zozeizeizeizeifaho wrote:
> ues wrote:
> > But I don't see how some of the other ones added
> > match it, like "rirny'utka". "rirn" is a four letter rafsi belonging
to
> > "rirni". It also seems to end in a fu'ivla (I think). The CLL says
that
> > rafsi fu'ivla can only start a lujvo. I'm still very confused as to
what
> > is going on here.
>
> This is another (quite old - 2006 or so?) innovation of the BPFK
> morphology.
>
> All fu'ivla can act as a final rafsi after "y", and all fu'ivla have a
> fallback non-final rafsi of "[fu'ivla]'y" that again has to be preceded
by
> "y" if not initial. A "'" is added before non-initial fu'ivla rafsi that
> start with a vowel.
>
> So:
> fuky'utka is fukpi + .utka
> barduku'yjvi is barduku + jivna
> (bardukyjvi being bartu + dunku + jivna)
> ka'ortyglauka is ka'orta + glauka
> norypru'ei is no'e + pru'ei
> (norpru'ei being a single fu'ivla)
>
> You can find the grammar here:
> https://mw.lojban.org/papri/BPFK_Section:_PEG_Morphology_Algorithm


Thanks for this. I am glad that fu'ivla are allowed to be used in initial
and non-initial states now. Somehow I had never ran across any like this
before. One thing though. Doesn't the removal of a final vowel in fu'ivla,
as in "ka'orta" -> "ka'orty" require a final vowel blocking rule similar
to gismu? Is this just for certain fu'ivla?

Comment #6: Re: fu'ivla can now be used in lujvo?

la du kooi noi zo zei zei zei zei fao ku ne vou biu pei (Mon Jan 4 06:42:53 2021)

ues wrote: > Thanks for this. I am glad that fu'ivla are allowed to be used in initial > and non-initial states now. Somehow I had never ran across any like this > before. One thing though. Doesn't the removal of a final vowel in fu'ivla, > as in "ka'orta" -> "ka'orty" require a final vowel blocking rule similar > to gismu? Is this just for certain fu'ivla? This hasn't been formally decided, but afaik, ccV[']VCV and CV[']VC/CV spaces have been kept free of final vowel conflicts so far. Shapes like ccVCVCV and CVC/CVCV don't need a rule like this since their final vowel is never removed.

Comment #7: Re: fu'ivla can now be used in lujvo? Curtis W Franks (Mon Jan 4 23:17:35 2021) zozeizeizeizeifaho wrote: > ues wrote: > > Thanks for this. I am glad that fu'ivla are allowed to be used in > initial > > and non-initial states now. Somehow I had never ran across any like this > > before. One thing though. Doesn't the removal of a final vowel in > fu'ivla, > > as in "ka'orta" -> "ka'orty" require a final vowel blocking rule similar > > to gismu? Is this just for certain fu'ivla? > > > This hasn't been formally decided, but afaik, ccV[']VCV and CV[']VC/CV > spaces have been kept free of final vowel conflicts so far. > > Shapes like ccVCVCV and CVC/CVCV don't need a rule like this since their > final vowel is never removed. I personally think that zevla should be prohibited from matching except for final vowel clusters (after either the final consonant or the final on-glide). Comment #8: Re: fu'ivla can now be used in lujvo? Wesley Wilson (Wed Jan 6 01:45:16 2021) krtisfranks wrote: > zozeizeizeizeifaho wrote: > > ues wrote: > > > Thanks for this. I am glad that fu'ivla are allowed to be used in > > initial > > > and non-initial states now. Somehow I had never ran across any like > this > > > before. One thing though. Doesn't the removal of a final vowel in > > fu'ivla, > > > as in "ka'orta" -> "ka'orty" require a final vowel blocking rule > similar > > > to gismu? Is this just for certain fu'ivla? > > > > > > This hasn't been formally decided, but afaik, ccV[']VCV and CV[']VC/CV > > spaces have been kept free of final vowel conflicts so far. > > > > Shapes like ccVCVCV and CVC/CVCV don't need a rule like this since their > > final vowel is never removed. > > I personally think that zevla should be prohibited from matching except > for final vowel clusters (after either the final consonant or the final > on-glide). I go beyond this. I demand a vowel-blocking rule for all fu'ivla to which this vowel-dropping feature applies. Otherwise we would have to introduce an assigned rafsi system for fu'ivla similar to three-letter rafsi, which in me induces nausea, or there would be certain grammatical ambiguity in any cases where the only distinguishable facet of a vowel-dropping fu'ivla is the final vowel. Has this really not been addressed by the LFK? I honestly cannot even find a resource that discusses this, other than the code you provided, which goes beyond my paygrade. At the very least this rule should be made readily accessible on the lojban website for nintadni and a decision needs to be reached about the final vowel rule, namely whether to create a blocking rule or disallow vowel dropping. I see that xorlo is featured further down on the home page. This rule is comparably important if we don't want users to be confused when they come across rafsi fu'ivla in the wild.