Henrik Theiling wrote:
> Hi!
> R A Brown <[log in to unmask]> writes:
>>No - the unstressed, word final vowels seem to have lost quantitative
>>distinctions before this became general elsewhere.
> And did this happen together with a quality change when it happened
> earlier?  (Sardinian 2nd decl. nouns end in _u_, not _o_, indicating
> no quality shift.)

There does not seem to have been a true qualitative change. The evidence 
  we have strongly suggests that the short vowels were lower (and laxer) 
than the corresponding long vowels, except for [a] ans [a:]. In other 
words, the long and short vowels were pronounced much the same as the 
same long and short vowels in Modern German (standard variety  :)

What seems to have happen is this:
1. In the earlier period, quantity (vowel length) was phonemic, with 
non-phonemic qualitative difference.
2. In the later period, under a stronger stress accent, vowel length 
ceased to be phonemic, and the former non-phonemic qualitative 
ddifferences now became phonemic (if you see what I mean).

In the case of Sardinian, [U] merged with [u] and with [o] as it 
generally did in western Romance.

>>In 1st decl. the two cases would just be -a; in the second the [U]
>>of the acc. & [o] of the abl. were practically the same in final,
>>unstressed position. The ablative -e in the 3rd decl. was short in
>>anyway, so again the acc. & abl. would be homophonous.
> Yes for 3rd decl.  I see this immediately.  Although for the 2nd
> decl. this assumes the qualitative vowel system already, right?
> Otherwise we'd get acc. in -[u] vs. abl. in -[o].  

No, its acc. [U] and abl. [o] which.....

> This is still close
> enough for a merger, especially if unstressed. even closer to get a merger.

> I don't doubt the two cases were collapsed by 1st cent. CE, but I'd
> like to understand how this happened (if that is known).  

Phonetic attrition (see above), and growing use of prepositions. Even in 
Classical Latin, with the exception of the ablative absolute 
constructions, noun & pronouns referring to persons could _not_ be used 
in the ablative unless preceded by a preposition. It is but a short step 
to extend this to all nouns. A separate ablative simply became redundant.

> Further, for interesting sound changes I had hoped to have a length
> distinction in the final vowels.  Now you say that that is also
> infeasible?

Personally, I doubt this was maintained in popular speech.

> I see three variants.  In both, I'd assume that the qualitative vowel
> system happened after the split-off of my conlang.  From your
> explanations, I also assume that acc/abl. have merged, and final -m
> has disappeared without traces (i.e., no compensatory lengthening).
> Variant 1:
>     - short/long vowels have collapsed in final syllable, without
>       quality change

Not exactly - see above.

> Variant 2:
>     - short/long vowels have collapsed in final syllable, together
>       with quality change.

Kept their qualities, rather, not changed them - see above.

> Variant 3:
>     - short/long vowels have not collapsed at all, but still, acc/abl
>       have merged
> I like variant 3 most, but you say it is too 'unrealistic', right?

We can't be entirely certain exactly when these things happened - you 
might even have had a particularly conservative dialect. I would 
hesitate to say "too unrealistic". More cautiously, I would say "less 

> So I assume variant 1 is most likely?

Well - 1 and 2    :)

> **Henrik
> --
> Relay 13 is online:
> --
> This email has been verified as Virus free
> Virus Protection and more available at

[log in to unmask]