Re[4]: [tied] PIE *a -- a preliminary checklist

At 12:17:41 AM on Thursday, February 14, 2008,
fournet.arnaud wrote:

>>> I think the difference is glottal stop was a phoneme in
>>> proto-Germanic

>> Evidence?

> Some sequences H2 + k / t become g / d
> Some don't

Examples, please.

>>> Now do you think bu?er for butter is part of the
>>> language? If yes, then it's phonemic in English too.

>> Is it? Or is it just an allophone of /t/?

> I don't think so different sounds can be allophones.

It's not all that dissimilar. There's no doubt that in many
varieties [?t] (simultaneous articulation) is an allophone
of /t/, and [?] is just a short step away.

> Otherwise /ng/ is an allophone of /h/
> Complementary distribution.
> People who make a regular use of -?- in bu?er
> probably have other words where -?- contrasts
> with -t-.

'Probably' isn't good enough, I'm afraid.

Brian