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

4 views
Skip to first unread message

Jorge Llambías

unread,
Jun 9, 2006, 11:13:20 AM6/9/06
to loj...@yahoogroups.com
On 6/9/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> For
> example, I think that the following is grammatical under your rules,
> but I find the interpretation infeasible (or awkward), since it uses
> both "markers":
>
> lu'o ro lo tadni cu sruri lo dinju

For me that's equivalent to {lu'o lo ro lo tadni} and it marks the slot
that the sumti fills as non-distributive. It is only the outermost marker
that concerns the distributivity of the slot.

> > But now the disagreement seems to occur at an earlier step.
> > We no longer seem to agree about the meaning of (1). For example,
> > if there is a chain of rocks around the building, with a gap filled by
> > Alice, who is a student, you would say that (1) is true and I would
> > say it is false.
>
> loi rokci cu sruri le dinju
>
> What would you say of that, assuming that the surroundment wouldn't
> work if Alice wasn't there? Would it, too, be false?

Right. It is not the case that rocks surround the building.

> As I've said, this is an issue of pragmatics. The speaker would likely
> say "[the rocks and Alice] surrounded the building" in the first
> place.

Yes: {lo rokci joi la alis cu sruri le dinju}.

But apparently under your current interpretation, from that it follows that
{loi rokci cu sruri le dinju} and also that {lu'o la alis cu sruri le dinju}
(= {lai alis cu sruri le dinju}?). But neither of those follow at all, the
way I understand it.

> > For me it is false because it is not the case that
> > students surround the building in that case. For you it is true because
> > it is the case that a group of things that includes at least one student
> > surrounds the building.
> >
> > Do we at least agree on what we disagree about?
>
> I think so

OK, it is a definitional matter then. You may want to argue for your
proposed definition, proposed as a change, but I don't think you can just
assert that it is the one that Lojban currently has or ever had.

(And this new argument is independent of the one we were having previously,
on whether or not the distributivity marker should be obligatory or not.)

mu'o mi'e xorxes


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

Jorge Llambías

unread,
Jun 5, 2006, 1:49:25 PM6/5/06
to lojba...@lojban.org
On 6/5/06, John E Clifford <cliff...@sbcglobal.net> wrote:
>
> > > As you may recall, my suggestion mirrors the
> > > English for "individually" and "collectively" (or
> > > "together"), attaching as convenient to sumti or
> > > predicate place (so, I suppose that UI is about
> > > the only selmaho that will work -- unless we
> > > invent a new one).

Here is a mathematical proof that it cannot be done with a
single word. Consider the sentence:

ko'a ko'e ko'i ko'o ko'u broda gi'e brode

It consists of eight words, so there are exactly nine places where
a word meaning "together" could be added.

But there are ten argument places it could be modifying:
each of the five slots of broda and each of the five slots of brode.

Therefore, a single word, no matter how strategically well placed, could
not be used to do the job of unambiguously marking any sumti slot
as collective.

John E Clifford

unread,
Jun 6, 2006, 3:52:55 PM6/6/06
to lojba...@lojban.org
Thanks. I have to say that seeing the
convolutions involved convinces me yet further to
work for some one-word solution.

--- Jorge Llamb�as <jjlla...@gmail.com> wrote:

> On 6/5/06, John E Clifford
> <cliff...@sbcglobal.net> wrote:
> >

> > On the
> > other hand, the present system leaves some
> cases
> > not expressible at all or at least I don't
> see
> > how to do it in a regular way. How do you
> > distinguish in "The carriers of the pianos
> wore
> > green hats" between individual carriers and a
> team?
>
> I suppose something like:
>
> ro da poi lu'o ke'a bevri le pipno cu dasni
> lo crino mapku
>
> ro da poi ro ke'a bevri le pipno cu dasni lo
> crino mapku

Maxim Katcharov

unread,
Jun 9, 2006, 11:22:59 PM6/9/06
to lojba...@lojban.org
On 6/9/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 6/9/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > For
> > example, I think that the following is grammatical under your rules,
> > but I find the interpretation infeasible (or awkward), since it uses
> > both "markers":
> >
> > lu'o ro lo tadni cu sruri lo dinju
>
> For me that's equivalent to {lu'o lo ro lo tadni} and it marks the slot
> that the sumti fills as non-distributive. It is only the outermost marker
> that concerns the distributivity of the slot.

1) {lu'o lo ro lo tadni}

You can potentially have

2) {lu'o ci lo ro lo tadni}

can you not? If you have (2) in mind, you can just say (1). So (1)
could be (2). (1) could even potentially be:

3) {lu'o ro lo ro lo tadni}

If your grammar says that an outer {ro} on a {lo} marks it for
distributivity, why is (1) exempt from this rule?

The explanation you gave was that you see it as "students (surrounded
the building)". Is that an explanation of {lo tadni}, or {lo ro lo
tadni}?

I asked you what {lu'o ro lo tadni} was. You said {lu'o lo ro lo
tadni}. So an outer {ro} automatically gets a second {lo}? Or is that
only when it's needed to excuse an infeasible aspect of your position?

>
> > > But now the disagreement seems to occur at an earlier step.
> > > We no longer seem to agree about the meaning of (1). For example,
> > > if there is a chain of rocks around the building, with a gap filled by
> > > Alice, who is a student, you would say that (1) is true and I would
> > > say it is false.
> >
> > loi rokci cu sruri le dinju
> >
> > What would you say of that, assuming that the surroundment wouldn't
> > work if Alice wasn't there? Would it, too, be false?
>
> Right. It is not the case that rocks surround the building.
>
> > As I've said, this is an issue of pragmatics. The speaker would likely
> > say "[the rocks and Alice] surrounded the building" in the first
> > place.
>
> Yes: {lo rokci joi la alis cu sruri le dinju}.
>
> But apparently under your current interpretation, from that it follows that
> {loi rokci cu sruri le dinju} and also that {lu'o la alis cu sruri le dinju}
> (= {lai alis cu sruri le dinju}?). But neither of those follow at all, the
> way I understand it.

As I've said, this isn't English, you don't need the same pragmatics
and verbatim translations. If I respond (1) "he came to work sober
today" to "was he drunk?", it does not follow that he was drunk on
other days. Even if I said (1) without provocation, it does not
logically follow that he must be a drunkard, it only pragmatically
'follows'.

Now, I could say {lu'o la alis cu sruri lo dinju}, but usually I
wouldn't. I'd say {lo rokci joi la alis cu sruri lo dinju}. However,
the former would still be true - Alice is a part of the
surrounder/surroundment of the building.

>
> > > For me it is false because it is not the case that
> > > students surround the building in that case. For you it is true because
> > > it is the case that a group of things that includes at least one student
> > > surrounds the building.
> > >
> > > Do we at least agree on what we disagree about?
> >
> > I think so
>
> OK, it is a definitional matter then. You may want to argue for your
> proposed definition, proposed as a change, but I don't think you can just
> assert that it is the one that Lojban currently has or ever had.
>
> (And this new argument is independent of the one we were having previously,
> on whether or not the distributivity marker should be obligatory or not.)
>
> mu'o mi'e xorxes
>

I'd still like to have that explanation of distributivity that I've
been asking for.

1) {lu'o la tadni cu sruri lo dinju}
2) {loi tadni cu sruri lo dinju}
3) {la tadni cu sruri lo dinju}
4) {la tadni cu dasni lo mapku}

What is Alice's relationship to each relationship? I've explained
exactly what it means for all 4 - 1&2 would be expanded in the manner
described (which you still seem to be at odds with), and 3&4 would not
expand, and could not mean 1 nor 2. 3 would probably be false. If I
were to be lazy, and in my mind thought of them as a mass that wore a
mass of hats ("oh look, they're all wearing hats"), I could very well
say {loi tadni cu dasni loi mapku}. That is my position.

You still have not offered an explained how Alice, a student, relates
to anything in a useful way. The explanations offered apply to
most/all of these at the same time, and so aren't useful in terms of
explaining what each is. I've offered an explanation of my position in
English, and since this pluralist view of yours /clearly/ isn't
axiomic, you should either be able to do the same. If not, then I
think that the sensible person can assume that it's not the proper way
to look at things. The part that I find particularly strange is that

{lo tadni cu sruri lo skori}

can expand to either

{[da poi sruri lo skori] cu gunma [la alis]} or
{lo tadni cu sruri lo skori}

where the second one, by your methods, is actually described by

{lo lo tadni cu sruri lo skori}

If it were

{lo tadni} >
{lo pu tadni} or
{lo ba tadni}

then, yes, that's perfectly fine. But

{lo tadni cu sruri lo skori} >
{[da poi sruri lo skori] cu gunma [la alis]} or
{lo lo tadni cu sruri lo skori}

? I can't imagine the rationale for /that/.

Maxim Katcharov

unread,
Jun 9, 2006, 9:36:23 AM6/9/06
to lojba...@lojban.org
On 6/9/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 6/8/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > On 6/8/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > > Suppose Alice and Betty are the only students
> > > in the group, and the rest of the people surrounding the building are
> > > professors, is "Students surround the building" true?
> >
> > Yes, it becomes much more apparent (to me, at least) once written out
> > in Lojban. It's not {la alis cu sruri lo dinju}, it's {lu'o la alis cu
> > sruri dinju}.
>
> My understanding of our disagreement so far had been as follows:
>
> (1) loi tadni cu sruri le dinju
> (2) lu'o lo tadni cu sruri le dinju
> (3) lo tadni cu sruri le dinju
>
> We both agreed (I thought) in which situations (1) would be true and
> in which situations it would be false. We both agreed that {loi} could
> be decomposed as {lu'o lo}, so that (2) says the same as (1). We

Yes, I think

> disagreed on whether the {lu'o} component could be optionally omitted.
> I held that {lu'o} was optional, and so could be dropped when context
> made it unnecessary, as in the example of students surrounding a
> building, and you held that {lu'o} could not be dropped because if it
> was dropped the predicate {sruri} had to be necessarily taken as
> distributive.

Yes. We also disagree on how non-distributivity should be marked. For


example, I think that the following is grammatical under your rules,
but I find the interpretation infeasible (or awkward), since it uses
both "markers":

lu'o ro lo tadni cu sruri lo dinju

>


> But now the disagreement seems to occur at an earlier step.
> We no longer seem to agree about the meaning of (1). For example,
> if there is a chain of rocks around the building, with a gap filled by
> Alice, who is a student, you would say that (1) is true and I would
> say it is false.

loi rokci cu sruri le dinju

What would you say of that, assuming that the surroundment wouldn't
work if Alice wasn't there? Would it, too, be false?

(In my mind, I find it hard to picture Alice and the rocks as forming
a group in the first place. Masses aren't "forced" in the same way
that members of a mathematical set can be. Just naming a random
collection of objects and calling them a mass doesn't seem right,
unless you can clearly visualize that they together "are". But this is
all mostly speculative/intuitive.)

As I've said, this is an issue of pragmatics. The speaker would likely
say "[the rocks and Alice] surrounded the building" in the first

place. For the case of the many students, and the one professor, Zoe,
the speaker would likely say "the people surrounded...", or "the
academics surrounded..." in the first place. Using the example of the
single professor and the many students (equivalent to the Alice/rocks
example), the following would be false:

la zox cu sruri lo dinju "Zoe surrounds the building"
[la zox e lo tadni] cu sruri lo dinju "Zoe and the students surround
the building"
lu'o po'o la zox cu sruri lo dinju "Zoe is the only thing the
surrounder/surroundment of the building consists of"

and the following would be true:

lu'o la zox cu sruri lo dinju "Zoe surrounds the building"
lu'o [la zox e lo tadni] cu sruri lo dinju "Zoe and the students
surround the building"
loi tadni cu sruri lo dinju "the students surround the building"
loi ctuca cu sruri lo dinju "the professor surrounds the building"
lu'o po'o [la zox e lo tadni] cu sruri lo dinju "Zoe and the students
are the only thing the surrounder/surroundment of the building
consists of"

> For me it is false because it is not the case that
> students surround the building in that case. For you it is true because
> it is the case that a group of things that includes at least one student
> surrounds the building.
>
> Do we at least agree on what we disagree about?

I think so

>


> mu'o mi'e xorxes
>
>

Adam D. Lopresto

unread,
Jun 7, 2006, 11:10:34 AM6/7/06
to lojba...@lojban.org
On Tue, 6 Jun 2006, Maxim Katcharov wrote:

> "Mass"/"together" expands to "x1 is a mass with components x2". This
> is an actual relation. I consider that as significant in terms of
> content as you can get.

That's pretty much exactly the definition of {gunma}. "x1 is a
mass/team/aggregate/whole, together composed of componets x2, considered
jointly." It seems pretty clearly that x1 is a mass, and that x2 is a plural
reference.

