RE: [lojban] The Pleasures of goi (was: zipf computations & experimental cmavo

[Invent Yourself <xod@sixgirls.org>]

On Fri, 28 Sep 2001, Bob LeChevalier (lojbab) wrote:

> At 10:54 PM 9/28/01 +0100, And Rosta wrote:
> >John:
> > > Invent Yourself wrote:
> > > > no'u probably works like you think goi already does:
> > > >
> > > > ko'a goi la djan. .i li'o .i la fred. no'u ko'a
> > > > ko'a is John. Fred is John.
> > >
> > > By no means: saying "ko'a no'u la djan." is bogus if ko'a is not
> > > *already* defined. The whole point of goi (and cei) is their
> > > defining nature.
> >
> >But you seem to be failing to recognize that ko'a, like cmene,
> >do not have to be defined by goi. They can get their referent
> >from context.
> >
> >English:  It is big.
> >Lojban:   ko'a barda
> >
> >In neither case must it or ko'a have an antecedent in previous
> >utterances.
>
> This may be the case in present Lojban, but when ko'a and goi were defined,
> ko'a was expected to always be explicitly defined with goi at first
> use.  Other usage was *allowed* but at the risk of confusion if nonsense
> resulted.




The possibility of goi-less ko'a should not be allowed to crowbar apart
the simple solution that goi should be taken to define ko'a, and no'u be
taken to refer to an existing ko'a without reassigning it. It is enough to
establish a fail-safe usage that has solid meaning; it is not necessary to
make sure each cmavo can never be used in a dumb manner. That's where this
discussion is going. If somebody wants to intentionally be confusing and
use undefined ko'a, no'u that they feel reassign ko'a, goi that they think
don't reassign existing ko'a, or invisible ink on the back of a
watermelon, that's their option.





-- 
It's said that Mullah Omar has met two non-Muslims in his life. Others say
even that's not true.

Sami ul-Haq, Osama bin Laden's closest friend in Pakistan, runs the
"University for the Education of Truth," a fundamentalist institution that
educated and trained nine out of the Taliban's top 10 leaders.