Message-Id: <199510232149.RAA03493@locke.ccil.org>
Date:         Mon, 23 Oct 1995 17:32:22 EDT
Reply-To: jorge@PHYAST.PITT.EDU
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: jorge@PHYAST.PITT.EDU
Subject:      Re: Incredible!
To: John Cowan <cowan@LOCKE.CCIL.ORG>
Status: OR

> A LOT of alternatives were considered at that point, and I have yet to
> see a single idea that meets all crtieria of the current language - you
> always seem to need to drop at leats one requirement as "unimportant".

Here is one possible idea:

Define as rafsi all syllables of the form CCV, CCVN, and KVN, where
CC is any of the 48 permissible initials, N is one of l,m,n,r and K
is any of the other 13 Lojban consonants.

There are 240+960+260 = 1460 such syllables, enough to cover all gismu.

(If those were not enough, the 3380 KVKN are available. They just don't
make very nice lujvo.)

Make gismu from those by adding any CV at the end of the CCV and KVN
forms, or a single vowel at the end of the CCVN forms. (Not necessarily
the same vowel or CV for all. The choice may be arbitrary or follow some
rule, classifying words in some way.)

All the gismu so obtained are morphologically like the ones we have.
The rafsi for each is unique and automatically obtainable from the
gismu, and what is even better, no additional 'r's or 'y's are ever
needed, and the rafsi are always one-syllable, except when in final
position. Lujvo are trivially decomposed because each syllable
always corresponds to a different rafsi.

What criterion would this idea have failed to meet?

Jorge