.i lei tadni cu gunma le tadni
The mass of the students is a mass formed from the students.

But if you don't permit plural reference, then what can possibly fill the x2
there? It certainly can't be each student considered seperately (ro le
tadni). And it can't be the mass of the students, since that's exactly
what's in the x1. So what is it?

And in case you're tempted to say it's actually the set, consider the
definitely of mei

x1 is the mass formed from set x2 whose n member(s) are x3.

lei re xa tadni cu re xa mei le'i re xa tadni le re xa tadni
The mass of the 26 students is a mass with 26 members, formed from the set of
the 26 students whose members were the 26 students themselves.
(Yes, that's horribly redundant. Oh well.)

So as far as your asking for proof that the language permits plural
predication (predication on several entities together, but distinct from a
distributive predication on individuals, or a predication on a mass), that
certainly seems like it to me. I'm curious how you interpret that, or what
you could possibly see filling the x2 of {gunma} or the x3 of {-mei}.
--
Adam Lopresto
http://cec.wustl.edu/~adam/

Life is like a simile.

Maxim Katcharov

unread,
Jun 11, 2006, 10:24:03 AM6/11/06
to lojba...@lojban.org
On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> On 6/11/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > > On 6/10/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > > > An outermost quantifier is distributive,
> > > > an outermost mass-marker is non-distributive. It's that simple.
> > >
> > > No doubt it's a simple rule, but the question is if it's the proper
> > > way to describe how we communicate.
> >
> > It's a proper way to describe how I understand Lojban works.
> > That's all I can offer you.
> >
> > > > la .alis .e la betis = ro le re prenu
> > > > la .alis joi la betis = lei re prenu
> > > > la .alis jo'u la betis = le re prenu
> > >
> > > Are you offering these definitions as suggestions, or
> > > as explanations of how it really is?
> >
> > As explanations of how I understand them to work in Lojban.
>
> pagbu and gunma, based on the information (from CLL and the gismu
> list) that I provided, seemed to be the corresponding ideas. What are
> you basing your interpretations on?
>
> >
> > > So you would disagree that

> > >
> > > loi tadni cu sruri lo dinju
> > >
> > > expands to
> > >
> > > [da poi sruri lo dinju] cu gunma [lo tadni]
> > >
> > > which you had previously described as a legitimate interpretation,
> > > yes?
> >
> > Using your place structure for {gunma} with distributive x2, I
> > would disagree, yes. I agreed when I thought we were using
> > {gunma} with the same meaning.
>
> This is an expansion of non-distributivity. By its nature, one should
> assume that it doesn't use yet more non-distributivity to explain
> itself.
>
> >
> > > {loi tadni cu sruri lo skori} and {ro lo tadni cu
> > > sruri lo skori} are different in some way, yes?
> >
> > Yes.
> >
> > > I'm asking you to explain the difference.
> >
> > I've already already done that. (And besides you understand
> > the difference perfectly well.)
>
> I understand *my* difference perfectly well. So far you've given me
> "differences" that apply to both distributive and non-distributive.
>
> The surest way to show that I'm a fool for asking this 30th time is to
> point me to an explanation that I haven't rightly shown to be
> unexplanatory.
>
> >
> > > Alice is part of X
> > > X surrounds the building
> > >
> > > Alice wears a hat
> > >
> > > Is the difference not apparent there?
> >
> > Yes. In the pluralist view one would say instead:
> >
> > Alice is one of X
> > X surround the building

Wow, I sure let that one slip by.

"one of" is *completely* different from "part of".

>
> In which case you'd no longer be using first order logic. Now, if the
> axiomic nature of first order logic is not available to you, then you
> should use something else. But you can't just say that your pluralist
> interpretation is axiomic itself, because it clearly isn't.
>
> And what's the difference between that and
>
> Alice is one of X
> X wear the hats
>
> Clearly, Alice does not relate in the same way to the wearing of hats
> as she does to the surroundment of the building, but you seem to gloss
> over this *critically* important point. What's the difference between:
>
> Alice is one of X
> X surround the rope
> (e.g. they're standing around a rope)

Difference is:

Alice is one of X
X are parts of Y (single thing)
Y surrounds the rope

>
> and
>
> Alice is one of X
> X surround the rope
> (e.g. they're playing tug-of-war)

and

Alice is one of X
X each surround the rope

>
> >
> > In the singularist approach, X is a single thing, distinct from the
> > things that surround the building and consisting of them, while
> > in the plural approach X are many things, just those very things
> > that surround the building.


> >
> > mu'o mi'e xorxes
> >
> >

John E Clifford

unread,
Jun 5, 2006, 8:34:17 PM6/5/06
to lojba...@lojban.org

--- Jorge Llamb�as <jjlla...@gmail.com> wrote:

> On 6/5/06, John E Clifford
> <cliff...@sbcglobal.net> wrote:

> > Very nice. But, since the terms might be
> > differently predicated on the two brivla, we
> > would have to place the markers on the
> predicate
> > not the term.
>
> Yes. Or, we can do things like:
>
> le tadni cu ckaji lo ka ro ce'u dasni lo mapku
> kei .e
> lo ka lu'o ce'u sruri le dinju
> "The students have the properties that each of
> them wears a
> hat and that they together surround the
> building."
>
> This effectively separates the term (le tadni)
> from the slots (ce'u),
> and so we may mark the slots directly, at the
> cost of having to
> introduce the {ckaji} predicate, of course.
>
> > This seems to mean that these
> > markers would themselves be strung out like
> terms
> > with the predicate, to be properly
> correllated
> > with the terms in the term list, probably
> > something like (using d and c for distrib and
> > collect)
> > ko'a ko'e ko'i ko'o ko'u broda d c c d c *
> brode
> > c d d c d (* because I suspect the
> conjunction
> > at this point will change since we are
> connective
> > bridi tails not selbri). Not yet a very
> comely
> > system, though it would be rare that we would
> > need to show all the markers. Still, it does
> > circumvent the problem you not (and, no doubt
> > creates other for the parser or interpreter).
>
> Yes, that would be quite doable, but it forces
> you to mark
> all terms up to the one you are interested in
> marking. For example,
> if you wanted to mark the distributivity of the
> third slot, you
> would be forced to mark the distributivity of
> the first and second
> slots too. You could also just use a subscript
> or something like
> fa/fe/fi/fo/fu to choose which slot was being
> marked.


I think I had fee-fi-fo-fum in mind.

> In any case, it would seem that any system
> would require some
> wordiness, so it is not obvious to me that the
> current one is
> particularly inefficient.

Well, precision does require some wordiness. I
don't expect that this system would be used all
that often, certainly not for 10 places. On the


other hand, the present system leaves some cases
not expressible at all or at least I don't see
how to do it in a regular way. How do you
distinguish in "The carriers of the pianos wore
green hats" between individual carriers and a team?

Jorge Llambías

unread,
Jun 5, 2006, 3:56:22 PM6/5/06
to loj...@yahoogroups.com
On 6/5/06, John E Clifford <cliff...@sbcglobal.net> wrote:

In any case, it would seem that any system would require some


wordiness, so it is not obvious to me that the current one is
particularly inefficient.

mu'o mi'e xorxes

Jorge Llambías

unread,
Jun 6, 2006, 9:01:36 AM6/6/06
to lojba...@lojban.org

I suppose something like:

ro da poi lu'o ke'a bevri le pipno cu dasni lo crino mapku

ro da poi ro ke'a bevri le pipno cu dasni lo crino mapku

mu'o mi'e xorxes


John E Clifford

unread,
Jun 10, 2006, 11:28:46 AM6/10/06
to lojba...@lojban.org
<<On 6/7/06, John E Clifford
<cliff...@sbcglobal.net> wrote:
> --- Maxim Katcharov <maxim.k...@gmail.com>
> wrote:
>
> > On 6/6/06, John E Clifford
> > <cliff...@sbcglobal.net> wrote:
> > > Well, I suppose that Alice's relation
> > surrounding
> > > the building (when she is one of the
students
> > > surrounding the building)is
"participation."
> > I
> >
> > Participation in an event? xorxes already
> > offered this. Consider "the
> > students surround the students". What is
Alice
> > participating in?
>
> Well, is Alice among the surrounding or the
> surrounded? Those seem to be the two events in
> which she could participate. In the one case
she
> is (more or less) on the outside looking in, in
> the other on the inside looking out.

Sure, I guess. I don't think that this helps much
in terms of
explaining it, though. She participates in the
wearing of hats too,
after all.>>

Not exactly; participation is the flip side of
doing things together. In the usal case, wearing
hats is done individually. She might, however,
particpate in a hat wearing demonstration, say,
and do that by wearing a hat. The comment was
just to point out that you had incompletely
specified the question, making an answer
difficult.

<<>
> > > suppose that giving it a name is not going
to
> > > satisfy you (quite rightly) but if I lay
out
> > the
> > > formal specifications of the relation, you
> > will
> > > just say "Oh, that's just membership in the
> > > group."
> >
> > Yes, that's exactly what I'll say, because
> > that's exactly what it is.
> > It's a mistake to think that masses can only
be
> > physical lumps of
> > something. For example, 1000 people can be
> > foolish each (by gathering
> > fools together, and inciting them each to do
> > foolish things), or
> > together they can "participate" in a
> > large-scale foolishness, without
> > being foolish each. What this is saying is
that
> > they're component
> > parts of an action, the action of being
> > foolish. Same thing, different
> > perspective, still a mass.
>
> This is beginning to look like your sense of
> "mass" or "group" or whatever is less about the
> things involved and more about what they are
> involved in. That is dangerously close to
making
> the distinction between distributive and
> collective predication but in (as in Lojban)
> misleading terms.

It's equally about things involved and what
they're involved in. But
in the end, it's the thing that the students
compose that does the
surrounding, and not the students themselves. I
don't care which one
of

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

lu'o le tadni cu sruri lo dinju

expands to, and I see the difference between the
two.>>

Huh? Which two � you offer three? Do please
decide whether there is something besides the
students involved here. If there is an ontic
group, then one line of chat is appropriate; if
there is only the students considered in a
certain way or some such locution, then another
is. In short, please finally give an explanation
of what �a group of students� means in real terms
(i.e., without falling back on the �group, mass,
set,�� idioms, which are question-begging).

<<Perhaps you could make this disinction between
"distributivity" and
"non-distributivity" in a way that (usefully!)
explains what relation
Alice has to the surroundment of the building?
The way I see it,

[da poi sruri lo dinju] cu gunma [la alis]
"alice is part of that which surrounds the
building" or
"alice is part of the surroundment of the
building"

I think that that's perfectly reasonable.>>

Depending on what it means, it is or is not, and
the latter depending upon what theory you have of
bunches. It MAY mean just what �Alice
participates in surrounding the building� (or
�Alice is among the surrounders of the building�)
means or it may say something about an entity (a
mass, apparently)and that entity may be either a
bunch in the technical sense or some other sort
of entity. If it is a bunch in the technical
sense, then again it means the same as �Alice
particpates in surrounding the building� or
�Alice is one of the surrounders of the
building.� If not, then it is just plain unclear
what it means and so how it is related to the
pluralist version. Which is it?

<<>
> > > Or if I try to specify it in extension,
> > > spelling out how she particpates (standing
> > NEbyN
> > > of the building at the same time as others
> > are
> > > standing at the other points of the
compass,
> > say)
> > > you will relate that to being a member of
the
> > > group as well.
> >
> > Well, yes. This is the method of
participation.
> > For example, I can say
> > "together the three men lifted the piano, by
> > method of one man
> > directing, and two men bearing".
>
> This tells me what each does by way of
> participating, but I still don't see anything
> like a group here unless it is just the fact of
> the particpation being described in some
> organized way. And that is just what a
pluralist
> would mean by "together," more or less.

What lifts the piano? The three men, right? What
relation does Avery
have to the lifting of the piano then, if he
doesn't lift it himself,
and he's not part of the mass that lifts it?
(This is the same thing
as with Alice, and the things that should be
noted there should be
noted here.)>>

Well, it appears that Avery has no relation to
the lifting of the piano, except, perhaps, for
non-particpation. Or is Avery the one directing
� in which case he is, apparently, part of the
mass that lifts it. Since you have yet to explain
what a mass is in these cases, I am unsure what
your intended answer is here. Alice�s role in
participation was just given, was Avery�s? Your
masses are beginning to sound a lot like the
oldest definition of sets: �things considered
together� (though, of course, you can�t say
that).

<<There is no real explanation>>

Well, not from you anyhow � repetition is not
explanation.

