Re[2]: [tied] Latin -idus as from dH- too => and the accent of Grk.

At 4:49:48 AM on Tuesday, March 18, 2008, alexandru_mg3
wrote:

> --- In cybalist@yahoogroups.com, Miguel Carrasquer Vidal
> <miguelc@...> wrote:

[...]

>> Kiparsky does offer an interesting theory about why the
>> Nsg. form has recessive accent. I wonder how you can have
>> missed it (it's on page 11).

>>> Another demonstration of the generalization that accent
>>> in simple words is assigned on the basis of the
>>> pre-contraction syllable structure comes from the
>>> process of IAMBIC RETRACTION. This process, first
>>> identified in Bartoli 1930, deaccents a final iambic
>>> sequence (^ -') in polysyllabic words, resulting in
>>> recessive accentuation.

[...]

> (te:r in pate':r, being a heavy syllable too isn't it, as
> in dHuga'te:r too...)

<paté:r> isn't polysyllabic. Patrick Bye, 'Accentuation and
Contraction in Attic Greek: An Axiom-based account',
referring to Wheeler's Law and Bartoli's Law:

These retractions have in common that they are restricted
to polysyllabic words, i.e., stress domains of at least
three syllables. The effects do not surface in
disyllables.

Brian