


(4)  s and h are of course also clearly distinguished.

old *s is is only preserved in some cases (asti = ved. asti, etc.)

All of these various development resulted in phoneme status. No longer
just simple phonetic allohones of s/h/s`' sounds

This is not just phonetics or a specialty of priestly  recitation
You cannot substitute these in pairs.


We thus have  by now 4 phonemes: s, h s`', s`' ',


and one could add the dotted  s`' .  <   rt  (as`'.a ~ ved. Rta) Older
pronunciation probbaly [hl] (lateral l) as shown in mpers ahlavan for
Avest. as`',uuan-. Medieval (in India) > s`'.

Plus the xi(i) case from sy, plus ngh < s  in -asa-/AsA- surrounding,
plus xv  < sv etc etc


I sum: many phonemes. Plus  some phonetic variants, mostly due to dialects


 ==========================================================================
Michael Witzel                       witzel at fas.harvard.edu