<<- none that I can think of, and none that
you've provided (correct me if I'm wrong). Yes, I
can see how,
intuitively, one may think of it that way, but
there are a lot of
things that we sense intuitively that are wrong.
In order to
understand it, I need an explanation. This isn't
an axiom we're
talking about here. You should be able to explain
it. (By it, I mean
my question about Alice.)>>

What question about Alice is unanswered? I have
said in what her participation consists � as well
as the participation of all the other involved.
Will you now show me the mass that you say is
involved?

<<>
>
> > > To which I can only say
> > > "Precisely" -- singularist and pluralist
> > > languages are two different ways of stating
> > the
> > > same facts.
> >
> > Not quite. The pluralist view asserts that
you
> > don't introduce masses.
> > Instead, there's a special "bunch-together"
(or
> > something - it hasn't
> > exactly been elaborated upon) that supposedly
> > handles the questions
> > raised by the removal of "mass".
>
> Well, you haven't introduced any masses yet
> either (aside from assuring me that they are
> there). Back to the students around the
> building. Each student occupies a place wrt
the
> building and other students, roughly (let's
say)
> that if simultaneously each student joined
hands
> with their neighbor on each side the result
would
> be a closed loop and the footpad of the
building
> (and little else?) is entirely inside the loop.
> The way I am reading the claim, I think it
> requires that each student intends to be part
of
> surrounding the building, but there are other
> readings which don't demand that.

"part" implies being part of something to me.
Does it not to you?>>

Well, not really in the sense you mean it. But,
if you insist, then �being a part� here means
�occupying one of the positions on the loop.�

<<> There are
> problably more conditions but this seems to me
to
> be the essential one. The "together" of the
> pluralist is just the fact that this pattern
> requires all the students involved (which is
> trivial) and perhaps that with many fewer
> students similar patterns (that formed closed
> loops around the building) are not possible --
> certainly that no one student can form such a
> pattern. Does the groupiness consist of
anything
> other than this? You've already said it is not
a
> thing over and above the students, so that the

You have the two "over and above"s confused. One
refers to a thing
that something is by nature of being what it is
(a dog is an animal,
since dogs are animals by nature).>>

I don�t use �over and above� in this sense.
Sorry. By me, �over and above� is just an
expression (taken over from acconting, I think)
that means �in addition.�

<<The other refers to things that are
aggeregates of other things. "Forest" being over
and above "trees" is
different from "animal" being over and above
"dog". You can say
"animal" when you have a dog in mind, but you
can't say "forest" (or
"grove") unless you have a forest (or grove) in
mind - which you
usually do, assuming that there are more than
6-10 trees.>>

Which I usually do when? When I am thinking of
forests, I think of � and frequently see �
forests. When I am thinking of trees (as I did
for several summers when I ran surveys for the
Forest Service) I see trees � even in the midst
of a multi-county national forest. Are you
saying, then, that there IS something over and
above the components? Well, again, 1) show it
to me (and presumably as somehow different from
the things together)and 2)(probably in the
process) tell me about its formal properties. I
don�t get the �in mind� notion: I can say
�animal� and have a paradigm picture of an animal
which happens to be a picture of a dog. If I say
�forest�, my paradigm picture can�t be of a
single tree. Because a single tree is not a case
of a forest, though a dog is a case of an animal.
I can, of course, have a pine forest in mind or
even a willow one (with associated problems)
because they are cases of forests. And the point
of this is? I can, of course, say �forest� when
I have trees in mind, and �trees� when I have a
forest in mind. All of which proves (or suggests
or illustrates) what?

<<'s not very often that someone gives you a
specific answer when more
than X things are involved (X being perhaps
10ish). "What's going on
there?" "Some kids are carrying a bunch of chairs
to the garden". But
"some kids" is clearly some sort of special
plural predication, since
you don't mention the words "mass" or "group"!
No, it isn't. The
average human will think of, say, a group of 20
kids massively, and
won't actually summon-to-mind 20 instances of
"kid". "Some kids" in
this case refers to a mass of kids.>>

Well, as usual now, �some kids� is not a
predicate (and predicates aren�t plural). All
that aside, what I take this to be aiming at
saying is that, like averages, plurals are a
convenient talking about a number of conveniently
similar things (kids,say) and noting something
common or collective about them without going
into details. We could, for example, go through
the fifty students surrounding the building and
say exactly what relevant thing each was doing
(standing at angle 67.5 at a distance 5� 3� from
the nearest surface, say) and then summing up by
plotting the points on a map to show that they
amounted to surrounding). We sum up by saying


that they together surround the building.

Similary, we could say exactly (well, as exactly
as needed) what each of the twenty kids is doing.
Instead we sum up (maybe without even counting)
�some kids�, �a bunch of kids� etc. are doing
whatever covers their various activities (enough
for our purpose). It is a special way of talking
(I�d say two special ways, but that is a later
matter) and so like averages in many ways. We
don�t think that there is a person, the average
man, separate the various men whose average was
computed. Why then would we think there is a
bunch of kids separate from the kids being summed
up? It is a figure of speech, as is � in another
way � �together� and the like. For all that,
these figures have their own logic and, in this
case, the logics are the same.

<<> students form a pattern seems to be the most
> obvious next choice. But that, of course,
means
> that for reality, it just says what the
pluralist
> says but in differnt words. If it is something
> else, that you need to say what and demonstrate
> that it really is there. It seems that the
> pluralist says "there are these students and
they
> form this pattern" and the singularist says
> "there is this pattern and the students for it"
> Why this stife there be/'twixt Tweedle-Dum and
> Tweedle-Dee?

My position is that you need an at least implicit
group/mass, so that
you can expand (i.e. explain using more axiomic
terms) things like
"lu'o" or "loi".

You seem to be contrary to this.>>

I am indeed. You can, of course, use masses or
whatever, but you don�t have to. You can do the
same work with just the things and the notions of
distributive and collective predication (which
you need anyhow to deal with {lo} and {lu�a}).
Have you looked at my stuff on the wiki about the
logic of these expressions? I forget the
reference, but the index should bring it up
pretty quickly (though it may be less detailed
than I would now like, it being rather old � back
to the earlier discussion of plural predication).

<<> > > >
> > > > Elaborate? To me, "among" has
implications
> > of
> > > > being "among a group such that".
>
> Well, of course it would; you are a believing
> singularist. For a pluralist, "x is among y"
> just means that x is one of the ys.

x is a referent of ys. Yes. But even for a
pluralist, Alice is also a
referent of/among "the students (wearing hats)".
Again, this doesn't
say anything of the difference between
distributive/non-distributive.>>

Nor was it meant to. That � for a pluralist at
least (and I think for a singularist as well) --
is about predication, not reference. For the
rest, x is not a referent of y, it is, at best, a
referent of whatever expression is used to refer
to y (so �y� properly understood). Alice is a
referent of �the students� (or better, is among
the satisfiers of �the students�) but is not
among �the students,� rather is among the
students (use-mention confusion).

<<>
> > > And so it does -- when used by a
singularist.
> > > When used by a pluralist, it doesn't. But
> > the
> > > properties of "among" are the same for
both.
> >
> > But in the pluralist view, there's still a
> > group there, you just don't
> > choose to acknowledge it, right?
>
> Where? Go through the whole pluralist
semantics
> and nothing like a group turns up, just things,
> one or several as the case may be. At the end
of
> it all, it is hard to say where the
> unacknowledged group might be.

Ok, then use these pluralist semantics to
(usefully!) explain the
relationship between Alice and "surrounds a
building", as opposed to
Alice and "wear hats".>>

Alice is a member of the extension of �wears a
hat,� Alice is among a member of the extension of
�surrounds the building.� The members of an
extension need not be single things but may be
several things at once (plural predication). I
suspect (Hell, I am sure) that this is where your
mind starts (and completes) to boggle: several
things at once and yet not a set/group/mass! To
which the answer is just, �Yup! That�s the way it
goes.� On your side is the fact that those
several things behave formally as if they were in
an L-set. On the pluralist�s side, they get the
same results without having anything other than
the amongers involved. You can call it an
implicit group, if you want to, but the pluralist
is then justified in saying that you are making
up extra things to no purpose whatever.

<<> What you just responded to wasn't so much an
> argument as a challenge.
> Fact is, explanations of how "bunch-together"
> differs from "mass"
> aren't really available. I attribute this to
> there being no
> explanation of "bunch-together" that is
different
> from "mass".>>
>
> I attribute it to the fact that there is no
> difference except verbiage. You seem to think
> that the mass form the explanation is right and
> the the other wrong, which is odd if they are
the
> same explanation. However, this is all empty,
> since we have neither explanation at hand yet
(I
> have tried to supply one but I don't know
whether
> you will buy it).

They're not the same explanation. One says that
there is no mass:

Go through the whole pluralist semantics
and nothing like a group turns up, just things,
one or several as the case may be.

Right? So, no, not the same.>>

It doesn�t say there is no mass. It just says
everything that mass talk does but without ever
using masses. But you, by the way, were the one
that said they were the same (OK, you said they
were not different).


<<> > > > > Ok, then when I say "group of
students",
> I
> > > too
> > > > > am "referring to many things".>>
>
> I agree, but you seem to think that you are
> actually referring to one thing, the group. At
> least you talk that way.

Yes. That, or I'm saying "the students are [part
of a group such that
that group surrounds the building]". Doesn't
matter which, but both
involve "group"/"mass".>>

If you talk that way, of course it does. But you
don�t have to talk that way to describe the
situation pointed to by �the students surrounded
the building.�

<<> > collective predication. Even {loi} does
not
> > appear to be just collective predication --
it
> > seems clearly to involve corporate and Urgoo
> > cases as well. And there are cases which
> cannot
> > be dealt with using gadri.>>
>
> Examples? I see no practical differences
between
> corporate masses and
> regular masses, and I'm not familiar with Urgoo
> cases at all. >>
>
> Corporate masses (I don't much like that
> terminology since it suggests more similarity
> than I think justified)continue to be the same
> even with a change of components; they also
> inherit properties from their components
> directly: if a component (acting as such) does
or
> is something, the corporation does or is, too.

This seems to have more to do with the details of
when a person stops
calling something a mass. I don't think that we
need to categorize
every mass into a certain type of mass in order
to use them.

As for doing something, and having the
corporation do it too, this too
I think has to do with unimportant details. It's
not a fixed rule. To
give an example (that's perhaps more similar than
justified), if a
salesperson makes a sale, the corporation makes a
sale, but if that
salesperson gets the flu, the corporation doesn't
have the flu. It's
not perfecly certain how each thing works out.>>

The salesperson makes a sale as a component of
the corporation, she gets the flu on her own (why
do you think I put the �acting as such� bit in?).
The only points about corporations is that 1)
they have a different logic from masses as most
commnly (I think) understood and yet 2) they are
still thought (at least sometimes) to be
represented by {loi} expressions. I am not
really categorizing masses but rather {loi}
expressions, which are habitually called
mass-expressions. As far as I am concerned,
corporations are not masses at all: that have a
totally different logic.

<<These things are details, and are not critical
to the concept of a mass.>>

Agreed. Except for contrast � what is and is not
a mass � of course.

<<> Corporations also have properties in which
some
> components do not participate. I suppose there
> are other charateristics but these are enough
to
> separate then from ordinary (collective
> predication) masses. Urgoo is the stuff of
which
> some kind of thing is made: all dogs are chunks
> of Dog, for example -- as are dog organs and
the
> mixture that results from a steamroller rolling
> over a pack of dogs. This is an actual
mass-noun
> concept.

Well, if your conception of "dog" extends to
that, then sure. For me,
something stops being a dog when it gets rolled
over - it becomes
"paste formed from a dog corpse". I'd still say
"that dog has been
squished", or "we'll bury the dog", but it would
be in the sense of
{lo pu gerku}, and not {lo ca gerku}.>>

Quite right; it ceases to be A dog and becomes
just Dog (as in �I got Dog all over my steam
roller�). I think that this Urgoo case can be
solved in Lojban as �a mass of bits of dogs� but
I don�t see any consensus on that yet.


<<> So far as I can tell, Urgoo is like
> corporations in some respects: it remains the
> same even if its representations change, it
> inherits properties from its manifestations.
It
> differs in that it is homogenous, does not have
> components, although the manifestations play a
> somewhat similar role, but an Urgoo can exist
> without any manifestations at all.

This is like an ideal mental form of something,
that all things that
are it are composed of: {loi ro gerku}, or
something of the sort.>>

As noted above, {loi spisa be lo gerku} or so.
But clearly not a mental form, since disgusting
concrete and external.

<<> I think these
> two are enough different to justify some
separate
> consideration but both have been folded into
the
> muddle that is CLL mass.

