[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [lojban] tertirxu
On Fri, 28 Jul 2000, Jorge Llambias wrote:
> I am not saying stripes and spots are not important, but
> {tirxybarna} is perfectly good for that. What I don't get is why
> the relationship between the animals and their marks is so
> exotic for these particular animals that instead of having
> a regular animal place structure for them we get this complicated
> one...
>
> What I am finding difficult about Lojban is learning the right
> place structures. Any patterns that can be discovered are always
> full of exceptions, and the result is that I am usually not sure
> whether or not a given gismu has some mysterious trailing places
> that I am forgetting...
Hear, hear! There's a lot of value to having regular place structures. I
put a fair amount of work into regularizing the place structures in Old
Loglan (an effort that didn't fly), and a lot of semantic categories are
very amenable to a "bed of Procrustes" approach. I was disappointed when
Lojban drifted off in the direction of custom-crafting each predicate's
definition according to which arguments are most important to it.
To my mind, every predicate relates a plethora of arguments both through
the numbered places and through BAI. Most of these arguments are of very
marginal usefulness, and anyone trying to write a computer program to
"understand" Lojban or related predicate logic languages would be well
advised to not try to represent all of them in the database; however, some
of them (like tense) can be very important. I am perfectly willing to have
an important argument lack a numbered place if there's a good BAI to stick
it on, and the value (for learners) of having regular place structures
leads me to be less than generous in welcoming special numbered places for
special predicates.
James F. Carter Voice 310 825 2897 FAX 310 206 6673
UCLA-Mathnet; 6115 MSA; 405 Hilgard Ave.; Los Angeles, CA, USA 90095-1555
Internet: jimc@math.ucla.edu (finger for PGP key)
UUCP:...!{ucsd,ames,ncar,gatech,purdue,rutgers,decvax,uunet}!math.ucla.edu!jimc