[lojban] Re: cmegadri valfendi preti

[Pierre Abbat <phma@webjockey.net>]

On Sunday 08 December 2002 18:34, Nora LeChevalier wrote:
li'o
> However, you have not considered how it interacts with the rest of the
> speech stream.  For example, if mlongena + nelci becomes mlongenynei (with
> stress, mlongEnynei), then 2 of them is "re mlongenynei", which in the
> speech stream is {remlongEnynei}, which is a lujvo
> (rem+lon+gen+y+nei).  Adding new capabilities for fu'ivla is not as simple
> as it looks.

Okay, so you got me on {mlongena}. Let's try a few others:
bauksita: rebauksitytu'a. Not a lujvo made from gismu, since {auk} isn't a 
rafsi. {rebauksita} isn't a word, since it falls apart. Looks OK.
baknrto: rebaknrtyjirna. {akn} isn't a rafsi, {rebaknrto} isn't a word. OK.
bakrto: rebakrtyjirna. Looks fine to me.
ma'agni: pama'agnyni'e. No consonant cluster in the first five letters, so it 
couldn't possibly be a lujvo. OK.
spatrteace: respatrteacycripu. This starts to break as res-pat- but then 
{rte} isn't a possible rafsi. It can't be {respa trteacycripu} either because 
a word can't begin with {trt} unless it's a name. Similarly for other 
type-3's.

So there's still a wide variety of rafsi fu'ivla, including all type-3's that 
end in one vowel, as well as {bauksita} and {ma'agni}.

phma