Re: [tied] PIE Stop System

On Sat, 13 Sep 2003, P&G wrote:

>
> For /ph/ there are 5:   *deph, *eph, *kapho, *k^iph, *phol
> For /th/ there are 12:  *kantho, *kenth, *kwenth, *math, (x2) *menth (x2)
> *perth, *rek^th, *ske:th, *wieth, *wreth
> For /k^h/ there is 1:  *k^onkho
> For /kh/ there are 4: *khakha, *makhos, *skhel, *(s)meukh
> For /kwh/ there are none.
>
> Is this "ample",

Perhaps not, that's a semantic question.

> and does it "demand" the reconstruction of */ph th k^h
> kh
> kwh/?

Yes, certainly.

> For all of these another explanation is possible.  Either:
> (1) the aspirate is required by just one language, so it can be seen as a
> particular development within that language (e.g. *deph), or

Not if there are no such rules for that particular language. Is /h/ just
an invention of Anatolian?

> (2) we have the sequence CH, where the laryngeal has produced the
> aspiration (e.g. *math), or

That seems to be irrelevant, even if it is certainly correct in some
cases: What I have seen of assessments of typological admissibility has
been based on pure *phonetics* and not *phonemic oppositions*. The
criticism levelled at the traditional IE stop values appears to be grossly
exaggerating the sophistication of its own field. Some self-criticism may
be in order here.

> (3) it is an onomatopeoic word, (e.g. *khakha)
>
> The only reason anyone reconstructs the *Th series is because the
> three-way
> system */t d dh/ is so improbable.  Let's not pretend it is reconstructed
> for etymological reasons.

Nonsense, I would posit aspirated surds even if they *caused* typological
worries.

Jens