Re: Exceptional rules

--- In cybalist@yahoogroups.com, "fournet.arnaud" <fournet.arnaud@...>
wrote:

> From: Richard Wordingham

> The key is the Arabic word : bayd?a "egg"
> Root : b_y_d?
> d? is an over-featured phoneme : voiced + emphatic dental stop.
>
> The rough equivalent of b_y_d? should be *w_y_H2.
> but Gotic shows that this -d?- must have survived in early PIE.
> -d?- sometimes pops up as -dh- as is the case in *addja.

I don't see how it does. North Germanic and Gothic had the hardening
sound changes jj>ddj (Gothic), jj>ggj (North Germanic), ww > ggw
(Gothic), ww > ggv (North Germanic).

> I don't believe in the playground of weak rules and lazy explanations.
> Work harder. You'll find what happened.

If only! English _blood_, _good_ and _mood_ suggest otherwise.

There's also the case of Rustic Latin ae > e:, which survives in a few
words. The mechanism here is straightforward enough. Phoneme mergers
tend to initially proceed word by word until finally children stop
acquiring the distinction. If the merger should be halted for some
reason, you end up with an irregular rule, of the type commonly
attributed to 'dialect mixing'.

However, it is right to see weak rules as admissions of defeat.

Richard.