Re: [lojban] fu'ivla tarmi preti

[Pierre Abbat <phma@webjockey.net>]

On Saturday 14 December 2002 19:12, Robert LeChevalier wrote:
> Nora is of the opinion (which I would probably agree with if I thought
> about it) that Cowan (and I before him) were not considering fu'ivla when
> writing most of the morphology rules, because then as now, fu'ivla were
> considered unimportant fill-ins that did not need careful definition (and
> indeed such definition was probably impossible).  Actually, I think we
> thought of fu'ivla as being closer to names in terms of rules than brivla.
li'o
> ></dl><cx "fu'ivla, consonant clusters in">Note that consonant triples or
> >larger clusters that are not at the be­ginning of a fu'ivla can be quite
> >flexible, as long as all consonant pairs are permissible. There is no need
> >to re­strict fu'ivla clusters to permis­sible initial pairs except at the
> >beginning.

Okay, so fu'ivla can have any length of medial consonant clusters, and 
{mastststststststaka} is valid. But the specification left out a couple of 
rules:

If removing the initial syllaboid from string S results in a fu'ivla, and the 
initial syllaboid is a cmavo, then S is not a fu'ivla (e.g. {packankua}, 
{ickankua}, {raumlongena} are not fu'ivla).

If, however, removing the initial syllaboid from S results in a 
valsrslinku'i, and the initial syllaboid is a cmavo, and S has a consonant 
cluster in the first five letters (not counting y'y and ybu), then S is a 
valid word: either the syllaboid is CV, in which case S is a lujvo, or the 
syllaboid is CVV, CV'V, or all vowels, in which case S is a fu'ivla (e.g. 
{eskrima}, {fauspa'i}).

phma