[lojban] Re: A (rather long) discussion of {all}

[John E Clifford <lojban-out@lojban.org>]

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

> To clear something up: I don't want a
> situation-independant antecedent.

I'm not quite sure what you mean by "antecedent."
Your examples later in this post show, however,
that you mean that your description is complete
only in a given situation (indeed, a given
description of a situation), not that it will
work everywhere.  But then the question is -- as
it always was -- how is this not dependent on the
situation (or its description)?  How is this an
advance on ones that feed off the situation
implicitly (but effectively) rather than
explicitly (and perhaps ineffectively -- they are
as fallible as any other)?
 
> On 5/15/06, John E Clifford
> <clifford-j@sbcglobal.net> wrote:
> > --- Maxim Katcharov
> <maxim.katcharov@gmail.com>
> > wrote:
> > > There were no other stones on the table
> > > mentioned, and so I was
> > > operating under the absolutely reasonable
> > > assumption that there were
> > > none.
> >
> > The fact that they are not mentioned does not
> > mean that they are not there; thjey may be
> 
> If I thought that they were there, I would
> properly restrict to them.
> But they were not mentioned, so I assumed that
> they were not there.

See above.  So you want completeness within a
given situation description.  Thus you will leave
out anything not mentioned in tht description,
even if it turns out to play a role in actual
practice.  But, of course, you cannot anticipate
what unmentioned (or indeed what mentioned) thing
 may play a role in locating a referent.  This is
why descriptions sometimes fail and I see no
reason to think yours are significantly  better
than others, for all their extra bulk.

> You're saying that the situation can vary, so
> the same restrictions
> would not apply to every situation. This is
> obvious to me. My complete
> restriction applies to /this/ situation, where
> there are /no/ extra
> stones on the table, and the players were /not/
> playing three
> simultaneous games, and [...]. If any of these
> complexities existed, I
> would account for them by adding one or two
> additional restrictions.

We don't know that they were not playing three
simultaneous games.  If we assume they weren't
then me may have a description that does not fit
the actual situation because we worked from an
incopmplete relevant description.  That is, our
description may not work in reality.  On the
other hand, the conversants would presumably know
the reality and tailor their descriptionsd
accordingly/  But how likely are they to tailor
"complete" ones when ordinary ones will do (for
those who know the situation)?
 
> Stop changing the situation, adding to it, or
> assuming that it's
> vague. It isn't vague, and it doesn't change,
> and it's perfectly
> reasonable for me to assume that if there were
> no other stones on the
> table mentioned in any part of aleks's
> discussion, that there were
> indeed none in that situation.

Reasonable, though possibly wrong -- if he were
describing an actual situation.  As a made-up
case, it is reasonable to think that everything
relevant is in there.  And thus that the ordinary
descriptions will work just fine.  You really
need an open case to make the need for a
"complete" description apparent -- you have to
fend off hidden traps (such as I have suggested
but which you fail to fend off).
 
> Ironically, your persistance in providing
> multiple interpretations of
> aleks's writing stems from the fact that aleks
> was not (could not be)
> precise in describing exactly what the setting
> was, and what his
> speakers were saying. If he had some effective
> means of doing this (as
> I propose) you wouldn't be telling me of all
> the ways that his
> description could be interpreted.

I doubt that he could have presented things so
completely that it would not be possible to slide
in some unmentioned feature.  It is very hard to
predict what might go wrong.

> > irrelevant to what the speaker wants to say
> and
> > still be in the environement, a possible
> source
> > of confusion to the unwary.  Note, by the
> way,
> > that "stones" here presumably means "playing
> > piece," since, depending on the game and the
> > quality of the set, some of these "stones"
> may be
> > shells or nuts or seeds or all may be
> plastic.
> 
> But in /this/ situation, they are not.

This we really don't know.  They are called
"stones,"  which, in a game context is simply
ambiguous. *You* can specify what you want in
your interpretation, but the original does not.


> > If your description is meant to be complete
> as it
> > stands, it may well lead to picking up things
> > that ought not be picked up and leaving
> things
> > that ought be picked up.  Of course, the
> hearer
> 
> Only if I, as a speaker, have no idea what I
> percieve the situation to
> be. I'm going to make this clear: in my
> perception, there are no other
> random stones on the table. They're playing Go
> or Renju, so they're
> using stones, not shells. 

(Actually many pricy Go sets use shell for the
white stones.They have a pleasingly different
sound when snapped down. And, of course, this may
be a cheap plastic set.)