I don't think that the distinction of corporate
vs. non-corporate
entities needs to be made on such a raw level.>>

What raw level; I am just making it within the
category of things expressed by {loi} phrases.


<<>
> "Mass"/"together" expands to "x1 is a mass with
> components x2". This
> is an actual relation. I consider that as
> significant in terms of
> content as you can get.>>
>

> But you offer no evidence that it applies here.
> "Together" is a real situation as well and I
have
> offered an explanation of what it means in
> different terms. What does "is a mass composed
> of" mean in different, neutral, terms. Failing
> that we are just talking by one another, since
we
> are using language radically differently.

The evidence is a sensible explanation of what
"the students surround
the building" means: "the students are part of a
group that surrounds
the building".

Is that wrong? How is it wrong?>>

Well, exactly in the sense part of sensible: I
can see the students but I don�t see something
else, the group, but just the students together.
Show me the group or explain what it means in
neutral terms. (It is not necessarily wrong, by
the way, but it is presented in a way that makes
it look wrong � and may actually be misusing it
to be wrong).

<<How is "the students surround the building"
different from "the group
of students surrounds the building"? Actually
different, and not in
terms of English frames or English pragmatics.>>

What do the two of them mean. I have said what
the first means; does the second mean something
different? If not, then we arre arguing about
nothing (as we are, of course); if yes, then that
way. I am on the one hand, not sure how you
think English frames and pragmatics comes into
and, on the other, what else you would expect to
come into it.

<<>
> <<> They, on the other hand,
> > would find it odd that you cannot understand
> > such a straightforward English expression as
> "the
> > students" (especially since you seem to
> > understand the mysterious "the mass of
> students").
>
> It's about as mysterious as "the building for
> students" - that is, not
> mysterious at all. "the students", on the other
> hand, is ambiguous: it
> can refer as in "the students wore hats" or
"the
> students (as a mass)
> surrounded the building", and then, of course,
> there's also "the
> students (as a bunch-together) surrounded the
> building", which nobody
> has really explained or demonstrated as being
> different from "as a
> mass", though copious flat assertions of the
sort
> have been made.>>
>
> But you, of course, have nowhere demonstrated
> that "as a mass" is different from "together"
> nor explained what it meant. You have asserted

I've explained many times what it meant.

"together the students surround the building" :

X is a mass, and each student is a component part
of that mass

X surrounds the building

the students are part of a mass such that
surrounds the building>>

That just repeats that there is a mass here.
Since no one has pointed to one, you need to show
us where it is or what claiming there is one
comes down to in reality. The existence of a
mass is the crucial one but you just assume it
(question-begging).


<<I don't think that I need to prove that such a
thing as "a
relationship between certain things" exists.>>

Well, of course relationships between certain
things exist, indeed there is at least one
relationship between any two things. But ewhat
does that have to do with the issue at hand,
which is not about relationships but about
masses.

<<> it is superior, but that is just your say-so.
On
> the other hand, if you really believe, as you
> seem to be saying here, that the two
expressions
> mean the same thing, what is the argument all
> about?
>
>
> <<> Note that, if you do write pages explaining
> the
> > differnce, the pluralist can take it, make a
> few
> > uniform changes and provide you with the
> > explanation you want for the difference
between
> > "the students individually" and "the students
> > together."
>
> Please, do it then! Do it with the crude
> paragraphs I've offered. What
> are you arguing this with me for, when simply
> demonstrating this would
> solve everything?>>
>
> Gladly. Please provide the explanation for the
> mass-talk form. Note, this will require saying
> it without assuming masses or giving a fairly
> complete formal system for masses.

A mass is a relationship like any other. Do you
deny that such a
relationship ("x1 is a mass/aggregate/composition
of x2 / x2 is a
component part of x1") exists? Do you deny that
such a thing as a
(predicate) relationship exists?>>

A mass is not a relationship. What you cite is
the relationship between a mass and its
components, which, obviously, presupposes that
there is a mass to begin with. What is it � and
where is it in the situation of studnets around a
building. As for predicate relatioships, that
seems another issue altogeher but I agree that
they exist, for whatever that is worth here.

<<I'm using perfectly established structures, in
English or otherwise,
to do the explaining of how Alice relates.
Predicate relationships,
and the idea of "x1 is a composite of component
parts x2". Both are
established in both our minds, right?


You did say "Gladly". Could you now do this?>>

Sure, as soon as you give the explanation. By
the way, if you have problems with doing that,
you might try just revising slightly the
explanation of �the students together.� The
ideas are clear; the problem is whether these
things have any relevance to the situation at
hand.


<<> > You will no doubt take it that way; how are
you
> > sure the speaker meant it that way or even
that
> > he can sense the difference?
>
> Uh, because "bunch" doesn't have the definition
> that we've assigned it
> (for the sole purposes of this argument) in
> common use. Bunch is
> simply "group", with implications of the things
> being close together -
> "bunch of twigs", etc.>>
>
> Well, it does seem to have that meaning in my
> dialect. That is, when I say "a bunch of
things"
> I am not implying that there is anything other
> than those things there (not even necessarily
> close together).

As I've said, you make it seem like I'm bringing
in the concept of a
baboon to explain away this thing.

You really aren't bringing in anything new,>>

Well, as a pluralist, I am not bringing anything
at all: there are just the students � and the
building, of course.


<<because your mind probably
has "mass of 20 students" 'loaded' (though of
course, the pragmatic
implications of "mass" and all the 'framing' that
"mass" entails are
not loaded), because humans don't usually like to
'load' each of 20
students when they don't have to. "Mass of
students, 20 component
parts" is good enough for most people.>>

I don�t think I have �mass� loaded � or indeed,
except a figure of speech � have it at all.

<<> I presumably have some reason
> for dealing with them together but that is
> nothing "out there" called "bunch," it is just
> how I am dealing with them.

So there isn't anything out there? Or just
nothing out there called
bunch? Because if there isn't anything out there,
I can't imagine you
explaining what Alice's relationship is. With
humor, I imagine
something like:>>

Nothing out there called a bunch.

<<Alice is ? ? ?
? surrounds the building.
where "?" stands for "magic happens here".>>

I have already filled in the question marks for
the pluralist view; what is the similar filling
for the singularists.

<<Well, it's not magic. The "rational
explanation", if you will, is that
Alice is part of a mass/group, the mass/group
that surrounds the
building. If you have a different rational
explanation, then please
offer it.

Calling it rational does not not make it
intelligible or accurate. I have, as noted
several time now, provided the alternative (well,
not an alternative, since you haven�t provided a
the first yet, and also because I suspect yours
will turn out to be about the same).

<<>
>
> <<If they said "the students surrounded the
> rope", then you might have
> an argument as to how it's meant. But if we say
> "the group of students
> surrounded the rope", then it's clear that we
> mean the *group* (of
> students), and not anything else.>>
>
> Not clear at all, since I don't see any group
> there, just students.

When I say "the *group* of students", you can't
imagine a group?>>

I can imagine all sorts of things, but I don�t
perceive one as necessarily there.

<<> If you mean "the group of
> students" to say, in different words, just what
> "the students together" says -- that is,
without

No, "the students together" in your mind for some
reason can't have
the same meaning as it does for me and the
dictionary (1. In or into a
single group, mass, or place), it seems. We don't
say the same things,
because your variant excludes any possibility of
"mass" in order to
describe the relationship (it seems).>>

Well, there are other definitions and some of
them sound quite like what I mean. It does occur
to me, howver, that your �group of students� is
not �students together,� but just �students�
plural. I can say �a group of students are
wearing green hats� (or �a bunch� or �a crowd�,
etc., though not �a mass�). So maybe you, too,
mean only to stress the plurality, rather than
the togetherness, the collectivity. That would
make it harder to align with the pluralistic
collective, but should make an exposition of what
you mean rather easy.


<<"The 50 students surrounded the building" and
> "the group of 50
> students surrounded the building" are
synonymous
> in meaning. It's just
> that one of them uses the word "group", which
> invokes a certain frame
> in your mind that the omission of the word
> wouldn't.>>
>
> Then what the Hell is this argument about? One
> person talks one way, the other the other, as
> their taste leads them. And, of course, that
is
> just what the formalism says: whether you give
a
> pluralist or singularist interpretation to the
> system, the logic is the same.

Ok, then you should have no problem telling me:

In "the students surround the building", Alice is
part of the
mass/group that surrounds the building. Of
course, we don't have the
same imagery invoked in our mind as is typically
associated with
English "mass" or "group", but yes, it's a
mass/group regardless. So
when I say "the students surrounded the
building", I mean that Alice
is part of a mass such that surrounds the
building. Same goes for
Bryce, Carol, David, etc.

Right?>>

Well, if that just means she is one of the
students doing it, then �Yes.� The problem is
that you sometimes talk as though it means
something more and it is that more that I am
trying to get you to explain. If you are saying
there is nothing more, than (aside from the
irritation that the expression �group� or
whatever causes) there is no problem.

<<>
> > Forests are
> > just trees, after all (with some exceptions
like
> > willow forests which are apparently just one
> > tree). (I don't of course, really mean this.
I
> > am just pointing out how useless taking what
> > someone says is in figuring out which of the
> > identical sides they are on.
>
> A forest is not the same thing as a set/"bunch"
> of trees, just as a
> human is not just a set/"bunch" of organs...
just
> as a crowd
> surrounding a building is not just a
set/"bunch"
> of students.>>
>
> And the difference is...? I suppose it is
> something that hold them all together, a common
> interest in them. That is something about us
> usually, although it is often helped by
> propinquity and short-chain causation and the
> like.

The difference between what? A tree and a forest?
A person and a crowd?>>

Between a forest and a bunch of tree, a human and
a bunch of organs, a crowd surroiunding a
building and a bunch of people � those were the
things you just said were not the same.

<<> > More than that too, an organism. That is,
the
> > organs in an organization. Without the
> > organization, the organs are just a pile of
> > specimens.
>
> That's what I mean when I say mass. I discussed
> this earlier using the
> example of a piece of graphite and a piece of
> wood not quite being a
> pencil. Search for the term "graphite" if
you're
> interested.>>
>
> Ah, that was the point of that story. It was
not
> very clear to me at the time. Your use of the
> term "mass" is adding yet another meaning to
that
> already overworked word; can we find another
word
> for you concept.

It's not an overworked concept in the same way
that "animal" is not an
overworked concept, because there are so many
types of animals.>>

I thought you just said that there were not many
types of masses. My point is that, in
Lojbanology, �mass� gets used for a number of
different concepts (�animal� presumably does not,
for, while there are many kinds of animals, they
are all animals under the same difeinition. That
does not appear to be the case with masses (in
the Lojban usage)).

<<Water, which can be combined or reduced into
more water is a mass. I'm
a mass. Just about everything is a mass. A crowd
that makes noise is a
mass, even if some people in it are quiet. Or
even if all people in it
are (relatively) quiet.>>

This makes the term �mass� � already rather
attentuated in Lojban � virtually useless. That
is why I suggested another term for what seemed
to me a minute ago a rther specific meaning. I
gather that that appearance was misleading and
you reallywant something as muddled � indeed,
more so � as Lojban usage. Water is a mass in the
mass-noun sense � continuous, individualized only
by portination, etc. You are a mass presumably
in the sense that you have components that fit
into a organizational scheme. Other things are
masses in other ways (though usually in this
latest one too). A crowd is a mass in the sense
that it has components that are joined together
to do some thing (make a noise, in this case).
And so on (as noted, Lojban mass includes all
these in various ways).

<<If you want to taxonomize and label all of
these different things that
a mass can exhibit, or when certain masses stop
being masses, or if
you can combine certain masses to form a mass
that is considered to be
the same thing as the two masses were, go ahead.
But it's still a
mass.>>

Only because �mass� has become so completely
broad as to make no useful distinctions at all.
Go ahead and use the term if you want, but don�t
be surpised if you get thoroughly misunderstood
as a result. Or give the term a more specific
meaning and thus help to make your point (there
is a point here somewhere, isn�t there?) clearly.

<<> But in any case, I don't see
> how this helps with the students: they do not
> compose an organism or an organic whole, and
> maybe not even an organization.

If 1000 people together do not compose a "crowd",
then what is a
"crowd"? Just a way to refer to the 1000
conceptualizations of people
that you have "loaded" into your mind? Even if
the crowd starts doing
things that none of the people do on their own?>>

Huh? 1000 people together would be a crowd
usually, especially if they are doing something
that needs the lot of them. �Crowd� is also a
way to refer to 1000 people (or more or less) who
are just milling about, perhaps with no common
goals or activities at all (I don�t generally
refer to conceptualizations, just to people �
certainly with the word �crowd� or �people�).
And I don�t see what you use of �loaded� does
here.


