Re: [lojban] Re: Usage of lo and le

[John E Clifford <clifford-j@sbcglobal.net>]

--- Maxim Katcharov <maxim.katcharov@gmail.com>
wrote:

> On 5/8/06, John E Clifford
> <clifford-j@sbcglobal.net> wrote:
> >
> > --- Maxim Katcharov
> <maxim.katcharov@gmail.com>
> > wrote:

> 
> > it is a dumb distinction or that it can be
> > covered without using {le}/{lo}, then,
> assuming
> > you understand what the distinction is, we
> can
> > have a discussion (but note that the
> specificity
> > of {le} is from one of the oldest strata of
> the
> > language, going back to 1960 or so, so --
> unless
> > you can show how to reproduce it using
> material
> > already at hand, you are unlikely to win this
> 
> I have no solution, though I dislike the idea
> of having a dual purpose
> (veridicity/specificity) to such fundamental
> words.

{le}'s only *purpose* is to indicate specificity.
 It happens that fulfilling this purpose also
allows (maybe even requires) that the
descriptions used need not be veridical.  {le}
cannot be used for sloppy descriptions (referring
to things that don't literally fit the
description) unless the referent are already
specified.
 
> Regarding specificity, what is it useful for?
> The utility of
> veridicity is illustrated by the 'man that you
> thought was a woman
> from a distance' example.

The importance of specificity goes back
historically to the definite descriptor in Logic.
 That is a term-maker which selects the one and
only thing that satisfies the description (and if
there are none or more than one yields a false
claim).  Now we often want to refer to the onnly
two or the only three, etc.  that satisfy a
description  but that could not be done in the
logic of the time (it can be done in plural
logic) so Jim Brown (creator of Lojban's
ancestor, Loglan) modified the notion to
encompass the larger possibility.  He also notice
that, in English, we often use descriptions
loosely (the Juno case, for example: "That woman
is a man") and so he wanted to allow this also. 
When he looked at how logic dealt with this
problem he found (or would have -- I don't think
he really dug in this deeply, but rather intuited
the answer correctly) he found that this kind of
description was possible only if the referent
were already picked out -- that such a
description was in effect a name ("purely
denotative description"),  He therefore declared
that {le} worked in this way.  That gave him the
problem of referring to things not specified and,
after playing with using only quantifiers, he
came up with {lo}, which in Loglan was rather
different from its Lojban form (and a lot less
clear, encompassing at one time or another all of
Lojban {lo} and much of {loi} and a few other
things as weell -- it changed a lot over time). 
So, nonveridicality did play a role in the
development of {lo}, but, once the concept of
specificity was established, its role was
completely derivative.

> Perhaps I should now explain how I see
> quantification, and how I
> incorporate specificity into my understanding
> of how Lojban works.
> 
> Keep in mind that I may explain things in a way
> that isn't similar to
> strict definitions. This is because I strongly
> believe that those
> sorts of definitions shouldn't be used to
> explain anything. Saying
> that something is 'specific' or 'veridical'
> means nothing to someone
> who isn't familiar with the uses of those words
> under the given very
> narrow context. If readers disagree, then point
> out with which
> statement. The veridicity of {_e} is not taken
> in account for the sake
> of clarity.
> 
> {__ __ ro cribe} refers to all bears. At the
> very least all things
> that were, are, and will be bears, everywhere
> (maybe even imaginary
> bears, story-bears, dreamt-bears, or
> hypothetical
> I'm-afraid-a-bear-will-eat-me bears) -
> henceforth "all bears".

Well, no.  {_ _ ro cribe} refers to all the bears
relevant in the present context, which may be
anything from the couple specifed to all the
actual and possible bears.  In a neutral context,
it usually means all current actual bears.

> Now, let's say that you want to say "the bears
> that chase us are brown".
> 
> {__ __ ci cribe} refers to three bears (of
> potentially all bears) -
> but you're not restricting it (any further than
> 'are-bears').

I think the parenthesis here is not necessary and
may be confusing.  It is enough that they are
bears, what pool they are drawn from is not (at
least so far) relevant.
 
> {__ __ ci cribe poi jersi} refers to three
> bears that are chasers
> (are-bears and are-chasers) (of potentially all
> bears)
> 
> {ro __ ci cribe poi jersi mi'o cu bunre} refers
> to three
> that-are-bears and that-are-chasers, and says
> that exactly each of
> them (incidentally three) is brown.

OK
 
> {pa __ ci cribe poi jersi mi'o cu bunre} refers
> to three
> that-are-bears and that-are-chasers, and says
> that exactly one of them
> is brown.

Yup.
 
> {so __ ci cribe poi jersi mi'o cu bunre} is as
> good as ungramattical.

Well, it is contradictory, but it is grammatical
(we can say nonsense very clearly in Lojban.) 
> 
> {ro __ ci vi ca cribe poi vi ca jersi mi'o cu
> bunre} refers to three
> that-are-bears-here-now and
> that-are-chasers-here-now-of-us, and says
> that exactly each of them is brown. (Other
> bears-h-n+chasers-h-n-of-us
> may very well be brown.)

The parenthesis here applies to the earlier cases
as well.  I think the tradition is to attach the
tense markers to the gadri even though they go
with the descriptor's predicate (that is, {lovica
ci cribe}.  But I get confused on these more
often than not. 

> {ro __ ro vi ca cribe poi vi ca jersi mi'o cu
> bunre} refers to all
> that-are-bears-here-now and
> that-are-chasers-here-now-of-us, and says
> that exactly each of them is brown. (all of
> them must be brown)

Yes.
 