>There are two bags,
> and not 20. When A says
> "in the white bag", they mean "in this white
> bag", and not "in a white
> bag". Etc. My strategy for percieving the
> situation is simple: don't
> add any excess.
> 
> And my description, even by your "one situation
> is actually 100
> different situations" perception, could not
> leave out things that
> ought be picked up, though it could pick up
> what is unintended.
> 
> > > No. I see that you're trying to argue that
> you
> > > can never restrict
> > > absolutely, though you avoid explicitly
> saying
> > > this for some reason.
> >
> > That is not exactly my point, which is that
> any
> > restriction can be restricted farther to meet
> a
> > possible problem and also that many
> restrictions
> > tht you take to be "more complete" are
> simply
> > superfluous in a given case (as here). You
> > imagine problems in this case and so restrict
> to
> > meet these imaginary difficulties but stop
> before
> 
> These 'difficulties' are not imaginary. They
> are quite clear given the
> circumstances: if there are or could be stones
> that are not on this
> table, and the speaker means the ones on this
> table, this is a
> problem, so he restricts it to those stones
> that are on this table.
> 
> > dealing with further ones that I imagine.
> 
> You have yet to present a further difficulty
> (as you term it) that is
> either neccisary, or doesn't stem from what
> seems to be a belief that
> the situation is vague in the mind of the
> speaker.

OK. Let's suppose the speaker has a perception
(description) of the situation and wants to
specify a referent.  He has an expression whihc
is adequate givne this situation description and
he may (maybe also, maybe the same) have a
complete{ expression for the purpose.  The
"complete" expression is adequate (I will grant
for the present purposes cerainly).  The adequate
expression is no longer than the complete one.
So, suppose he uses the adequate one and it
works, the hearer comes to the intended referent.
 In that case, the "complete" expression is
superfluous, if it is different, and useful, if
the same, because it is adequate, not because it
is complete.  On the other hand, suppose the
adequate expression fails, the hearer does not
come t the intended referent.  Since the
expression was adequate to the situation
description of the speaker, it must be that the
hearer is working with a different description,
one for which the speaker's expression is not
adequate.  But, since the "complete" expression
is complete only for the first situation
description, we have no assurance that it is even
adequate, let alone complete, for the hearer
description.  Since it is generally more detailed
than the adequate expression and thus more
closely tied to the original description, it is
likely to fail the new description even more.  So
the "complete" expression reaches the goal of
introducing a new referent only occasionally and
accidentally when the adequare expression fails. 
In the meantime, using the "complete" expression
always will result in generally longer
expressions. It is not clear that the need to
occasionally correct a missed referent results in
a longer average expression than always using the
"complete" expression.  On balance, then, using
adequate expressions seems the more natural
("better") choice. 

> > > You seem to be expanding the example
> towards
> > > including an entire
> > > rainbow of hundreds of stones that are now
> upon
> > > this table. But no,
> > > there are simply black stones, and white
> > > stones. If you want, we'll
> > > call them light stones (>=50% of some
> > > scientific color measurement),
> > > and dark stones (<50%). You're adjusting
> the
> > > situation to make my use
> > > of certain words incorrect, but this still
> > > doesn't show that you can't
> > > restrict completely. The restriction /was/
> > > complete, in accordance
> > > /with the example given/.
> >
> > But more than was needed for the example
> given,
> > therefeore, adding problems seems a
> legitimate
> 
> Not by changing the given situation. If you
> want to "add problems",
> give me a /new/ situation. Otherwise, you're
> just playing no true
> Scotsman with me: "ah, yes, but what was
> /really/ meant by aleks's
> description was probably...", or "but what if
> he actually meant..." -
> - well, I assume that he didn't mean something
> complex and obscure. I
> have to make an assumption based on context
> somewhere, because he
> wasn't (and could not be) specific as to what
> his speakers had in
> mind. Now, my assumption may be wrong, but the
> restriction that I gave
> based on that assumption is not.

As I have said, my additions were in the
direction of getting behind the siutation
description to the situation -- actually, of
course, a more detailed description.  The point
is, if your atempt to establish a new referent
fails, there must be something in the hearer's
situation description that is not dealt with
adequately in your exprssion.  I am suggesting
few things that might be.  You can't ever,of
course, guarantee that you have gotten all of
these in any expression, so any expression may
fail to do the referent introducing work.
  