<<> They each fall
> into a place in a pattern which we are taking
as
> significant and by virtue of which say they are
> together. Is it also by virtue of this that we
> say they are a mass? If not, what is involved?
> If so, why are we having this argument (or,
more
> accurately, what the Hell are we arguing about,
> since we seem to agree on everything except
what
> words to use and that is merely a matter of
style
> and not open to argumentation).

Well, at first you seemed to deny that the
concept of mass was used in
plural predication, but now you seem to deny that
the concept of mass
(or group) exists at all. So that's what we seem
to be arguing about.>>

I do deny that the concept of mass is used in
pluralistic predication; that is sort of the
definition of that kind of predication. The
whole semantics just does with individuals, no
masses, etc. at all (You can look back at the
couple of expositions I�ve given here or at the
wiki on Bunches). I am not arguing that the
concept of mass (whichever one you want � here I
suppose the one that does what the pluralist says
is done by things together) does not exist. I
merely ask you to explain what it means to say
that it does certain things. I have said what it
means for the students together to surround a
building; what does it mean for a group of
students to surround a building? You were
claiming that only this latter locution is
legitimate (or, at least, that it is the more
accurate locution); my response is to say that,
so afar as you have shown, it is not yet even an
intelligible locution and to suggest that, when
it is made intelligible, it will turn out to be
the same as the mean just the same as the
�together� locution. As you said in a different
context, this is less an argument than a
challenge, although the challenge is within the
frame about whether group talk is intellible and
more correct than the alternative. Occasionally
you say something that sounds like saying that
you recognize tht the two locutions say the same
thing, but then you seem to go back to the
position that only one of them is correct � an
odd combination, so I suppose they are two
separate points.

<<> > Set theory, which seems to be the model for
> talk
> > of masses,
>
> A mass is a relationship, it need not have
> anything to do with set
> theory. x1 is a mass of composite parts x2.>>
>
> Huh!? There is a relationship of composition
that
> defines a mass, but a mass is not a
relationship
> (notice, by the way, that {gunma} is not a mass
> of the sort you descibed earlier).

Is not mass of what sort?>>

I�m not sure; as I noted, your use of �mass� has
become rather diffuse (it seeems to me). The
point is tht {gunma} stands pretty clearly only
for the most specific kinds of {loi} expressions,
what I would characterize as cases of collective
predication, attribution to the whole rather than
to the components separately. You seem to mean
rather more than this (I may be wrong, of course,
the going is rather rough back there).

<<> It may also be
> that the fact that things stand in a certain
> relationship to one another is what gets them
> into the mass, but the mass is not that
> relationship either.

All (?) things (physical things, especially) are
masses. Maybe the
strings of string theory aren't a mass, but
everything else, we've
found it to be a mass. Tiny things, arranged
pencil-wise, form a
pencil. Can everything be broken down into
something else? Is
everything composed of something else? Yes.>>

As noted, this makes the notion of a mass
virtually useless for present purposes:
explaining how a mass of students surrounds a
building. The fact (if it is one) that each
student is also a mass and the building one, too
shed no useful light. If the point of this is to
convince me there are masses, I never have denied
that and so this chat is irrelevant. If the
point is somehow to show how masses are involved
in surrounding a building, it has so far failed,
largely because no effort has been made to get
behind masses to what is going on in mre neutral
terms.

<<The argument isn't really about this (I hope).
It's about whether or
not an entity with parts:students can exist. I
say that it can, and
frequently does. And I also say that this entity
is the thing that
surrounds the building.>>

Now this is just confusing (confused?). No one
denies that entities with parts exist; as you
say, I am one. I take it that you think that
�students� refers to such an entity rather than
to several entities (the individual students, the
putative parts of your entity). I am not even
denying that such an entity can exist; I am
merely asking what it means to say that it
surrounds a building. As a pluralist, I don�t
have to acknowledge that it exists or that it
surrounds a building, since I can account for
students surrounding a building without it. From
that point of view, I can wonder what it would
mean to say that a mass is surrounding a building
and ask you to explain in terms that I
understand. I have explained what �the students
together surrounded the building� means in terms
I assume you can understand, since the
explanation contained no troublesome words like
�together� or �set.� I invite you to do likewise
(or point out what about my explanation you don�t
understand so that I can readjust it to your
apprehension). Then we can examine whether there
is any reason to think that only your locution
involving masses is correct or whether only the
together version is correct or whether they are
equally correectr and maybe even identitical
behind the forms.

<<> Ok, then if it's not connected to the act of
> "surrounding the
> building" by way of a group, then how is it
> connected? What is the
> relation?>>
>
> Directly by each of them taking a place in a
> pattern which constitutes surrounding the
> building. You may call "taking a place"
"forming
> a group" but there is no necessity in doing so.
>

Ok, sure, that's another sensible way to think of
it.

[da poi sruri lo dinju] cu morna [la alis]

x3 can even be "surrounding-the-building-wise".>>

I am not sure Alice is appropriate for {morna2},
she isn�t a part of the pattern, after all, but
occupies a place in that pattern, an x2. And, in
general, what surrounds the building is not the
pattern but things in that pattern.

<<>
> <<> Of course,
> > you can mean that equally well using "the
group
> > of students," but it is harder to see. And,
by
> > parity of reasoning (since the two are
formally
> > identical) "the students" does refer to a
> group,
> > if you want to go that way, although it is
> clearer
> > if you say "the group of students."
> >
>
> What are formally identical? Thinking of them
as
> a group and not
> thinking of them as a group?>>
>
> Well, thinking of them as a group and thinking
of
> them as acting together.
>

Sure. So

[da poi sruri lo dinju] cu gunma [la alis]

is a correct/complete way to express your
pluralist "lo tadni cu sruri
lo dinju", right?>>

Well, no. The original says nothing about Alice
being in the group. As long as you don�t mean
anything ontic about it, I don�t immediately see
anything objectionable about {da poi sruri lo
dinju cu gunma lo tadni}, however; it is
roundabout but apparently equivalent. It is not,
note, a privileged form, more correct than some
other, nor does it constitute an explanation of
the original (except, perhaps, for someone who
heretofore spoke only in metaphysical
periphrasis). But more to the point, that these
two are equivalent is a hypothesis of mine (well,
I know it holds for formal systems, the issue is
whether it holds in normal language). It needs
to be demonstrated and to do that we need to know
what �A group of students surrounded the
building� means in the way that we know what �The
students together surrounded the building� means.
We can then compare. Clearly, if the existence
of an entity, group of students, distinct from
the students is essential to this notion, an
irreducible factor, then they cannot be
equivalent. But that will leave us with the
problem of whether there is in this situation any
such entity. A straightforward count of the
factors involved in fifty students surrounding a
building turns up 51 � the students and the
building, not either 52 (an added group) nor 2 (a
group and a building). What can convince us to
change this count?

Maxim Katcharov

unread,
Jun 11, 2006, 10:11:17 AM6/11/06
to lojba...@lojban.org
On 6/11/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > On 6/10/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > > An outermost quantifier is distributive,
> > > an outermost mass-marker is non-distributive. It's that simple.
> >
> > No doubt it's a simple rule, but the question is if it's the proper
> > way to describe how we communicate.
>
> It's a proper way to describe how I understand Lojban works.
> That's all I can offer you.
>
> > > la .alis .e la betis = ro le re prenu
> > > la .alis joi la betis = lei re prenu
> > > la .alis jo'u la betis = le re prenu
> >
> > Are you offering these definitions as suggestions, or
> > as explanations of how it really is?
>
> As explanations of how I understand them to work in Lojban.

pagbu and gunma, based on the information (from CLL and the gismu
list) that I provided, seemed to be the corresponding ideas. What are
you basing your interpretations on?

>
> > So you would disagree that
> >

> > loi tadni cu sruri lo dinju
> >

In which case you'd no longer be using first order logic. Now, if the


axiomic nature of first order logic is not available to you, then you
should use something else. But you can't just say that your pluralist
interpretation is axiomic itself, because it clearly isn't.

And what's the difference between that and

Alice is one of X
X wear the hats

Clearly, Alice does not relate in the same way to the wearing of hats
as she does to the surroundment of the building, but you seem to gloss
over this *critically* important point. What's the difference between:

Alice is one of X
X surround the rope
(e.g. they're standing around a rope)

and

Alice is one of X
X surround the rope
(e.g. they're playing tug-of-war)

>


> In the singularist approach, X is a single thing, distinct from the
> things that surround the building and consisting of them, while
> in the plural approach X are many things, just those very things
> that surround the building.
>

> mu'o mi'e xorxes
>
>

Maxim Katcharov

unread,
Jun 8, 2006, 10:42:18 AM6/8/06
to lojba...@lojban.org
On 6/8/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 6/8/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
>
> > My take on it is that
> > {cmima} concerns things seen more seperately - a squadron of planes,
> > family of bears, [...]. {gunma} would concern things that look like
> > they're quite close together - a pencil, a book, a car, a body.
> > {pagbu} would be the word to use when you don't care to detail if you
> > see them dispersed or visibly combined.
>
> If your take on {gunma} is that the x2 is distributive, then you cannot
> expand {loi tadni} as {lo gunma be lo tadni}. Consider this case:
>
> The building is surrounded by students and professors together. Then:

>
> loi tadni cu sruri le dinju
> Students surround the building.
>
> is false.

No, true. This is an issue of English pragmatics being brought in to a
language that doesn't need them. {[da poi sruri] cu gunma [lo tadni]}
contains no {po'o}. If you wanted to specify that only students
surround the building, then you'd do just that - and it's not done
here.

>
> lo gunma be lo tadni cu sruri le dinju
> A group that has students as components surrounds the building.
>
> would be true.

Yes

>
> da poi sruri le dinju cu gunma lo tadni
> Something which surrounds the building has students as components.
>
> would also be true.

Yes

>
> So you must either take the x2 of gunma to be non-distributive, or you
> need a different expansion for {loi tadni}.
>
>
> > But it doesn't really matter to me which exact interpretations are
> > given: all of these have the same format - there's one aggregate, and
> > component parts of it.
>
> {pagbu} and {cmima} are both like that, yes. If A and B and C are parts of D,
> then A by itself is a part of D, and B by itself is a part of D, and C by itself
> is a part of D. If A and B and C are members of D, then A by itself is a
> member of D, and B by itself is a member of D, and C by itself is a member
> of D. That's what we mean by saying that the x1 of {pagbu} and the x1 of
> {cmima} are distributive.
>
> But {se gunma] is different: if A and B and C conform D, then A by itself
> does not conform D, B by itself does not conform D and C by itself does
> not conform D. Only together, jointly, do A, B and C conform D.

No, A by itself is a component part of D. My lungs are organs that are
in my body at a certain position. They, by themselves, are a component
part of me. There might not be a D if there wasn't an A and a B, yes,
but that's beside the point, since there /is/ an A, a B, and a C. But
if there wasn't an A and a B, it wouldn't matter that they weren't
there, because the "group" wouldn't be there in the first place. The
relationship wouldn't even be a subject of discussion.

This seems a bit like McKay's "shipmates" line of reasoning - I went
into that subject in a bit more detail in my June 5th reply to aleks.

A generic statement of my objection is that I don't see why this
relationship should get a unique treatment.

>
> {gunma} means "x1 consists of x2".
> {se gunma} means "x1 conform x2".
>

"consists" is "includes"+po'o. {gunma} means "x1 is a
mass/team/aggregate/whole, together composed of components x2,
considered jointly" - mass, team, aggregate, whole, together,
composed, components, jointly - all of these are gloss words for the
relationship. Saying that it means a mere "consists of" is... strange.

"conform" means to become or be similar, but I assume that you mean
the "together form" definition. Yes, they together form x2 - in the
sense of "they are something together: a group / mass / aggregate /
whole / compound / form"

>
> > I'd disagree with that. It's a "mass/team/aggregate/whole, together
> > composed of components...". "Consists" is a special term, it has very
> > specific pragmatics attached to it.
>
> {gunma} does mean "consists of". But even if you think {gunma} means
> something else, the predicate that you need to make your expansion
> of {loi} is one that means "x1 consists of x2", not one that means
> "x1 has component x2"
>

Ah. No, I don't need it to make my expansion.

Jorge Llambías

unread,
Jun 11, 2006, 11:16:50 AM6/11/06
to lojba...@lojban.org
On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> The surest way to show that I'm a fool for asking this 30th time
> is to point me to an explanation that I haven't rightly shown to
> be unexplanatory.

I don't think you are a fool.

