[tied] Re: Six, -ts- > -ks-

--- In cybalist@yahoogroups.com, "m_iacomi" <m_iacomi@...> wrote:

> --- In cybalist@yahoogroups.com, "tgpedersen" wrote:
>
> > And you have ironclad evidence that it didn't occur before that?
> > Lately I've come across many passages in linguistic literature in
> > which the author claims that the late appearance of some feature

is

> > caused by its having to work its way up from the subjugated

popular

> > deep. What am I to make of that?
>
> Nothing. Final -s in "some" Portuguese is pronounced /s^/ with
> other preceeding vowels than /u/.
>
> >>> 2) The usual "path of disappearance" for /s/ is > /s^/ > /h/ >
> >>> zero.
> >>
> >> I don't think so.
> >
> > I do. Nyah, nyah, nyah (that oughta take care of _that_ argument).
>
> Nope. You should substantiate your claim and that "usual" word.

What can I say? Where are the examples of direct /s/ > /h/?

> >> It didn't. It disappeared first before a voiced consonant (by

way

> >> of s > z > D > 0), then (11th. c) before voiceless consonants (by
> >> way of /h/), finally (13th. c.) in final position,
> >
> > Daintily steppping over the /s^/-puddle.
>
> Which was noted by noone. The spelling <cheva(u)x> did not stand

for

> /c^evaus^/, it is simply that /us/ was rendered by <x>: "...

équivaut

> en effet dans l'orthographe au groupe final -us indiquant le pluriel
> de certains noms: au lieu de chevaux, issu de caballus, on écrivait
> souvent chevax [...]", you might have noticed that also from other
> examples I gave.

Whoever this guy is he explains <chevax> as a corruption of
<chevaux>, the spelling he is used to seeing. That has nothing to do
with historical linguistics.

>This rendering was a simple graphical way to write
> faster. Restoring the "u" became a necessity once the diphthong
> reduced to /o/: /s^evo(s)/ [and that was 200 years later than 13th
> century when the graphy <ch> shifted its' pronounciation from /c^/
> to /s^/, providing thus a good graphical rendering for any /s^/
> which might have appeared], the <au> was there to stand for [o],
> but final x remained just as a graphical mark of the plural with no
> real etymological background.

I have no doubt this is the standard theory, but where is the
evidence?

> >> except in liaison, where it survives
> >> until today (as /z/, of course, not as /z^/).
> >
> > Oh! The final blow. But I don't think /s/ > /s^/ would force /z/
> > > /z^/.
>
> Of course not. But were the <x> pronounced [s^], the regular
> voicing of it would have yielded [z^], not [z]. So your theory has
> to include also an explanation for this.
> The simplest explanation is

that sandhi was generalised to [z]. The next simplest

>that your theory does not hold.

Torsten