> > (though "the price of exact
> > > precision is exact verbage" may
> > > apply)
> >
> > Do you mean something secial by the pleonasm
> > "exact precision"?  It does nto seem to apply
> to
> 
> Yes, like when I say "all the stones on the
> table", and mean exactly
> that. Not just the game pieces, not just the
> white ones, not just
> anything. "The group of all such that are
> stones that is supported
> directly by the top of this table". That's what
> I mean by "exact
> precision", by "complete restriction", by "not
> relying on context to
> determine the referent"...
> 
=== message truncated ===
<<Yes, like when I say "all the stones on the
table", and mean exactly
that. Not just the game pieces, not just the
white ones, not just
anything. "The group of all such that are stones
that is supported
directly by the top of this table". That's what I
mean by "exact
precision", by "complete restriction", by "not
relying on context to
determine the referent"...>>

And yet this depends upon at least the table
being clearly specified (is "this" enough? -- and
isn't it contextual anyhow).  That is, I don't
see this as context free.  I also don't think
that this failure is a real problem.  The issue
is to get the right referent and this looks as
likely as anything to do the job, though I think
probably simpler expressions would have worked as
well.

<<> . Regardless, this assumption (that you
> > cannot completely
> > restrict) is wrong.
>
> I await the evidence or the argument for that. 
I
> take it you mean "restrict to the point that
only
> one thing or group of things satisfies the
> description."  That may be true in a given
> situation, but you seem to want one that is
> situation independent and that seems a lot less
> likely.

Oh, I see the problem - I don't want a situation
independant one. I
don't understand where or how you got that idea.

 Your talk about not relying on context just
above, for one.

<< You can't have a
situation-independant antecedent.>>

What does "antecedent" mean here?

<< You percieve the situation to be a
certain way, and you'll restrict based on that.
Your perception might
even be wrong - maybe you didn't notice that
there were some non-game
stones on the table or whatever. But your
restriction would still be
right based on your perception.>>

Is your perception (situation description) what
you mean by "antecedent."  Then, of course, since
it is of the situation it cannot be situation
independent -- except that it may get the
situation all wrong (in some weird absolute
sense).

> > Adding those restrictions does nothing. It's
> > like saying "that which
> > is an elephant... and (poi) that which is a
> > mammal and that which is
> > an animal and that which is a thing...". No,
> > I've given a complete
> > restriction.
>
> Not at all.  You do not specify how the
captures
> are related to the last game. To be sure, in
this

aleks's description of of the situation does not
indicate that there
is more than one way for the captures to have
been related. And surely
you wouldn't call "sneaking a piece off the
board" a capture? If there
was an elephant on the table, it is enough for me
to say "a mammal has
broken the table" - I could, but I don't need to
actually specify
elephant.

> > But I assumed that the referants in the
> > speaker's mind were "all three
> > games of the last set that we played"/"all
> > three just-previous games",
> > so that's how I said it, and based on my
> > (perfectly reasonable)
> > assumption, my restriction was complete.
>
> But as redundant as you claimed my line above
> was.

Redundant to what? Where else was it specified
that "all three
just-previous games". Are you saying that context
has already
specified this? Because that's the exact problem
that I'm trying to
address: when context is the only thing that
"specifies" something,
and it (context) is ambiguous (as it always is,
to some extent).>>

Well, if it doesn't specify that (and I take it
"the last three games" does something along that
line) then where did you get it from.  and if it
does, then why repeat it?  These are not problems
in performance, merely in your description 0f
what is going on --and possibly in where you are
drwing the line between what is there and what
you are adding.  Why do you get assumptions but I
don't -- and may not question yours either?

<<> > You're correct, though it would probably be
> > "all the stones related to
> > the game set (board, bags...)" (here I'm
> > assuming that the speaker
> > doesn't consider that some stones may have
> > rolled into the grass long
> > ago).
>
> Though, if they should find one, it should
> probalby (you don't actually say in your
> "complete" description) go into the bag, too.

I don't see how that's relevant: the speaker is
saying "I think I got
all of the stones into the bag", he isn't saying
"I did my best to get
all of the stones into the bag".>>

But his thought may be wrong in some absolute
sense.  In that case, he has missed the referent
that the description gave if it was "all the
stones related to the game," but done OK with
"all the stones on the table," which is probably
all the speaker had in mind (so that the game
relation was unneeded).

