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

["Jorge Llambías" <lojban-out@lojban.org>]

On 5/26/06, Maxim Katcharov <maxim.katcharov@gmail.com> wrote:
> On 5/26/06, Jorge Llambías <jjllambias@gmail.com> wrote:
>
> > That's why I don't say that {loi so'i broda} refers to a mass. I only
> > say that it refers to many brodas. To them directly, not to an
> > additional thing on top which contains them. There are some advantages
> > to doing this, i.e. not bringing in unnecessary entities.
>
> I would say that it does refer to a mass, namely, the mass composed of
> brodas. The brodas are referred to in the same way that "the store" is
> in "the hats such that were in the store" - it's not the main
> referent, but a referent regardless. The main referent of a loi is the
> mass itself. I don't understand your comment regarding unnecessary
> entities, since the mass-entity is the most critical aspect.

That's the "singularist" view, the one McKay argues against in the
link I posted. Unfortunately, it appears that only the first chapter is
available online now, his second chapter, "Against Singularism", had
a more complete argument. The singularist view maintains that you
can only make a predication about single things, and the only way to
predicate something about many things is to either say that the same
predication applies to each one, or to group them all into a new
encompassing thing and make a predication about this single thing.
The non-singularist view holds that you can make a predication about
several things without having to distribute it or having to introduce
an encompassing single entity.

> > (Notice that I am not saying that you can never talk about
> > masses/groups/sets/etc., or that it is never useful to do so. You certainly
> > can, as in for example {lo gunma be lo prenu}, "a group of people". What
> > I don't do is use {loi} to intoduce a new entity.)
>
> What do you mean by 'introduce a new entity'?

I mean that I can say something about several people without having
to make reference to another entity that has them as members.

> > (1)    lo gunma be lo ci nanmu cu bevri le pipno
> >         A group of three men carry the piano.
> >
> > (2)    loi ci nanmu cu bevri le pipno
> >        Three men carry the piano.
> >
> > The difference I see between (1) and (2) is that in (1) there is a reference
> > to a new entity, a group, that has three members, whereas in (2) there
> > is no reference to any such entity, there is only reference to three men,
> > which carry the piano together.
>
> The entity is implicit. (2) expands to (1). The word {srana} (it was
> srana, wasn't it?) is not said every time we use {pe}, but it's still
> there.

That's the singularist view of {loi}, yes.

> I'd like to emphasize the difference between "group", {girzu}, and
> "aggregate", {gunma}. Many (all?) things are aggregates, relatively
> few things are groups. "Atom" would be {nargunma}, "that which is not
> composed of any other things" (of course, we've found that atoms in
> physics /are/ composed of other things, and so you would not call them
> this).

Yes, I'm not disputing that various distinctions can be made between
different types of encompassing entities. All I'm saying is that the view that
does not introduce any encompassing entity is, for me at least, the most
useful. That way I can say:

  le mu no tadni cu dasni lo mapku gi'e sruri le dinju
  "The fifty students wore hats and surrrounded the building."

without claiming that there was any single entity that both wore a hat
and surrounded the building.

> When {su'o} is used, you don't mean some special number called "some".
> You mean that the listener should make a best guess as to what the
> number could be (maybe it's five, maybe it's all of whatever number is
> the inner).

I think it would be more correct to say that {su'o} is a special quantifier
(if not strictly a number) called "some". For example, in:

   naku su'o prenu cu klama
   It is not the case that at least some person came.

you don't want the listener to make any best guess for {su'o}. Any number
you replace {su'o} with will make a claim different from the one intended.

> This is exactly the behaviour of the blank outer. If you
> want to say {pa}, but decide to leave it vague, then the blank or
> {su'o} are equally suited for your purposes.

Not really.

> I agree that an implicit {su'o} emphasizes "some, probably not all
> (are X)", which is probably not the intent. Aside from emphasis, there
> is nothing different between the blank outer and {su'o}. They're both
> vague in the exact same way.

