On 2006-03-15 02:51, alexandru_mg3 wrote:
> Same for Alb. dëborë 'snow' : finally we 'are obliged' to explain also
> why the intervocalic b in Albanian dëborë is still there (because
> Albanian & Romanian share the same rule b/intervocalic>zero...And
> there is a single answer: because it wasn't a simple b at that time...
No: what happen's at the prefix-stem juncture is different from what
happens word-internally. As a result, the beginning of a lexical
morpheme is often a "protected" environment. For example, Germanic
fricatives were voiced after an unstressed syllable, but NOT after an
unstressed prefix (except in some completely obscured prefixations, but
these, by definition, did not count as prefixations synchronically). Old
English voiceless fricatives had voiceless allophones in intervocalic
positions (e.g. hu:[s] 'house' but pl. hu:[z]as), but again not in words
like <geseah> 'saw' (with the prefix ge-).
Piotr