<<> > Here we're deliberating on the best way
> > to make a complete
> > restriction, but this doesn't indicate that a
> > complete restriction
> > cannot be made, which seems to be your
> > position.>>

Actually, my point is just that, even if
"complete descriptions" (in your sense) are
possible -- as they are -- they are not any
better than other devices for getting new
referents (the ultimate point).

<<>
> Well, insofar as restrctions you claim to be
> complete are not, it does count as evidence. 
At

As evidence of what? It just shows that there are
many ways to
complete a restriction.>>

But some leave stuff out that may foil the effort
to pick a referent -- and there is no way to be
sure to get all the possible flaws covered. 
(This is beside the point of course, but as long
as you keep having this notion of "complete"
descriptions, it is useful to point out how
inutile it is.)


<<> Yep, but if I was a speaker, I wouldn't make
this
> restriction. In
> fact, I may very well say "take care not to
trip
> on stones along
> paths" (which wouldn't restrict to ornamental
> stones or this certain
> path at all, and would have a better ring to it
> in Lojban).>>
>
> That is, the "complete" restriction you give
> originally is not the "real" complete
> description?

No, I'm just saying that I would say something
different, something
that has a different restriction on it.>>

So, a complete description is not unique in a
given situation (description).  Or did my
question significantly alter the situation
description? Or both, for that matter.


<<> <<
> These structures may be relevant, but they
would
> be unneccisary. If
> you mean that they are neccisary to provide a
> complete restriction,
> I'd like to know what you think they are.>>
>
> Well, I have set a particularly obtuse
conversant
> into the situation and the strictures are
needed
> to guide him.  On the other hand, your
additions
> are clearly not needed in the situation as
> presented, since the conversant get along fine
> wihtout them.

Who? You mean "a" or "b"? Yes, they got along
fine. I wasn't using
that as a demonstration of where my complete
restrictions are
neccisary, I was just showing what complete
restrictions were and all
that. If you want an example where a complete
restriction is
neccisary, look no further than aleks's
description of the situation.>>

Which is relevant how?  To be sure. since it is a
situation description, everything in it is
necessary relative to that situation description
and it is complete relative to that.  But that is
tautological.  It certainly does not bring us
down to a single situation, the referent of the
example.

<<> No, it is not the case. You'll notice that
both
> the left and the right
> have been taken out of their context, but not
out
> of their setting.
> The ones on the left become vague, the ones on
> the right mean exactly
> what the speaker intends them to mean.>>
>
> Against a more abstract situation, perhaps. 
Once
> you take them out of the situation, there is no
> longer any reason to object to my additions,
> since they were objected to only as not fitting
> the given situation.  But if we stick to the
> given situation, then your additions are
> unnecessary as well.  You can't really have it
> both ways, you know.

No, if we stick to the given context, my
additions are unneccisary
(unless one wishes to be clear in what they're
referring to). If we
stick to the /setting/, then my restrictions are
neccisary, if you
want the listener to know what the referent is.>>

I have probably forgotten what you call context
and what setting.  Context is the linguisrtic
stuff, setting is the mise en scene (what I have
been calling the situation -- or maybe situation
description).  If that is right, I don't see why
they are necessary (for fixing referent, the only
interesting topic).  Notice that the participants
do fine without them.  Therefore they are not
necessary (QED, but I expect you are on some
other point),


<<> <<
> Of the many examples that I've used just in the
> course of this
> discussion, the most prevalent one is probably
> how I'm always forced
> to either say "all (ever) bears", or "*all*
> bears".>>
>
> Are these meant to be examples of "all" used to
> show completeness of the description or are
they
> examples of adding on explanations.  They seem
to

The one with the brackets is the one with the
added (short)
explanation, the one with the asterisks is the
stressed.>>

I would take stress as an added something and so
count it in the same category more or less.  So,
then, what are the examples (which are the ones I
find harder to understand) of "all" being the
mark of completeness?  These seem to be complete
because they give descriptions that selimit a
single referent (admittedly, the emphasis case is
a bit vague), not because they contain "all". The
"all" just days you want to talk about all of the
delimited class, not some.

<<> <<
> The additional explanation is given when the
> restriction is
> incomplete: "all bears (ever, imaginary, past,
> present, future,
> hypothetical...)" - in brackets is the
additional
> explanation.>>
>
> So, if this comes after the "all" is used, how
> does "all" mark completion and, if this is part

"All" doesn't mark completion. "*All*", and "all
(ever)" do (usually,
English isn't good at being consistant). Because
we express that we
mean "*all* such that are...", the listener
doesn't think "all ...
based on context", and so he doesn't have to
determine what the
remaining restrictions are based on context, and
so the restrction
that we've given is a complete one.>> 

I don't suppose there is any point in it, but it
should not be too hard to construct a situation
where even all the conditions given still might
be used to refer to a restricted set of such
bears (the ones in this woods, say).  That is, if
you want independence from the verbal
surroundings, there may not be a surefire way to
do it.  Piling on restrictins is not obviously
enough.

