On 2008-02-12 09:36, fournet.arnaud wrote:
> It's quite absurd to write that a(:) is not conditioned by anything.
> obviously a(:) is eH2 and °H2.
Obviously?? It's one thing to observe that **eh2 > *ah2 > a:, and an
entirely different thing (a fallacy, in fact) to claim that _every_ *a:
comes from that source. To give you an illustrative example, most /a:/'s
in non-rhotic British English come from earlier /ar/ before a consonant
or word-boundary (/r/ behaves like *h2 here). But it would be wrong to
claim that all instances of /a:/ result from /r/-vocalisation, or that
there had been no /a:/ in British English before the loss of
syllable-final /r/. Words like <father> or <cast> don't go back to
+farð&r, +karst.
> You are unconvincing.
You are demonstrably wrong, which is worse. I will discuss the
alternation of _non-laryngeal_ *a:/*a in a separate posting.
Piotr