[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

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

On 6/7/06, Jorge Llambías <jjllambias@gmail.com> wrote:

On 6/6/06, Maxim Katcharov <maxim.katcharov@gmail.com> wrote:

> Ok, so what does "lo tadni" mean? "The bears" quite definitely implies
> there being some quantity of bears.

{lo tadni} has one or more students as referents, yes. That's different
from saying that the sumti has an outer quantifier, explicit or implicit.

What an outer quantifier, if present, would do is indicate the number of
those referents such that each of them has the property predicated by
the selbri. That's just as in first order predicate logic. But the absence
of a quantifier does not imply that there is at least one of the referents
such that it, by itself, has the property predicated by the selbri. It may
happen that none of the referents has, by itself, the property in question.



That's not the issue in terms of outer quantifiers. That's the issue
in terms of lo/loi in general.

What does {lo tadni} mean to you? To me, it means "the student(s)",
that is, "some [undefined] number of a certain [undefined] 'bunch' of
students". I can't get by without thinking of the students in some
quantitative sense - most, many, all, some, 5, around 20, undefined -
whatever. If there's no number at all, then what are you saying? "the
[no real number of] bears ran"?


> I guess the underlying difference here is that you think the slot
> ({tersum}?) is marked, while I think that the sumti is marked.

Yes, that's the key difference. (BTW, the lujvo would be {tersu'i},
all brivla must end in a vowel, so you can't use a CVC rafsi in the
final position of a lujvo.)

When we say something like {ko'a broda gi'e brode}, the sumti {ko'a}
is filling two slots at the same time, the x1 of broda and the x1 of
brode. This means that if we wanted to mark one of the slots as distributive
and the other as non-distributive, we would be forced to separate the two
slots. One way of doing this is by expanding the sentence:

     ko'a broda i je ko'a brode

now we can mark each slot correspondingly.


How would you mark those two slots?


Another way is to use {ckaji}:

    ko'a ckaji ge lo ka ce'u broda gi lo ka ce'u brode

and again we can mark each slot correspondingly.


brodaness? Why use this method when the above is available?



(I used {ke'a} instead of {ce'u} before. {ke'a} is actually to mark the slots
of relative clauses, and {ce'u} is the word used to mark the slots of
properties.
{ce'u} was one of the last cmavo added to the language and I often forget
it and mistakenly use {ke'a} for properties too.)

To separate two sumti at the same time we can use {ckini}:

  lo tadni lo stizu cu ckini lo ka lu'o ce'u [xi pa] bevri lu'o ce'u [xi re]


The students to chairs are related by [together-those-students carry
together-those-chairs]ness?

Ok, sure. I don't really see what this says though - you can say this
of just about any sentence. It's just a way of rearranging things.
Like saying "the students" refers to the students.

Before I understand how you're using {lu'o}, I'll have to understand
how Alice relates to the surroundment of the building. Right now, to
me {lu'o} expands to something that includes the word {gunma}.



(The subscripts are actually not needed, because each {ce'u} is
taken to indicate a different slot. This is different from {ke'a} which,
when repeated in the same clause, indicates that the slots are filled
by the same thing.)

If we want to separate three sumti, there is no gismu, but we can use
a {lujvo}, {cibyki'i}: x1 x2 and x3 are related by relationship x4
and so on with {vonki'i}, {mumki'i}, etc. Needless to say, this is never
actually done in practice, this is just a way of analyzing things.

> However, you seem to be against using gunma at all. We'll have to
> resolve that first.

I am not at all against using {gunma}, or {girzu}, or {bende}, or {selcmi},
or {kanmi}, or any other gismu that refer to groups of things as single
things. I don't have any problems with them, and I think they can be very
convenient words.


Against using them in this situation. You seem against having

 lei tadni cu sruri lo dinju

expand to

 [da poi sruri lo dinju] cu gunma [le tadni]

as a full, precise, and lossless (and gainless) form.



> I've asked you a question:
> explain how Alice relates to the students that surround the building,
> in a way that is different from how she relates to the students that
> wear hats.

And my answer is: she relates to them in exactly the same way, because
they are the same students in both cases. The relationship is that she is
one of them. Your question has a false assumption, so I can't give you
the answer you want.


So "the students wore hats" and "the students surrounded the building"
are the same? The students each wore hats in the same way that the
students surrounded the building?

No, they're different. In one, each of the students wears a hat - it
is true for each student that they wear a hat. Alice wears a hat.

In my interpretation, the other one would say that each student is
part of a mass-of-students that surrounds the building - it is true
for each student that they are part of a mass that surrounds the
building. Alice is part of a mass that surrounds the building.

You do not adhere to this interpretation. So in your interpretation,
what is true of Alice? That she is a referent of "the students" (that
surround the building)? Yes, but it's also true that she's a referent
of the students that wear hats. She has something to do with the
surroundment of the building? Ok, but that doesn't say anything of her
exact relationship with the surroundment of the building, and it's
also true of the wearing of hats. "...something {lu'o} Alice
something..."? Ok, but that doesn't help me at all. Pretend that I
don't speak in the pluralist variant of Lojban. It's like I don't
speak any french at all, and you have to explain what a french word
means. You shouldn't try to use french to explain it to me, you should
use english. "But there is no first-order explanation - we're on a
higher level here, one of pluralism" - I'm not asking you for a
first-order explanation. I'm asking you for an English one, or
whatever. An explanation that uses words in the way that they are
defined, and when words have more than one sense, this is noted and
explained, and the second sense of the word isn't used halfway through
the explanation.



