Re: [tied] Re: Earth and Thorn

From: Miguel Carrasquer
Message: 39200
Date: 2005-07-12

On Tue, 12 Jul 2005 06:38:54 -0500, Patrick Ryan
<proto-language@...> wrote:

> So, you are saying that "post-classical" studies hold that the plene writing of e-eš-zi means that we should interpret the spelling to mean that the word was pronounced /é:s-tsi/.
> Does up-to-date theory now hold that all plene spellings indicate vowel length?


>Or is the spelling immaterial?

Plene spellings were used inconsistently by the Hittite
scribes. But in cases where they were used, they indicate
vowel length.

> Does that mean we should accept /e:/ as an allophone of /e/ under special conditions (+ root, +stress-accent)?

Usually, [e:] is an allophone of [e], but not always.

The rules as stated by Melchert, simplified, are:

1) all vowels are long when accented in an open syllable
2) /e/ and /o/ (but not /a/) are also lengthened when
stressed in a closed syllable
3) unstressed long vowels were shortened in Proto-Anatolian
4) new unstressed long vowels (e.g. [originally] i-stem nom.
plural /-e:s/ < *-eyes) were created in Hittite.

> I have never seen *dhe:g^hóm. But that's not what Jens wrote. Jens wrote
> *dhég^ho:m - now *that* I saw many times in the literature. In Hittite *dh
> > t, *g^h > k, *e > e: under stress, unaccented *o: shortens regulary and
> yields /a/ and *-m# > -n. Nothing strange about it.
> ***
> Patrick:
> Is there then some other new rule, not governed by stress-accent, which lengthens the /o/ of *dhég^hom?
> ***

Two rules:

1) Szemerényi-lengthening in the nom. (*d(h)ég^hom-s >
2) "mm/wm-lengthening" in the acc. (*d(h)ég^hom-m >
*d(h)ég^ho:m (cf. accustaives like **dyewm > *dye:m, **gWowm
> *gWo:m).

Miguel Carrasquer Vidal