On 20 Oct 2014 22:43, "Jorge Llambías" <jjllambias@gmail.com> wrote:
>
>
>
> On Mon, Oct 20, 2014 at 10:02 AM, And Rosta <and.rosta@gmail.com> wrote:
>>
>>
>> Why must Lojban have onsets and syllabification? /Cia/ is problematic only if /Ci-/ must constitute an onset.
>
>
> There's no special need to realize /Cia/ as one syllable, but it must be counted as one syllable for penultimate stress rule purposes.
How come? Is it simply that because word-segmentation is sensitive to stress, and stress is sensitive to syllabicity, varisyllabicity risks undermining the consistency of the word-segmentation rules? If so, the stress or word-segmentation rules can be reformulated so that VV counts as a single metrical unit rather than a sequence of two.
> Formulating the rule without mentioning syllables would be more complicated.
It's not yet clear to me that that is so, but that could be because I don't know the rule.
> It also has to be distinguished from "ciia"
How come? Is "ciia" licit? Is it licit even if "cia" isn't?
>
>>> I see no problems with {ie'o} as a cmavo form.
>>
>>
>> How about {a'ua}?
>
>
> "a'ua" is currently not allowed by camxes, but "a'uua" is (it's two words)
Ah. So the two-word version needn't be "a'u.ua"? I had been thinking that all words must begin with a consonant.
I would allow "a'ua" (assuming I was not allowed to kill /'/), disallow "uu", and have all words begin with a consonant, but if "a'uua" must be a variant of "a'u.ua" then I see why "a'ua" must be forbidden.
--And.
--