Explanations are pointless at this point, because definitions cannot
be right or wrong. We are now working with different definitions.
At this point your {loi tadni cu sruri le dinju} and my {loi tadni cu
sruri le dinju} are applicable in different situations. For example, in
a situation where students are on one side of the building and
professors are on the other side, in such a way that students
and professors surround the building together, you can say
that {loi tadni cu sruri le dinju} = "a group that includes students
surrounds the building", and I can't say {loi tadni cu sruri le dinju}
= "students surround the building".

For me {loi tadni} means "students", just like {lo tadni}, and
the mass gadri in addition indicates that whatever is predicated
of the students is predicated non-distributively. For you it means
"a group of things that includes students among its members",
which is something quite different.

At the beginning of the discussion, I thought we both understood
what {loi tadni cu sruri le dinju} meant (more or less what it
has always meant in Lojban) and we were comparing
different ways of analyzing the sentence to get to that meaning. Now
it appears that we don't even understand the sentence to
mean the same thing. Comparing two ways of analyzing it as if
we thought it meant the same thing for both is pointless.

Maxim Katcharov

unread,
Jun 8, 2006, 7:12:47 PM6/8/06
to lojba...@lojban.org
On 6/8/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 6/8/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > On 6/8/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> >
> > > The building is surrounded by students and professors together. Then:
> > >
> > > loi tadni cu sruri le dinju
> > > Students surround the building.
> > >
> > > is false.
> >
> > No, true.
>
> Are you sure? Is "Alice surrounds the building" true? Is "Alice and Betty
> surround the building" true? Suppose Alice and Betty are the only students

> in the group, and the rest of the people surrounding the building are
> professors,
> is "Students surround the building" true?

Yes, it becomes much more apparent (to me, at least) once written out
in Lojban. It's not {la alis cu sruri lo dinju}, it's {lu'o la alis cu

sruri dinju}. As in other aspects of Lojban, saying one thing does not
preclude another. Usually, the speaker would say "Alice and the
professors surrounded" in the first place, so this wouldn't be so
strange. However, if asked "did Alice surround?", or better put "was
Alice a surround-mate?" (which is a more exact, though unorthodox
translation), one would say yes. The strict

"Alice and Betty surround the building" would be
{[la ales e la betis] sruri lo dinju}

which, yes, in this case is false.

>
> > > {gunma} does mean "consists of". But even if you think {gunma} means
> > > something else, the predicate that you need to make your expansion
> > > of {loi} is one that means "x1 consists of x2", not one that means
> > > "x1 has component x2"
> >
> > Ah. No, I don't need it to make my expansion.
>

> I think you do.


>
> mu'o mi'e xorxes
>
>

Maxim Katcharov

unread,
Jun 11, 2006, 12:30:46 PM6/11/06
to lojba...@lojban.org
On 6/11/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > The surest way to show that I'm a fool for asking this 30th time
> > is to point me to an explanation that I haven't rightly shown to
> > be unexplanatory.
>
> I don't think you are a fool.
>
> Explanations are pointless at this point, because definitions cannot
> be right or wrong. We are now working with different definitions.

I don't think that this is the issue. Definitions define concepts, or
explain how words are shortcuts for saying much longer things. If we
were talking about the same thing in two different ways, then yes, but
we aren't. And neither of us is saying that each other's definition of
their own concept is wrong - in fact we usually mark things as "your
X" or "my X".

But we're not talking about definitions. I'm asking you to explain or
define or (usefully) exemplify your *concept* itself. I know that
certain words refer to your concept, but I want to know how they are
explained using axioms - things that neither of us have to prove,
things that nobody should really argue against.

For example, both of us understand and agree with a 1to1 relationship:

Alice is inside the school

and both of us understand and agree with the basic plurally
predicative relationship

the 26 students are inside the school >>
Alice is inside the school
Bryce is inside the school
[...]
Zoe is inside the school

and both of us understand and agree with the "mass" relationship

the graphite and the wood are component parts of the pencil >>
the graphite is a component part of the pencil
the wood is a component part of the pencil

and both of us understand and agree with the use of 'variables'

the stones are inside X
X is on the table

and we see how these can be combined

the graphite is a component part of the pencil
the pencil is on the table

the graphite is a component part of X
the X is on the table

and so the explanation of my position is

Alice is a component part of X
X surrounds the building

(where X could be~ "the surroundment of the building", or "surrounder...")

Now, this doesn't prove that Alice actually *is* a component part, or
that X *does* surround the building, but it shows that if we were to
see it this way, then it would be perfectly workable.

I don't see your position as equally sensible. If I were to say "ok,
there's no mass involved", I would have nothing like this to rely on.
I would have

Alice is a referent of X
X surround the building < axiomic explanation of this is needed

but I would be taking this in on faith - it seems that this is
correct, and that there is no mass, so hey, why not?

> At this point your {loi tadni cu sruri le dinju} and my {loi tadni cu
> sruri le dinju} are applicable in different situations. For example, in
> a situation where students are on one side of the building and
> professors are on the other side, in such a way that students
> and professors surround the building together, you can say
> that {loi tadni cu sruri le dinju} = "a group that includes students
> surrounds the building", and I can't say {loi tadni cu sruri le dinju}
> = "students surround the building".

Sure. I'd be perfectly happy to say that

loi tadni cu sruri lo dinju

expands to

[da poi sruri lo dinju] cu gunma [[lo tadni] po'o]

for the purposes of this discussion, since really, the discussion
isn't much affected by it.

However, your position is that it *doesn't* expand using gunma, since
there is supposedly no concept of a mass invoked in one's mind when
one says "the students surround the building". If there's no concept
invoked, then it doesn't expand in that way. In the same way, {lo
gerku} doesn't expand to {lo danlu} (though it's an acceptable way to
see it) - though it may very well expand to something like {lo danlu
be la dog} (or what have you). "That's" would not expand to "that is"
if the concepts suggested by "that is" were not invoked. But if they
were, it would, even if the 'method of invocation' was different.

You treat this expansion like {lo gerku}>{lo danlu} - "you can see it
that way, but that's not quite accurate". I treat it as "that's">"that
is" - different words, but the very same concepts are used.
Specifically, the concept of "mass/parts".

>
> For me {loi tadni} means "students", just like {lo tadni}, and
> the mass gadri in addition indicates that whatever is predicated
> of the students is predicated non-distributively. For you it means
> "a group of things that includes students among its members",
> which is something quite different.

If you'd like, it can be "an entity composed of students", it doesn't
really matter. And yes, this would be quite different, because it
treats the students together as a different entity than each of the
students themselves.

Jorge Llambías

unread,
Jun 3, 2006, 4:50:59 PM6/3/06
to lojba...@lojban.org
On 6/3/06, John E Clifford <cliff...@sbcglobal.net> wrote:
> As you may recall, my suggestion mirrors the
> English for "individually" and "collectively" (or
> "together"), attaching as convenient to sumti or
> predicate place (so, I suppose that UI is about
> the only selmaho that will work -- unless we
> invent a new one).

In English those adverbs normally indicate how the predicate
applies to the subject. For example:

The men carried the pianos together.

would normally mean that each piano was carried by all the men,
not that each man carried all the pianos at once. Is that what you
have in mind, that when the predicate is tagged, it indicates how
it applies to the x1?

> The point of using {lu'o} and
> the like is that they would have no use in the
> mildly revised system and so could be used for
> something else -- in this case something related,
> even.

There has been resistance to changing utterly useless words, so
I don't expect a proposal to change something not totally useless
like {lu'o} to succeed. I wouldn't especially oppose it, but I know
others would.

> I think this would have a minimal effect
> on old text, since 1) {lu'o} and the like have
> scarcely ever been used and 2) the old forms are
> legitmate under the new dispensation and have the
> same meaning (or what now corresponds to the old
> meaning). These last two claims are
> impressionistic, so correct them if they are
> wrong.

{lu'o} has seen some use, I don't know about "scarcely", it is hard
to quantify. Some cmavo I would be pretty sure have seen no use,
but I don't think {lu'o} is one of them. If it's moved to UI, then the
old forms would change rather drastically because UIs attach to
the preceding word and {lu'o} applies to the sumti that follows. That
could be partly solved by putting it in BAhE instead of in UI, but the
scope would still be wrong, because {lu'o} can take any number of
connected sumti, whereas BAhE only takes the immediately
following word.

Maxim Katcharov

unread,
Jun 7, 2006, 6:17:32 PM6/7/06
to loj...@yahoogroups.com
On 6/7/06, Adam D. Lopresto <ad...@pubcrawler.org> wrote:
> On Tue, 6 Jun 2006, Maxim Katcharov wrote:
>
> > "Mass"/"together" expands to "x1 is a mass with components x2". This
> > is an actual relation. I consider that as significant in terms of
> > content as you can get.
>
> That's pretty much exactly the definition of {gunma}. "x1 is a

Yep, that's where I got the definition from.

> mass/team/aggregate/whole, together composed of componets x2, considered
> jointly." It seems pretty clearly that x1 is a mass, and that x2 is a plural
> reference.

No, it isn't a plural reference. It's a quite strictly a singular
reference when it's the expanded form. Think of {nenri}, "x1 is within
x2". Just as there's one container there (there could be two+, but we
assume a {pa}) that each of many things can be within (it is true that
each is within it), there's a single mass that things are components
of (it is true that *each* is a component).

>
> .i lei tadni cu gunma le tadni
> The mass of the students is a mass formed from the students.
>
> But if you don't permit plural reference, then what can possibly fill the x2
> there? It certainly can't be each student considered seperately (ro le
> tadni). And it can't be the mass of the students, since that's exactly
> what's in the x1. So what is it?

It is each student considered seperately. It is true that Alice is a
component part of, say, a crowd, or what-have-you. Same for each of
Bryce, Carol, Daved, Eliza, [...].

>
> And in case you're tempted to say it's actually the set, consider the
> definitely of mei

I'm not tempted. On this subject, I don't really like the way that
mathematical sets are given so much priority. I have a much looser
definition of 'set': "the three bears" pops three bears into my head,
and that's a set of 3 bears, with no commitment at all to "group" or
"mass" or whatever.

Maxim Katcharov

unread,
Jun 11, 2006, 7:47:57 AM6/11/06
to lojba...@lojban.org
On 6/10/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:

> On 6/10/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > > >
> > > > lu'o ro lo tadni cu sruri lo dinju
> >
> > If your grammar says that an outer {ro} on a {lo} marks it for
> > distributivity, why is (1) exempt from this rule?
>
> Because the presence of {lu'o} means {ro} is no longer the outermost
> marker. The distributive/non-distributive marker marks a slot, and no
> matter how many {lo}s and {loi}s and {lu'o}s and quantifiers are contained
> inside the sumti expression, the one that determines whether the slot is
> distributive or not is the outermost. An outermost quantifier is distributive,

> an outermost mass-marker is non-distributive. It's that simple.

No doubt it's a simple rule, but the question is if it's the proper
way to describe how we communicate.

>
> In addition to quantifiers, logical connectives are also distributive and
> work just like quantifiers, and {joi} works like {loi} and {lu'o}. The neutral
> connective, not marked for distributivity and which corresponds to {lo}
> is {jo'u}. So if we have two people, Alice and Betty:


>
> la .alis .e la betis = ro le re prenu

> la .alis .a la betis = su'o le re prenu
> la .alis .onai la betis = pa le re prenu
> la .alis na.enai la betis = no le re prenu
> la .alis na.anai la betis = su'epa le re prenu
> la .alis .o la betis = ro ja no le re prenu
>
> (The remaining logical connectives are not symmetric, and therefore don't
> have a corresponding quantifier.)


>
> la .alis joi la betis = lei re prenu

joi: in a mass with non-logical connective: mixed conjunction; "and"
meaning "mixed together", forming a mass

pagbu: x1 is a part/component/piece/portion/segment of x2 [where x2 is
a whole/mass]; x2 is partly x1

> la .alis jo'u la betis = le re prenu

jo'u: in common with; from CLL: A and B considered jointly

gunma: x1 is a mass/team/aggregate/whole, together composed of
components x2, considered jointly


Are you offering these definitions as suggestions, or as explanations
of how it really is?

>


> > > {lo rokci joi la alis cu sruri le dinju}.
> > >
> > > But apparently under your current interpretation, from that it follows that
> > > {loi rokci cu sruri le dinju} and also that {lu'o la alis cu sruri le dinju}
> > > (= {lai alis cu sruri le dinju}?). But neither of those follow at all, the
> > > way I understand it.
> >
> > As I've said, this isn't English, you don't need the same pragmatics
> > and verbatim translations.
>

> I'm not talking about pragmatics here, I'm talking about what follows
> logically from an expression. For all cases of ko'a and ko'e and broda,
> independently of their meanings, under your interpretation you have
> that from:
>
> (1) ko'a joi ko'e broda
>
> you can deduce:
>
> (2) lu'o ko'a broda

Right

>
> Just from knowing that ko'a and ko'e do X together, you can deduce
> that ko'a is part of a group that does X. There's no pragmatics
> involved there.

