R A Brown skrev:
> Joe wrote:
>> Roger Mills wrote:
> [snip]
>>> Ancient Greek shows *s > h/0 (at least in some environments), and 
>>> IIRC some of the Persian langs. do too. Of course, IE had no other 
>>> fricatives to lose... French lost *s in certain environments; Spanish 
>>> may be in the process. But it's noteworthy that this is not a common 
>>> change in IE languages.
>> Are you sure?  Sanskrit does it, at least word finally.
> It also occurred, under similar conditions to ancient Greek, in Welsh & 
> the other Brittonic langs.

It also happened *everywhere* in the Iranian/Persian branch.
However Iranian belongs to the Satem(1) branch, where *k_j > s
and *g_j/*gh_j > z, so all Ancient Iranian languages still had
an /s/ phoneme.

(1) The word _sat@m_ itself is an example of this, being
the reflex of *k_jm=tom "100" in Avestan, the oldest attested
Iranian language, and one of the oldest attested IE languages
overall, in spite of probably not being written down until
between the fourth and sixth centuries CE.  It was orally
transmitted by the Zoroastrian priests till then.

/BP 8^)>
Benct Philip Jonsson -- melroch at melroch dot se

    "Maybe" is a strange word.  When mum or dad says it
    it means "yes", but when my big brothers say it it
    means "no"!

                            (Philip Jonsson jr, age 7)