> {ro __ ro cribe cu bunre} all bears at all
> times/places are brown
> {ci __ ro cribe cu bunre} three and only three
> bears are/were/willbe brown

Well, all the relevant ones anyhow.

> An inner {ro} finalizes/commits your
> restriction in that it says "no
> other restrictions need apply". If you say:
> 
> {ro __ ro cribe}
> 
> you cannot say that you meant the same
> identity-group/referant/entity/"set" as
>
> {ro __ ro cribe poi bunre}
> 
> unless you show that
> 
> {ro __ ro cribe cu bunre}
> 
> ...good so far?

Let's see: if I say {_ ro cribe poi jersi mi'o} I
can't add further restrictions and be sure I am
still referring to the same thing.  But there is
nothing special about {ro} in this; anytime I add
new restrictions I am in danger of changing the
referent.  Is it that, if I claim to be talking
about all the whatsises, I cannot then change to
talking about the whatsises that are whatever,
whereas, if I talk about some other number of
whatsises, I could later "clarify" by saying "I
mean the whatsises that whtever."  This is true,
I suppose, but I don't see the point.  It is not
that {ro} completes the description in some way
that {ci} does not, it is merely that one form of
weaseling is precluded (though I suspect you can
get away with it even then).  And, I suppose, the
main prolbme is that I don't see what this has to
do with specificity.  You can confine the
description as much as you like, down to the
point where only one thing can satisfy it and
still not be specific (if you have no idea what
if anything satisfies it) and you can have as
open a description as possible "the thingies" and
still be specific.

=== message truncated ===

<<Now, when you have something in mind, like the
pen on my desk, you can
do one of two things:

let context restrict: {ro __ pa vi cu penbi}, {ro
__ pa penbi poi cpane}, ...
restrict it yourself: {ro __ ro vi cu penbi}>>

I am not clear what the point of this
"restricting" is.  If you want to refer to the
pen you have in mind, say {le penbi}; if you 
think that this is not enough to get your hearer
to the right thing you might add the {vi} and the
{pa] and {ca} and {poi cpana} and whatever else
you think is needed.  None of this makes it more
specific, it only helps others get to the right
thing. Nor do various quantifiers all over  the
place (they may in fact be confusing -- as may
any addition).  And, in this context, you could
add all manner of things to {le ro penbi} to help
the hearer and not make it more (or less)
specific.


<<In the second, I've 'fully' restricted it. That
is, >>

Not in the sense that, with {le}, it may not be
appropriate to add more clauses, if the audience
doesn't get it.

<<the restriction
that I've provided exactly matches the pen that I
have in mind
(there's only one pen here-now). Note that {pa}
is not used in the
second. This is because I'm not saying that there
is one pen that
exists here now. (What I was getting at with ro#,
e.g. {ropa}, was
that people just might ask me how many pens exist
by-me-now, and this
would let me preemptively answer that
question).>>

I am not sure whether {ro pa} is legitimate but,
if it is, it probably means about what you have
in mind: "this bunch is all of the relevant ones
and that comes up to one in number"  (I should
note that there is a case made that the internal
quantifiers in {le} descriptions are not
veridical either.)

<<{ro __ ci cribe cu bunre} some three bears (out
of all) each of which
is/was/or will be brown. I'm letting context
restrict. Filling the
blank space:

le: I'm lazy, and I have 3 bears in mind. I don't
want to restrict it
fully, like I did in my pen example.>>

It need not be laziness; it may be that the short
description is adequate, in which case the longer
one is confusing and mildly insulting ("Do you
think I am too dumb to figure which pen you
mean?").

<< Maybe restricting them is as easy
as saying that they're in my back yard here-now
(and those three and
no others are, so I just put in the inner {ro}
and restrict/{poi} it
using 'in my backyard here-now', and I'm set),
maybe it's hard. I just
don't want to do it. So I leave it up to context.
But I do have some 3
bears in mind (that hypothetically can be
restricted-to). Anyway,
those 3 bears are brown.

lo: I don't have three bears in mind. But, I want
to say that three
bears are brown.>>

The {lo} part is no problem, but I still don't
see what ever the quantifier discussion or the
bit about added predicates has to do with the
discussion at hand.


<<ok, I've lost the specificity that you
mentioned, or perhaps I never
had it. See, for that latter one, I'd just say
(in your terms):

{cisu'o le ro cribe cu bunre} (or maybe exactly
ci)>>

Which latter one, I've lost track of the lists.
Certainly {cisu'o le ro cribe} isn't specific
(isn't it {su'o ci}?) 

<<And shouldn't that be enough? Back to your
example:

> Not very clearly.  Stick to simple examples
until
> we get the fundamentals out of the way. 
Suppose

> I say {lo cribe cu citka le jbari}. When I go
and
> check, it does not matter which bears it is
that
> are eating the berries, the statement is true
if
> some bears are eating (have eaten, etc.) the

{su'o le ro cribe cu citka...} ? (or some
specific number, perhaps)>>

Now it does matter which bears are involved. 
This will be false if there are none of the
specified bears involved.

<<> berries.  On the other hand, if I say {le
cribe
> cu citka le jbari}, the statement would be
false
> even if some bears are eating the berries BUT
> they are not the ones I meant.  I know in
advance

{ro le su'o cribe cu citka...} ?

.uacu'i>>

I would take it that the {ro} and the {su'o} are
unnecessary, given that the predication is
distributive (it may not be, but in this case at
least, collective also implies "all") and that
{le} at least always has to have a referent.