On Fri, Sep 28, 2012 at 1:08 AM, John Cowan
<cowan@hidden.email> wrote:
Mike S. scripsit:
> I have gradually come to see the light on this idea. As long as there is a
> dictionary editor keeping track of coinages and guaranteeing uniqueness,
> then the mnemonicity of the component stems will surely be sufficient for
> the needs of speakers, and the added machinery needed for unique
> compositionality of form is rather pointless.
That was tried in pre-1983 Loglan, but it didn't work, precisely because
it was impossible to keep the community unified enough. It would be
easier now on the Internet, but unless people are writing with one eye on
the dictionary all the time, colliding coinages will appear. Livagian,
as an artlang, doesn't have this problem, because only And gets to make
up words in practice.
I believe that would happen again, which is part of why I think -y- = [j@, y] should be put into service as the nonce stem hyphen, and full stem coinages should be left in the hands of the designated dictionary committee (on which literally everyone who wants one could have a vote; the point being, there has to be at least a cursory review process before a word gets coined).
With 2-3 letter words, I really don't think you need any more collapsing.
In Gua\spi, where predicate words are about the same size as in Xorban,
lujvo are made by concatenation, and true tanru require a particle.
(The Gua\spi tone system allows the representation of three different
jvajvo relationships by tone alone, an advantage Xorban doesn't have.)
We don't really have the words yet. Out of the 1342 gismu, only 647 yield unique stems in Xorban. Of the others, 283 can be retained, and 412 will have to be reassigned (assuming we kept the list as-is). At the same time, perhaps 50-100 will be chosen for reduced CC form, and an unknown number of other needed predicates, probably 100-200, will be coming in from the cmavo list (nearly 100 have rafsi). Then there will be probably at least 500 (my conservative minimum rough guess) gaps and simple concepts that Lojban covers with lujvo that we might as well coin basic predicates for to round things out while we're looking at this. Essentially, we will be reforging the lexicon from scratch before this is over. And that would just give us a bare-bones vocabulary of 2000 predicates or so, hardly an OED of conlang vocabularies.