> We have 26 students in blue shirts stand on 26 marks (the letters A-Z)
> arranged so that they surround the building. We tell them to go
> inside, and call out 26 students in red shirts to stand on the marks.
> Now, both the blue students and the red students had the
> "non-distributive" property of surrounding the building, correct?

Correct.

> Now, who or what really had this property? Each of the 26 students?

No, the 26 blue shirted ones together had the property and also the 26
red shirted ones together had it.

> No: the 26 students together. Well, what does that mean? Why is it
> that I can't swap blue-A, with red-A, and then say that "[the 25 blue
> students and the 1 red student] surrounded the building" ({...sruri lo
> dinju})? Because they're not of the same group/mass.

Right. Or in other words, without mentioning masses, because 25 of the
blue shirted students and one of the red shirted students together did
not have the property. It's not any 26 students at all that have the property
it's 26 particular students that have it.


I think I'll have to wait on your explanation of how Alice relates to
"surrounding the building" vs. "wear hats" before I can understand
what you mean by "together".



> > > Ok. So it's the very same students that surround the building
> > > individually, and that compose the group that surrounds the building.
> > > Right?
> >
> > They wear the hats individually and surround the building together, yes.
>
> No, I mean that they compose the *group* that surrounds the building.
> I expect you to disagree with this.

I don't disagree with that. Both ways of describing the situation are valid:
the students surround the building together, or the group surrounds the
building by itself. It amounts to the same thing. There is nothing strange
about having more than one way of describing one and the same situation.

> > > The issue is what lo means, what lu'o means, what loi means.
> >
> > lo: converts a selbri into a sumti, selecting the selbri's x1 argument.
> > loi: like {lo}, and in addition it marks the slot which the sumti occupies
> >      as non-distributive
> > lu'o: It marks a sumti so that the slot it occupies is non-distributive.
>
> (wait, which one is the one that marks it as distributive?)

An outer quantifier, as in ordinary first order predicate logic.

> Here is a way that the "marking of a slot for distributivity" can be
> explained/expanded:
>
>  lu'o le tadni cu sruri lo dinju
>  lei tadni cu sruri lo dinju
> expands to
>  [da poi sruri lo dinju] cu gunma [le tadni]
>
> Perhaps you have an equally sensible way to expand it? Perhaps even a
> way that does not involve {gunma}?

In terms of singularist first order predicate logic? No, there is no other
way, because first order predicate logic does not have plural reference,
so the only way to do it is the singularist way of introducing a separate
single thing, the mass or group.


There's the very strong implication that this method of expressing it
in "first order predicate logic" is lacking something (whatever
benefit plurality gives). Are they exactly synonymous? If they're
synonymous, then you should have no problem with me translating
anything that you say plurally using that gunma expansion.



> So you would consider
>
>  [da poi sruri lo dinju] cu gunma [le tadni]
>
> to be a correct and complete expanded form of
>
>  lu'o le tadni cu sruri lo dinju
>  lei tadni cu sruri lo dinju
>
> correct?

They would describe the same situation, yes.


Is it a correct, and (especially important) a complete expanded form?



> > > I'm happy with perhaps {loi nanla cu bevri
> > > loi stizu le pa purdi}, because I can choose to see them as a
> > > group/mass that carries over another group/mass.
> >
> > Including a situation where each boy carries just one chair?
> > What do you think of {loi tadni cu dasni loi mapku}?
>
> If each boy carried one chair, I would likely see it as {lo nanla cu
> bevri lo stizu} instead. Otherwise, I can see it in a general sense,
> not care how many boys or how many chairs there really are, just think
> of both as masses, and say {loi nanla cu bevri loi stizu}.

Suppose you just don't know the details. Someone told you, in English,
that the boys took the chairs to the garden, but they didn't say whether
each boy carried one chair, or whether some boy carried more than one,
or whether some boys carried a chair together. How do you report that
in Lojban? If you use {loi}, then it would appear that your {loi} matches
my {lo}: it indicates nothing about the distribution of chairs among the
boys.


No, it does indicate something. It indicates that I'm looking at the
boys as a generic mass of kids. A crowd, if there are that many. It's
not that it *could* mean either one or the other, it's that it means
"mass". If I chose to look at them individually, and see that yes,
they're doing it some specific way, then I could perhaps give a full
expanded form. But, just as I wouldn't care to say the full "A carries
1 and 2, B, C, D with E directing carry a tower of 3 4 5 6, etc.
etc.", I usually wouldn't like to think that way either. Perhaps if I
was some sort of autistic savant, I would. But that's just not the way
people, or at least I, think. This is why we have words like "crowd"
and "forest", and why even learned physicists don't call pencils "a
rod of [insert chemical description of graphite here] surrounded by
[insert chemical description of wood here]".



> But if I
> clearly saw that each boy carried a chair (as, say, each soldier in a
> formation might carry a rifle), then why wouldn't I say it with {lo}?

Of course I can do that too if I want:

    ro lo nanla cu bevri pa lo stizu

The question is, am I forced to specify a particular kind of distribution when
I don't need to specify a particular kind of distribution?



No, you aren't forced to do something you don't need to do. You state
that there really isn't a mass involved in "the students surround the
building", and I disagree with that. I think of it as "Alice is a
component part of that which surrounds the building". It's easy to
understand, I don't quite mention "mass", and it's English. I don't
have an equivelantly easy way to define the (useful!) relationship
between Alice and the building in the pluralist view, even though I
don't have trouble thinking of "the students" (as a bunch) surrounding
the building. (However, when I think deeper of the students as a bunch
surrounding the building, I think of them as being part of a mass.)