Re: le/lo

[John Cowan <cowan@locke.ccil.org>]

la bab. cusku di'e

> I understand you are saying that when you refer to `a cat', you do
> *not* want to say, `something that I am describing as a cat, but may
> be something else.'

Not so.  When I say "The cat is black" (referring to a cat that
is not well-known to me) I merely mean that it appears to me to
be a cat, that I believe it to be a cat, but *not* that I am
certifying that it is a cat.  The truth value of my sentence
does not depend on the actual cathood of the referent.

Saying "All cats are mammals" does depend on referring to actual cats
only, and not crows, e.g.
 
> Well ... that is the defined use of {le}; veridicality is for {lo}.
> It may be extraordinary, but that how {le} and {lo} have been defined
> for some years now.  Specifically, {le} is non-veridical.  Here is the
> list of LE cmavo:

That part of the cmavo list has not been updated since the discovery
of the "specificity" category, but "le" has always been used for
specific things.  Because of the specificity, the truthful
applicability of the referential predicate has become logically
irrelevant, as in my example above (which does not mean that it
is pragmatically useless).

> For example, someone might say, `this discussion is a quagmire'.  The
> person is not trying to trick you.  She is using a metaphor.  She is
> saying
> 
>     this discussion is not a veridical quagmire, but I am describing it
>     as such, to indicate resemblances.

Ah, but that sentence isn't translatable using "le/lo" before
"quagmire": it is

	le -discussion cu -quagmire

and not either of

a)	le -discussion cu du lo -quagmire
b)	le -discussion cu du le -quagmire

where (a) says that this discussion is identical with some actual
quagmire (false), and (b) says that this discussion is
identical with some specific thing I describe as a quagmire
(not provably false, but bizarre).

>    ... The non-veridicality of {le} seems really only to be there to
>    enable you to use figurative language more easily in Lojban, ...

I think that "le" enables imprecision in description.  Without "le",
we would be compelled to describe everything by its qualities,
as in SQL.  "le co'e" is really "da" with a hint about the
referent.
 
>    ... You don't need
>    to be in the dark about {le} because it should only be used when you will
>    know what the speaker does have in mind for a referent in the first
>    place, or at least when you will know that you won't have to worry about
>    what it is if you haven't got it quite yet.
> 
> Knowing the specified referent does not tell me the referent is true.

Referents are not true or false; only statements are true or false.
Referents just exist (or fail to exist).

> Your default assumptions in English, as indicated by:
> 
>     ... you will tend to describe a cat as a cat.
> 
>     If she really has two cats, there will be strong pragmatic
>     pressure on her to describe them as cats.
> 
>     Basic veridicality, with certain commonsensical side constraints
>     such as what I have already discussed, seems only logical.
> 
> suggest to me that you are misleading yourself and others when you use
> `that which I describe as'.  It appears to me you more likely mean
> `that which really is'.

Again, "really is" is a strong claim; it means that the truth or
falsity of the description-selbri is *essential* to the truth or
falsity of the full bridi:

	mi viska lo cinfo
	I see (lo) lion

is false if what you are seeing is not really a lion.

-- 
John Cowan	http://www.ccil.org/~cowan		cowan@ccil.org
			e'osai ko sarji la lojban