Right, there are no pragmatics involved in my interpretation.

> But for me (and also the way Lojban has always
> been, as far as I can tell) {lu'o ko'a} is not "some group that has the
> referents of {ko'a} as components, possibly among other components",
> it means, in singularist terms, "a group that consists of the referents of
> {ko'a}, no more and no less" or in pluralist terms it means that the
> referents of {ko'a} do something together.

"Consists" is handled by {po'o}. Aside from pragmatics ("well, the
speaker wouldn't say 'the students surrounded the building' if it was
students /and/ professors, so it must be just students"), there is no
reason to assume that {lu'o la alis cu sruri lo dinju} would be false,
or rather, to assign that po'o.

>
> > Now, I could say {lu'o la alis cu sruri lo dinju}, but usually I
> > wouldn't. I'd say {lo rokci joi la alis cu sruri lo dinju}. However,
> > the former would still be true - Alice is a part of the
> > surrounder/surroundment of the building.
>

> Under your reinterpretation of {lu'o}, that's correct. Under the usual
> interpretation, that's not correct. This is independent of whether
> you take the singularist or the pluralist road. The singularist and
> pluralist roads take you both to the same final place, but this new
> spin that you want to put on {lu'o} changes it to something else.

So you would disagree that

loi tadni cu sruri lo dinju

expands to

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

which you had previously described as a legitimate interpretation,

yes? I agree, this is somewhat aside from the discussion. If it were
true, then the expansion would be

[da poi sruri lo dinju] cu gunma [[lo tadni] po'o]

which would be sensible, and if it was the case, {lu'o la alis cu
sruri lo dinju} would be false.

However, I can't see any reason to introduce po'o. (Lack of a good
reason should be enough to default to the simple version.) It doesn't
harm anything to omit po'o - you could always use the expanded form,
or a different shortcut. However, in other less simple cases, the full
enumeration of components may prove infeasible (is a sports team
really just the players that are on it? Is a human just those
molecules? etc.), so the introduction of po'o is harmful there.

>
> > I'd still like to have that explanation of distributivity that I've
> > been asking for.
> >
> > 1) {lu'o la tadni cu sruri lo dinju}
> > 2) {loi tadni cu sruri lo dinju}
> > 3) {la tadni cu sruri lo dinju}
> > 4) {la tadni cu dasni lo mapku}
> >
> > What is Alice's relationship to each relationship?
>

> (I assume you mean {le} or {lo} rather than {la}.)
>
> In all cases, Alice is one of the referents of the sumti that appears in x1.
> There's no more to it than that.

There is more to it. {loi tadni cu sruri lo skori} and {ro lo tadni cu
sruri lo skori} are different in some way, yes? I'm asking you to
explain the difference. Does saying "Alice is one of the referents of
both lo tadni" contribute /at all/ to an explanation of the
/difference/? No. So there is more to explain - more to it.

> The answer to your question "If Alice and
> Betty do something together, what am I saying that Alice is doing by
> herself?" is "I'm not saying anything about Alice more than that she is
> one of the two people doing something together."
>

If I didn't know any English, and asked you to explain "dog" to me,
would you do it in a way that relies on the English words "hound" or
"canine"? No, you wouldn't.

The meaning of the word "together" is the thing under dispute. If
you're going to offer an explanation, don't make it circular - I have
no idea what you mean by "together", the meaning of your "together" is
what you're trying to explain - so why would you use your definition
of the word? The only way that I understand "together" is "as parts of
a group" (or "reciprocally"/"simultaneously", which are not
applicable).

Here is, again, my explanation for as a mass and individually, respectively:

Alice is part of X
X surrounds the building

Alice wears a hat

Is the difference not apparent there?

> mu'o mi'e xorxes
>
>

Maxim Katcharov

unread,
Jun 11, 2006, 9:47:25 AM6/11/06
to lojba...@lojban.org

lo tadni cu dasni
lo tadni cu sruri

Assuming that Alice is one of those tadni, she "participates" in both
of those. So explaining it in that manner doesn't explain anything,
unless you explain what "participates" means. I don't see how the
question is incompletely specified. I don't see how participation is


the "flip side" of doing things together.

>
> <<>

> Huh? Which two � you offer three? Do please

which one of [a, b] [c] expands to.

> decide whether there is something besides the
> students involved here. If there is an ontic
> group, then one line of chat is appropriate; if
> there is only the students considered in a
> certain way or some such locution, then another
> is. In short, please finally give an explanation
> of what "a group of students" means in real terms
> (i.e., without falling back on the "group, mass,

> set,�" idioms, which are question-begging).

Think of a student in your mind. 1 concept. Now think of a mass of
kids that are putting away a mass of chairs. 2 concepts. A group is a
single thing - it's composed of many things, but seen together. Think
of a sports team. Do you think of each player individually? No. When
you look at a forest, do you think of every tree individually? No.
When you look at a crowd of people of sufficient size to surround a
building, do you think of every person individually? No. You think of
one thing, incidentally composed of players/trees/people.

>
> <<Perhaps you could make this disinction between
> "distributivity" and
> "non-distributivity" in a way that (usefully!)
> explains what relation
> Alice has to the surroundment of the building?
> The way I see it,
>
> [da poi sruri lo dinju] cu gunma [la alis]
> "alice is part of that which surrounds the
> building" or
> "alice is part of the surroundment of the
> building"
>
> I think that that's perfectly reasonable.>>
>
> Depending on what it means, it is or is not, and
> the latter depending upon what theory you have of
> bunches. It MAY mean just what "Alice
> participates in surrounding the building" (or

Useless, you'd have to explain how she participates, and that brings
us back full circle.

> "Alice is among the surrounders of the building")

Useless, she's among the wearers of hats too.

> means or it may say something about an entity (a
> mass, apparently)and that entity may be either a
> bunch in the technical sense or some other sort
> of entity. If it is a bunch in the technical
> sense, then again it means the same as "Alice
> particpates in surrounding the building" or
> "Alice is one of the surrounders of the
> building." If not, then it is just plain unclear
> what it means and so how it is related to the
> pluralist version. Which is it?

Which is what? My theory of technical bunches is that they're always
predicated individually. And a "mass" is 1 entity, such that has
component part "Alice" (and another component part "Bryce" [...]).

Yes, none by your view. When you think about it, and force yourself to
abandon the idea of a "mass", Avery ends up floating somewhere out
there with no real relationship to what you were predicating about
ultimately him.

> Or is Avery the one directing

> � in which case he is, apparently, part of the


> mass that lifts it. Since you have yet to explain
> what a mass is in these cases, I am unsure what
> your intended answer is here. Alice's role in

Mass: one thing, with component part: Avery (including but not limited to).

> participation was just given, was Avery's? Your
> masses are beginning to sound a lot like the
> oldest definition of sets: "things considered
> together" (though, of course, you can't say
> that).

Sets are different. A set doesn't lift a piano.

>
> <<There is no real explanation>>
>

> Well, not from you anyhow � repetition is not
> explanation.

Repetition of explanation is. Every time I explain something, I find
that you require an additional explanation for seemingly obvious
things, things that you had no trouble grasping before. Why are you
asking me what a "mass" is? Or what a "group" is? It's been clear for
no doubt the last 30 messages, and I've elaborated upon it many times.
I've (repeatedly) expanded

lu'o le tadni cu sruri lo dinju

to

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

and I'm sure that you know what something like /that/ means.

A is part of entity X
X surrounds the building

It's a lot like

A is nenri X
X is on the ground

in terms of how the predication should be working.

>
> <<- none that I can think of, and none that
> you've provided (correct me if I'm wrong). Yes, I
> can see how,
> intuitively, one may think of it that way, but
> there are a lot of
> things that we sense intuitively that are wrong.
> In order to
> understand it, I need an explanation. This isn't
> an axiom we're
> talking about here. You should be able to explain
> it. (By it, I mean
> my question about Alice.)>>
>
> What question about Alice is unanswered? I have

> said in what her participation consists � as well


> as the participation of all the other involved.
> Will you now show me the mass that you say is
> involved?

Where? Can you repeat it here, or point me to where you've done this?

No, that's not what I insist on. Alice could stand on position X, then
walk away, and Bryce could stand on position Y, and then walk away,
etc., until we have 26 students who have stood on the requisite
positions. Will they have surrounded the building? No.

I'm curious as to the definition of "part" that you use. Perhaps you
could describe it? It would be helpful if also you gave the
relationship involved (e.g. "killer" is part of the relationship
"[killer] murders [victim]").

"...which one usually does..."

> When I am thinking of

> forests, I think of � and frequently see �


> forests. When I am thinking of trees (as I did
> for several summers when I ran surveys for the

> Forest Service) I see trees � even in the midst


> of a multi-county national forest. Are you
> saying, then, that there IS something over and
> above the components? Well, again, 1) show it
> to me (and presumably as somehow different from
> the things together)

"Over and above" meaning "in addition to"? Yes, but it's not something
as strange as you seem to think.

It is the things "together". It is your assumption that "things,
together" is somehow different from "mass", not mine.

> and 2)(probably in the
> process) tell me about its formal properties.

Its formal properties are whatever properties it exhibits. What formal
properties does a fork have? How about a person? A squadron of fighter
planes? The solar system? A sports team? A rioting mob? It's not like
there's some formula to this. Each of those is a composite entity,
with certain components.

> I don't get the "in mind" notion: I can say
> "animal" and have a paradigm picture of an animal
> which happens to be a picture of a dog. If I say
> "forest", my paradigm picture can't be of a
> single tree. Because a single tree is not a case
> of a forest, though a dog is a case of an animal.
> I can, of course, have a pine forest in mind or
> even a willow one (with associated problems)
> because they are cases of forests. And the point
> of this is? I can, of course, say "forest" when
> I have trees in mind, and "trees" when I have a

No, you can only say forest when you have trees in mind that are
component parts of the same forest. Just having (somehow) 1000 trees
in various locations in mind, which very clearly satisfies "trees",
will not usually satisfy "forest". "Trees" is a bunch, and you can't
help but treat each distributively; forest is mass, and you can't help
but treat it distributively too, it's just that it's composed of each
of that bunch of trees.

The points themselves don't add up to surrounding, and even doing it
at the same time doesn't add up to surrounding (or "reading"). It has
to be, aside from all that stuff, seen as an entity.

> We sum up by saying
> that they together surround the building.
> Similary, we could say exactly (well, as exactly
> as needed) what each of the twenty kids is doing.
> Instead we sum up (maybe without even counting)
> "some kids", "a bunch of kids" etc. are doing
> whatever covers their various activities (enough
> for our purpose). It is a special way of talking
> (I'd say two special ways, but that is a later
> matter) and so like averages in many ways. We
> don't think that there is a person, the average
> man, separate the various men whose average was
> computed.

...obviously. The entity isn't some magical and contrived "man". The
entity is akin to "crowd". Look at a crowd of people, what do you see?
Each one of the people? No, you probably see a "crowd".

> Why then would we think there is a
> bunch of kids separate from the kids being summed

> up? It is a figure of speech, as is � in another
> way � "together" and the like. For all that,

How do you need them anyway? Whatever collective predication is, if
not related to "mass", is not needed for my explanation of how Alice
fits into "the students surround the building" vs. "the students wear
hats".

Alice is component part of X
X surrounds the building

vs.

Alice wears a hat

> Have you looked at my stuff on the wiki about the
> logic of these expressions? I forget the

I haven't. Could you link me?

> reference, but the index should bring it up
> pretty quickly (though it may be less detailed

> than I would now like, it being rather old � back


> to the earlier discussion of plural predication).
>
> <<> > > >
> > > > > Elaborate? To me, "among" has
> implications
> > > of
> > > > > being "among a group such that".
> >
> > Well, of course it would; you are a believing
> > singularist. For a pluralist, "x is among y"
> > just means that x is one of the ys.
>
> x is a referent of ys. Yes. But even for a
> pluralist, Alice is also a
> referent of/among "the students (wearing hats)".
> Again, this doesn't
> say anything of the difference between
> distributive/non-distributive.>>
>

> Nor was it meant to. That � for a pluralist at


> least (and I think for a singularist as well) --
> is about predication, not reference. For the
> rest, x is not a referent of y, it is, at best, a
> referent of whatever expression is used to refer
> to y (so "y" properly understood). Alice is a
> referent of "the students" (or better, is among
> the satisfiers of "the students") but is not
> among "the students," rather is among the
> students (use-mention confusion).

Alright, then what is this "among" relationship that you speak of?

