>> Merge the syntactic classes cmevla and brivla, yes. Brivla would still
>> be divided in four morphological classes: gismu/lujvo/fu'ivla/cmevla,
>> but there is no need for the last class to have its own separate
>> syntax.
>
> I guess the merging would involve changing the stress rules for cmevla a
> little bit.
No, the phonological/morphological rules are totally independent of
what I'm saying. All the phonological rules accomplish is break down a
string of phonemes into a string of words. Once we have those words
identified, how they are then structured to form a text is independent
of the previous step.
Yes, I understand that. What I'm failing to see in spite of my 3 years of Lojban experience is how a listener (not a reader) is supposed to recognise the 'pause after the final consonant' of a cmevla. How exactly is the pause supposed to be realised orally, so as to differentiate the sequence of sounds "
min.ci" from "minci"?