Re: Re[4]: [tied] Re: *a/*a: ablaut

On Sat, 1 Mar 2008 17:16:29 +0100, "fournet.arnaud"
<fournet.arnaud@...> wrote:

>[mcv]
>Merely a misunderstanding on your part. Synchronically,
>Hittite stressed /e/ is always long [e:], unstressed /e/ is
>always short [e]. Nothing strange about that.
>
>=============
>It's a highly strange feature, on the contrary.
>Why is it that
>Anatolian

Hittite!

>allows *a/a: *i/i: *u/u:
>in stressed vowels
>but no short *e in stressed vowel.
>Phonologically insane.

No, all short vowels are lengthened under the stress in open
syllables. In closed syllables, only *e and *o are
lengthened (to [e:] and [a:]), but not *a, *i and *u.

>Now the fact that Hittite
>has e:s-mi out of H1es-mi
>makes your case hard to plead.

PIE *h1és-mi > PA *ésmi > Hitt é:smi.
What's the problem?

>See Melchert's Anatolian Historical Phonology. The most
>important evidence for PA *e: is of course that it shows
>distinct reflexes in Luwian, Lycian and Lydian as opposed to
>*eh1. Some of the evidence given by Melchert:
>
>(With *e: > i:)
>*ne: 'not' => Luw. ni:, Lyc. ni, Lyd. ni-d
>=====
>Unclear.
>The root might be *n_?

This is PA *ne: < PIE *me:

>A.
>===========
>*He:rut- 'curse' => Luw. hi:rut-
>*ke:s-ah2-ye- 'to comb' => Luw. ki:s^a(i)-
>*e:lH-ah2-ye- 'to wash' => Luw. e:lha(i)-
>=========
>Why e: instead of i: ??

It's a cuneiform Luwian form: e/i-graphemes are largely
indistinct. Read [i:lxa(i()-].

>Because the graphic system
>has only -e-l- but -i-r- ?

The graphic system has only -e/iC-.

>A.
>==============
>*e:ptro- > Luw. i:ppatarri(ya) 'distrain'
>*HVpe:riye- 'to sell' > Lyc. epirije-
>*pe:r 'house' > Lyd. bira
>*se:rmó- 'of the precinct' > Lyd. sirma
>
>=========
>Two-thirds of your examples
>have the structure C-i-r
>with -r ending.
>Contextual realization ??
>Not a phoneme ?

No. Only two of the eight examples have tautosyllabic -e:r-.

>==============
>
>(With *eh1 > a:)
>*h1yeh1- 'to do, make' > Luw. a:-, Lyc. a-
>*seh1- 'to let go' > Luw. sa:-, Lyc. ha-
>*dheh1- 'to put' > Lyc. ta-, Lyd. ta-
>etc.
>
>=======================
>Why is it not from *oH1 ??

Because all these forms had *eh1 in PIE.

=======================
Miguel Carrasquer Vidal
miguelc@...