Message-Id: <m0qnTgS-00005XC@xiron.pc.helsinki.fi>
Date:         Wed, 21 Sep 1994 16:20:19 BST
Reply-To:     C.J.Fine@BRADFORD.AC.UK
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         Colin Fine <C.J.Fine@BRADFORD.AC.UK>
Subject:      Re: TECH: Any old thing whatsoever (mi nitcu lo tanxe)
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 4780
Lines: 122

coi ken.
..ue .ui rinsa lo se gugdrtaiwana


> I (hi everyone; this's my first post I suppose, apologies if I make
> errors since I'm the opposite-of-experienced-type-of Lojban-thinker)
> see this problem as a need for clarification of what a description
> does when nothing fits the description.
>
..i so'e lo ca jbopensi cu na'e certu .ije so'i ri cu to'e certu


....


> Case 2:
> However, if there are boxes in this world but none fits my need,
> then the sumti "lo tanxe" is valid, but the assertion that "mi nitcu
> lo tanxe" (i.e., >=1 of all boxes in this world is needed by me")
> is false, since I need none of the existing boxes.  Rather, I need
> a box *which does not exist now*.  The problem is, there seems to be
> no description for "something which does not exist".
> Perhaps "loba tanxe"?
>

> Case 3:
> Suppose that there are no boxes in this world.  Then the description
> "lo tanxe" becomes a source of problem itself.  Is the description
> "lo tanxe" == "su'o lo ro tanxe" still valid when no boxes exist?
> I suppose the "ro" part is still fine, but the "su'o" invalidates the
> description (how do you choose >=1 things out of nothing?).

Your analysis is OK, but you're missing the fact that tense is
optional in Lojban. "lo tanxe" doesn't mean "at least one of all the
boxes that happen to exist in the world right now" - it certainly
covers all the boxes that ever have or will exist, and probably (I get
a bit hazy about this) all those that could exist too.

So in your case 2 (and ignoring all the other discussion I haven't
followed about nitcu)

mi nitcu lo tanxe .i je ku'i mi nitcu no lo ca tanxe

If there are no boxes and never could be any, then "lo tanxe" is a perfectly
valid description: it just happens to select no items.

>
> Changing the description to "su'ono lo tanxe" == "su'ono lo ro tanxe"
> makes the sentence a tautology, since it is trivially true that ">=0
> boxes in this world is needed by me".

I am not convinced that this is a tautology: LOGICALLY there is nothing
wrong with the claim

mi nitcu ni'upa lo tanxe        - I need -1 of the boxes

even though it may be hard to assign a real-world meaning to it (I have
one too many boxes ?), and PRACTICALLY it may be entirely moot whether
I need a box or not.

>
> =======================================================================
> I see an analogy here between abstractions and descriptions.
>
> Lojban is able to express false/uncertain statements.  To my
> understanding, saying "lenu..." does not mean that the event took place.
> Nor does saying "ledu'u..." automatically claim that the abstraction
> statement is true.
>
 It is crucial to Lojban to maintain the
distinction between 'bridi' - predications (which have truth values) and
'selgadri' descriptions (which like other sumti designate - or fail to
designate - entities). So while it is true that Lojban can express false
or uncertain statements, eg
        mi ca toldi "I am now a butterfly" (false)
or
        mi ba patfu "I shall be a father"  (unknown)

this has nothing to do with abstract descriptions, where the embedded
bridi is 'mentioned' rather than 'used'.


> So Lojban is capable of handling "nonexisting abstractions", so to speak.
> However, I don't see Lojban handling "nonexisting descriptions".  Maybe
> this is why many of the proposed solutions to the "mi nitcu lo tanxe"
> problem involves abstractions; abstractions can be false while descriptions
> can't; using a description automatically claims that there is something
> which fits the description.  (Am I right here?)

No. An abstract description (such as "lenu ...") designates an
event without making any claim as to whether the event exists or not, and
"ledu'u ..." designates a claim ('predication') without claim that it is
or is not false. Equally a non-abstract description designates an entity
without specifying whether there are in fact any entities which meet the
description or not.

>
> On the other hand, how do I explicitly claim that an abstraction is true?
> I can (1) say so in a separate sentence; or (2) use "lenu...kei noi jetnu".
> So it is also possible to claim the truth of an abstraction, even if you
> have to do it explicitly. (of course!)

Precisely! Only an independent bridi (a sentence) is capable of making any
claim whatever!
Bringing up 'noi' is interesting, because unlike 'poi' it does appear to
make a claim - but I'm not convinced.


>
> So how do I use a description without claiming that anything fits that
> description?  Something "hypothetical"?

Use lo/loi/lo'i and you are not making that claim. If you want to make the
claim, introduce the entity in a separate sentence, or (extra-logically)
use the discursive "da'inai" (not irrealis).

Keep on asking!

        Colin Fine

>