gadri

[Robert LeChevalier <lojbab@lojban.org>]

Nora and I talking have agreed that one solution to the gadri problem 
should be considered, that I don't think has been mention (though I may 
have implied it). It would probably also move us closer to the TLI 
Loglanists, and it would be consistent with other aspects of Lojban.

That change would be to make "le" (and perhaps the entire "le" series) the 
'basic gadri' which would be +nothing-in-particular and -veridical. The lo 
series would add +veridical. Individual cmavo would be added for +specific 
+opaque +intensional or any other feature deemed important enough to 
warrant making a logical distinction, but "le" makes no such distinctions 
and is metaphysically null. (lei, le'e and le'i would each be adding 
features to the null article consistent with their current definition, but 
le could be used to cover a lei or a le'i.)

The English usage that I once made "speaker in mind" would be interpreted 
the way I intended, which is
-specific in that the speaker may not necessarily have specific object(s) 
in mind, or even extant in some world, but merely that there is a reference 
being invoked, which is defined within the speaker's mind, which he is 
trying to evoke in the listener.

I believe this is consistent with usage if not with CLL wording (and I'm 
not sure it is far from the wording), matches the null-default practices of 
tense and other portions of the language. I am not sure that I much care 
what happens to the "default quantifiers" in all this except that lo/loi 
should assume a reference to a minimal subset of those referenced by the 
description and with the relevant feature (i.e. su'o/pisu'o outside 
quantifier) and the "in-mind-described" le and "in-mind-named" la should 
refer to all of those that are in-mind (ro/piro outside quantifier). It 
also allows us a clear and logical solution to all the problems that we've 
come up with using one cmavo apiece for each added feature. The only 
negative is that a very complicated set of features will be a long 
gadri-string (the added feature cmavo could probably be UI), but this is 
consistent - an elaborate tense claim is also long-winded.

Only slightly related to this, and I am less sure that it is supported in 
CLL/usage:
I think that if du'u abstractions refer to facts/relations in the real 
world i.e. being realized, si'o abstractions might be used to refer to 
facts/relations that might or might not be realized or realizable. This 
would clearly allow "le sidbo" a si'o abstraction, to refer to a possibly 
unrealizable, but imaginable relationship. I'm sure I've talked of si'o in 
that way before, but I think I've also talked about si'o as the neutral, 
least restrictive featured abstraction, and I would have to withdraw that 
understanding (again, no idea how this would match CLL policy.)

lojbab

--
lojbab lojbab@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org