Re: Goats' legs and counting

[John Cowan <cowan@SNARK.THYRSUS.COM>]

la mark. clsn. cusku di'e

> I've been mulling over John's comments about how in Lojban something like
> {lo'e remna cu se tuple reda} has to be a complete enumeration.  Something
> about it didn't sit right with me, and I think I have some worthwhile
> comments on it.
>
> John says the enumeration must be complete, that is, you can't say that a
> goat has two legs, even though it has, because to speak completely, you
> have to say it has four.

I say that if and only if you use an exact numeral.  Of course you can say
it has at-least-two legs with no problem.

> Well, let's play with some other expressions:
>
> "He has small hands" can be expressed prettily and lojbanically as {ko'a se
> xance lo cmalu} (I will not be forced to say {lo xance be ko'a cu cmalu}).
> What if I want to say "He has one blue eye"?  John will forbid me from
> saying {ko'a se kanla pa blanu} if in fact he has more than one eye.

I finally grasp what's going on here, after talking with lojbab and pc and
reading a book or two on logic.  No, I don't forbid that, and here's the
explanation:

        ko'a se kanla pa blanu ->
        ko'a se kanla pa lo blanu ->
        ko'a se kanla pa da poi blanu ->
        pa da poi blanu zo'u ko'a se kanla da

which reads:  "There is exactly one X, where X is restricted to blue things,
such that he1 has-as-eye X."  So if he has two blue eyes, then you are wrong;
but if he has a blue eye and any number of not-blue eyes (including zero),
then all is well.

> And don't start telling me it's only with {da}-series words.  Would John
> permit me to say {lo'e kanba cu se tuple re tuple}?  I doubt it; it's
> incomplete.  (I suppose I shouldn't have use {re tuple}, given my
> discussion above; how about {re datci}).

zo datci cu se basti .ei zo dacti

I would reject this, based on the same reasoning:

        lo'e kanba cu se tuple re da poi tuple ->
        re da poi tuple zo'u lo'e kanba cu se tuple da

which reads "There are exactly two Xes, where Xes are restricted to being legs,
such that the-typical goat is-be-legged-by Xes."  Obviously false.

> Nor will I accept that it only works with numerals.  For one thing, that
> breaks a lot of the Lojban model, and for another, everything is implicitly
> quantified anyway.

True, but the implicit quantifications are things like "su'o" and "pisu'o"
for exactly that reason -- so that you don't wind up making claims about
number which are required to be correct.  If you introduce a number into
a bridi, however, you had better mean it.

> Besides, what about "There were many things on the
> floor, among them my foot (one of my feet)": {so'ida cpana le loldi .ije
> pada jamfu mi}.  I'm told the ijek doesn't reset {da}'s, so this should be
> right.

This is okay, because the first use of "da" binds it, whereas in the second
use it is already bound and "pada" means "one of the Xes".

> This seems to be another incarnation of the "only" discussion we had a
> while back, with our different translations of "only".  John would have me
> believe that {ko'a se kanla pa blanu} means "he has one blue eye *only*",
> and that just doesn't seem to work for me.

No, it means "he has only one blue eye."  As explained, he may be be-eyed
(urgh) by any number of other things that are not (contradictory negation)
blue.

--
John Cowan      cowan@snark.thyrsus.com         ...!uunet!cbmvax!snark!cowan
                        e'osai ko sarji la lojban.