> The CLL also implies that an inner {ro} is an
> assertion regarding how
> many bears exist. My position wants nothing to
do
> with that.>>
>
> So you are proposing something new, as has been
> clear all along.  Further, you want {lo ro
cribe}
> to be of a different order from, say {lo ci
> cribe} or even {lo rau cribe},where the
internal
> quantifier tells how many things are in the
> referent.  Even when {lo ro cribe} referred to

Unless I'm unbelievably mistaken, saying that an
inner {ro} refers to
*all* - well, that tells how many things are in
the referent.>>

But I thought you just said that what {lo ro
cribe} did was signal that the desciption was
complete.

<<> the set of bears, it did not, of course, say
how
> many bears exist, since "all" is a
> merelytautological answer.  It does say how
many
> out of the posible referents are in the
referent,
> however.

Sure, that's the use I want, and it has nothing
to do with that whole
"inner pa means only one exists" confusion.>>

But no one has seriously held to that since a
week after CLL was published (and damned few
before it). It was a careless carry-over from a
problematic situation in Lolgan.  So, why bring
this in here?

> It /is/ about the number of things in the
> referent. When I say "all
> bears", I damn well mean all bears - "all such
> that are bears". Do I
> mean "all relevant things such that are bears"?
> No. I mean "all such
> that are bears". This extra "relevant things"
> that has been tacked on
> seems to /me/ perverse.>
>
> Okay, we can leave off the "relevant" bit (and,
> indeed, usually do).  that does not mean that
we
> are not talking only about relevant bears.  You
> say you damned well mean all bears -- which
ones?

Uh. All bears. I don't mean some bears, I don't
mean just "all the
bears in context", I don't mean just "the ones
that exist". I don't
mean "just" anything. I mean all bears. All
bears. Each and every
single thing such that is/was/etc. a bear. When I
say "all bears", I
always mean this. I had previously stated that I
always meant this.>>

Ah, the philosophical bear set.  I really rarely
mean that except when I am in the philo mode.  I
am surprise that you mean that unless you are in
that mode.  Most often I mean some
spatio-temporarily delimited set of bears.. I
rarely care about talking bears, winged bears and
the like, just about the ones that might affect
me.

<<>  all the ones around here?, all the ones that
> currently exist?, all the ones that have ever
> existed? all the ones that can turn up in
> philosophical ponderings?  If you "obviously"

No, I mean "all bears", which includes all of
these.>>

OK.  You are serious in what you say.  My problem
is, as you will have noted, to figure out when
and for what purpose you will want to refer to
these.

<<> mean one of these, how do you distinguish
this
> case from the others? Do you always mean the
same
> ones?   This seems to me to be the weakest part

Yeah, I do mean the same ones. How could "*all*
bears" ever not mean
the same ones?>>

Different context (or maybe it is setting),
different bears.  You see, speakers of real
languages use context/ setting all the time.  I
think, indeed, they have to to be understood.

<<> <<If we have the clear and simple "all such
that
> are bears", then the listener doesn't have to
> worry that the
> restriction is incomplete, that the speaker has
> ommited anything.>>
>
> Yeah, IF we had that; but, of course, we don't.

Uh, right. If we had it, then I wouldn't be
arguing for its inclusion
into your version of the language.>>

So, since we don't have that -- anywhere, why are
you arguing for its mythical inclusion?  

<<> <<I urge you to try to construct some
> sentances using my proposed definitions, to get
> some idea of how
> they're used. I've done the very same using
your
> definitions.>>
>
> I take it that your examples of your usage are
> correct and they do not do what you want them
to
> do, hence I conclude that my constructing
> examples are even less likely to succeed.

Ok, then don't. I'll continue assume that you
don't understand my position.>>

In which case, since I have given it my best bona
fide effort, I suggest that you reformulate it
with greater care, since what you have said so
far hardly seems contradiction-free, let alone reasonable.


To unsubscribe from this list, send mail to lojban-list-request@lojban.org
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
you're really stuck, send mail to secretary@lojban.org for help.