That's one of the differences, but in some cases introducing a {su'o}
changes the claim completely. For example:

  lo tadni pu sruri le dinju
  "Students surrrounded the building."

  su'o lo tadni pu sruri le dinju
  "At least one student surrounded the building."

Or:

   lo cribe pu citka ro le mi jbari
   "Bears ate all my berries."

   su'o lo cribe pu citka ro le mi jbari
   "At least one bear ate all my berries."

> > > Do you simply "not commit
> > > to any interpretation"? If one meant {no} as the outer, would it be
> > > acceptable to use the blank?
> >
> > No, one would have to use {no}. Similarly if one meant {ro}, or {su'o},
> > or {re}, etc. one would have to use them.
>
> ...unless one wanted to be ambiguous. Let's say that I had {no}, or
> {pa} in mind. Would it be ok to leave it blank?

Probably not. With some very strong context, perhaps:

A: no ma cu nenri le tanxe
B: lo cribe

A: No what are in the cage?
B: Bears

In that case, the single {lo cribe} will probably be understood as
{[no] lo cribe [cu nenri le tanxe]}, but I can't think of any context in which
you could omit the {no} without omitting the predicate at the same time.
A negation is hard to get from context alone. I suspect an outer {pa},
"exactly one of", would be almost as hard.

> > You can't assume that inserting a {su'o} where no outer quatifier has
> > been used will leave the meaning unchanged, no. In some cases you
> > will get a very similar meaning, in other cases a very different one.
>
> How will it be any different? When I say {lo ci cribe...}, I mean
> "some unspecified number of the bears are [...], and the rest are
> not". I find this exactly equivalent to {su'o lo ci cribe...}.

That's how CLL defines {lo}. Defined like that, they are equivalent, yes.
But when *I* say {lo ci cribe}, I don't mean that.

> > > > {ro lo mapku} = each hat
> > > > {lo ro mapku} = all hats
> > >
> > > All of what hats?
> >
> > All things that count as hats.
>
> All things that count as hats are All hats (existing now, future,
> past, hypothetically, etc.). Your "all" is very different. Your "all"
> means "all of some relevant/contextually-sensible group".

No. As I said before "all hats" is contextually sensible, because
what counts as a hat is contextually sensible, but it does not
mean "all contextually sensible hats" because when you mention
"contextually sensible hats" you immediately bring into the picture
hats that are not contextually sensible as well. In other words, context
sensibility is something that you can discuss in the metalanguage,
when discussing what a phrase means, it is not something that you
can incorporate into the phrase without changing its meaning.


> > > "All of the bears" (in
> > > the zoo), "all of the hats" (in the store). There's a group there.
> >
> > When you make the restriction explicit, yes.
>
> There is no explicit restriction. You said that {lo ro cribe} meant
> "all bears that are in the zoo".

I hope I never said that. "All bears that are in the zoo" would be something
like {lo ro cribe poi zvati le dalpanka}.

> > > So omitting
> > > the outer means that it is ambiguous?
> >
> > Only in the sense that remaining silent is ambiguous between all
> > the things I could have said but didn't. But that's not what "ambiguity"
> > normally means.
>
> Yes, it is. "Open to more than one interpretation" is the standard
> definition, even in the context of linguistics.
>
> http://www.answers.com/ambiguous
> http://www.m-w.com/dictionary/ambiguous
>
> If you want to say that a certain *word* is ambiguous, you say "ambiguous word".

In that case, according to you, is every sentence (in any language) ambiguous?
For example "That's an animal" is ambiguous between "that's a cat" and
"that's an elephant", "that's a cat" is ambiguous betwen "that's a white cat"
and "that's a black cat", "that's a black cat" is ambiguous between "that's
a large black cat" and "that's a small black cat", and so on. Construed that
way, it is not very clear what use the concept of ambiguity may have.

mu'o mi'e xorxes


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.