This is not an axiom. It's not apparent to everyone. So no, you can't
just say "Yup! That's the way it goes". You have to be able to explain
it, and explain how it affects relationships.

> On your side is the fact that those
> several things behave formally as if they were in
> an L-set. On the pluralist's side, they get the
> same results without having anything other than
> the amongers involved. You can call it an
> implicit group, if you want to, but the pluralist
> is then justified in saying that you are making
> up extra things to no purpose whatever.

The problem with this is that these "extra things" are requisite. You
cannot explain how Alice relates to the surroundment in any other way
than

Alice is part of X

X surrounds the building

or at least none that you've demonstrated.

But that's so first-order! Yes, I guess it is. This sort of
predication is both simple and fundamental, and both of us know that
it works.

Now, we come upon a situation that is interesting. We have a phrase
like "the students surround the building". To explain this phrase, we
part ways:

I state that "the students" actually refers to "the students seen as a
mass" - that is, one entity, of which each student is a part. Your
objection to this is that we're introducing an extra entity. My
response is that the entity is already introduced, and regardless,
what's improper about doing that? This isn't like I'm introducing a
gorilla - I need no further information, aside from the students. And
this isn't, as you may assume, "well, I don't know what's really going
on, so I'll say that it means 'is a component part of y' because it's
the only thing that seems to make sense". No. This idea that many
students form something isn't alien. Words like "crowd", "mob", etc.
are all indicative that people can and consistently do see an entity
"beyond" just each of 100 people. Now, I still recognize the students,
in the same way I recognize "the table" in "the stones such that are
on the table", but the actual predication concerns what the students
are together - or viewed another way, what group the students are part
of (the one such that surrounds the building).

You state that "the students" is predicated as
"bunch-together"/non-distributively/etc. That is, it's not predicated
as a mass (though for some reason that's still a valid
interpretation), and each isn't predicated individually. My objection
is that you should be able to explain what the relationship is: I
explained it, in English, using established concepts (distributive
predication, the idea of "x1 is composed of entities x2"), and you
should be able to do the same to explain "bunch-together".

My position may seem bulky (though I assert that this bulkyness is due
to the framing associated with "mass"), but your position is
inexplicable. It's an argument from zen, if you will: "this is a
higher level of thinking about it. I don't need to give an explanation
of this. It just 'is', and that is all, what is not obvious?". I find
this both frustrating and absurd. Every explanation given so far has
had either one of the following qualities:

1) they make no distinction between pluralism or singularism. For
example, saying that Alice is a referent of "the students" in "the
students surround the building" is like saying that the difference
between a dog and a cat is that a dog is an animal. Or perhaps
defining a dog as "non-cat". Alice just as much is a referent of "the
students" in "the students wore hats", and she 'participates' in both
relationships (the word 'participates' needs to be explained). In
short, explanations are just other names. What's a dog? A hound.
What's a hound? A canine. What's a canine? A dog.

2) they are highly indicative of the group relationship that I support.

I'm actually not too certain that I've encountered (2).

(They are the same. "So, no, not the same [by your standards]" )

You don't mention the word, but the concept is certainly used. It's as
if to say that "is" is not mentioned in "that's", so aha!, we can say
everything that "that is" says without mentioning "is", since we have
this surely completely different concept of "that's". Sure, you may
argue that "that's" is quite different. It is, after all: you could
see it, if you wanted to, as another way to express "that is" - this
is the very reason that some people fail to properly use an
apostrophe. However, the concept invoked is the same.

Neither of us can prove that our positions are correct, because no
complete and tried theory of how the mind works is available. The best
we can do is point out flaws in each other's explanations. You seem to
agree with mine, but I disagree with yours. My disagreement stems from
how no explanation not fitting (1) or (2) has been provided by you.
When asked for explanations, you simply assert "that's how it is", as
if this were simple predication, or the existence of a relationship
called "gives", or "a +b = b + a", which are things that we could all
agree on -- this of course seems to be after I've indicated that your
several given explanations were unhelpful, since they fit (1) or (2).

>
>
> <<> > > > > Ok, then when I say "group of
> students",
> > I
> > > > too
> > > > > > am "referring to many things".>>
> >
> > I agree, but you seem to think that you are
> > actually referring to one thing, the group. At
> > least you talk that way.
>
> Yes. That, or I'm saying "the students are [part
> of a group such that
> that group surrounds the building]". Doesn't
> matter which, but both
> involve "group"/"mass".>>
>
> If you talk that way, of course it does. But you
> don't have to talk that way to describe the
> situation pointed to by "the students surrounded
> the building."

Sure, you don't have to use the words "mass" or "group", and all the
bulk attached to those two. But you still use those concepts. People
are capable of and do use them constantly, and they're a perfectly
reasonable expansion (as in "that is" from "that's").

The very definition of "acting as such" is based on if it affects the
corporation or not. If she represented the corporation fully, and,
say, got the flu while talking to a client - but no, this doesn't
matter, because she's only "acting as such" if the entire corporation
is seen to have done it with her.

> The only points about corporations is that 1)
> they have a different logic from masses as most
> commnly (I think) understood and yet 2) they are
> still thought (at least sometimes) to be
> represented by {loi} expressions. I am not
> really categorizing masses but rather {loi}
> expressions, which are habitually called
> mass-expressions. As far as I am concerned,
> corporations are not masses at all: that have a
> totally different logic.

What sort of different logic?

>
>
> <<These things are details, and are not critical
> to the concept of a mass.>>
>

> Agreed. Except for contrast � what is and is not
> a mass � of course.

No, I wouldn't call it Dog. I would see no "dog" in a mass of bloody
mush, though I would, if the damage was not too great, or if I knew
that the mush was once a dog, see "pu dog", {lo pu gerku} - or even
"that which was Dog", {lu'o pu lo ro gerku} (?) (that which in the
past was a part of the mass of all dogness).

> I think that this Urgoo case can be
> solved in Lojban as "a mass of bits of dogs" but
> I don't see any consensus on that yet.

Well, not bits of dogs. But "mass of all dogs" seems right.

I'm saying that if you want to say "dog", you can say "animal"
instead. You don't need to say what kind of animal it is each time.
This assumes that a corporate mass is really just a mass.

>
>
> <<>
> > "Mass"/"together" expands to "x1 is a mass with
> > components x2". This
> > is an actual relation. I consider that as
> > significant in terms of
> > content as you can get.>>
> >
> > But you offer no evidence that it applies here.
> > "Together" is a real situation as well and I
> have
> > offered an explanation of what it means in
> > different terms. What does "is a mass composed
> > of" mean in different, neutral, terms. Failing
> > that we are just talking by one another, since
> we
> > are using language radically differently.
>
> The evidence is a sensible explanation of what
> "the students surround
> the building" means: "the students are part of a
> group that surrounds
> the building".
>
> Is that wrong? How is it wrong?>>
>
> Well, exactly in the sense part of sensible: I
> can see the students but I don't see something
> else, the group, but just the students together.

"The students together" seems to be a very pure/unburdened form of
group/mass/etc.

> Show me the group or explain what it means in
> neutral terms. (It is not necessarily wrong, by
> the way, but it is presented in a way that makes

> it look wrong � and may actually be misusing it
> to be wrong).

When someone says "the trees are burning", you don't think that each
tree is burning. You simply think of a bunch of trees, and you treat
them collectively, as one entity. The entity is burning.

>
> <<How is "the students surround the building"
> different from "the group
> of students surrounds the building"? Actually
> different, and not in
> terms of English frames or English pragmatics.>>
>
> What do the two of them mean. I have said what
> the first means; does the second mean something
> different? If not, then we arre arguing about
> nothing (as we are, of course); if yes, then that
> way.

If they're the same, then though you may not have the same concepts
come up in your brain when you say them, you use the same single
concept, as in between "that's" and "that is".

> I am on the one hand, not sure how you
> think English frames and pragmatics comes into
> and, on the other, what else you would expect to
> come into it.

"That's" and "that is" use the same concept. If someone were to say
that they were different, it would only be in terms of framing or
pragmatics (of course, those two don't differ so much in framing or
pragmatics). Likewise, "the students" and "the group of students" use
the same concept when used in "-- surround the building". A difference
of language - of the English language.

Yeah.

> Since no one has pointed to one, you need to show
> us where it is or what claiming there is one
> comes down to in reality. The existence of a
> mass is the crucial one but you just assume it
> (question-begging).

That's beside my point. You wanted a feasible explanation of my
singularist handling of "the students surround", and I gave you one.
Anyway, in response:

Sure, but you assume the non-existence in the exact same manner.

I've stated that a) people can and constantly do think of 100 students
as a thing called a crowd (for example) and b) that this
interpretation makes "the students surround" perfectly sensible, and
*explicable*.

You have not given a sensible interpretation of what "the students"
means in the surrounding example. You've agreed with mine, that they
'could' be seen as a mass, but said that they don't have to be. Well,
what the heck else could they be seen as? "Together"? Yeah, that's a
synonym for "as a mass/in a group" - what else could it be? Can you
explain this other thing that it could be, instead of just alluding to
it?

What do you want me to say, "A mass is anyt...hing composed of
component parts"? I'm sorry, but it seems that the idea of a mass
having component parts needs to be mentioned. That's what a mass is.
Strictly, no, it's not a relationship - nothing is really a
relationship.

> What you cite is
> the relationship between a mass and its
> components, which, obviously, presupposes that
> there is a mass to begin with.

Right. That masses exist. Where did I say that I was talking about a
mass formed of those students? No - that the relationship exists. That
it's possible for me to say "X is a mass formed of component parts Y".

Here's what you asked:

> > Please provide the explanation for the
> > mass-talk form. Note, this will require saying
> > it without assuming masses or giving a fairly
> > complete formal system for masses.

Yes? Yes. So here's my explanation for my understanding of a mass:

Alice is part of X
X surrounds the building

or

graphite-rod (component part) is part of pencil (mass/composite entity)
pencil is on the table.

Am I assuming masses? Yes. I "assume" them because this relationship
actually exists. I almost expect you to argue that there is no such
thing as anything that is a mass, or anything that is a component part
of a mass. Hopefully you won't.

> What is it � and

A mass. Some people call it a crowd. Other people call it "the
students". Whatever you want to call it, it's the (one) thing that
surrounds the building.

> where is it in the situation of studnets around a
> building.

What's the semantic difference between "the mass of students
surrounded the building" and "the students surrounded the building"?
One of them treats it like a mass... and the other one doesn't? It's
not enough to describe a "right hand" as "not a left hand". What does
the other one treat them as?

> As for predicate relatioships, that
> seems another issue altogeher but I agree that
> they exist, for whatever that is worth here.
>
> <<I'm using perfectly established structures, in
> English or otherwise,
> to do the explaining of how Alice relates.
> Predicate relationships,
> and the idea of "x1 is a composite of component
> parts x2". Both are
> established in both our minds, right?
>
>
> You did say "Gladly". Could you now do this?>>
>
> Sure, as soon as you give the explanation.

I've offered it many times. You seem to disagree with it, but I have
offered it. I disagree with yours because you haven't offered it.

You can very well give the explanation concurrently, I'm really not
interested in carrying on an argument of "you first". If it pleases
you, pretend that I've failed to offer an explanation, and go on with
offering yours.

> at all: there are just the students � and the
> building, of course.

General "you", as in "one really doesn't bring in anything new,"

>
>
> <<because your mind probably
> has "mass of 20 students" 'loaded' (though of
> course, the pragmatic
> implications of "mass" and all the 'framing' that
> "mass" entails are
> not loaded), because humans don't usually like to
> 'load' each of 20
> students when they don't have to. "Mass of
> students, 20 component
> parts" is good enough for most people.>>
>

> I don't think I have "mass" loaded � or indeed,
> except a figure of speech � have it at all.

So you aren't capable of recognizing that a pencil is a mass of
graphite and wood? Or what do you call that, a thing that is made up
of other things?

>
> <<> I presumably have some reason
> > for dealing with them together but that is
> > nothing "out there" called "bunch," it is just
> > how I am dealing with them.
>
> So there isn't anything out there? Or just
> nothing out there called
> bunch? Because if there isn't anything out there,
> I can't imagine you
> explaining what Alice's relationship is. With
> humor, I imagine
> something like:>>
>
> Nothing out there called a bunch.

...so there's something out there? This is my position, not yours.
Surely you don't mean that there is something out there, which is the
students together, but not called a "bunch" or "mass" etc.

>
> <<Alice is ? ? ?
> ? surrounds the building.
> where "?" stands for "magic happens here".>>
>
> I have already filled in the question marks for
> the pluralist view; what is the similar filling
> for the singularists.

Where?

You mean she participates in the surrounding of the building? Uh, no,
she participates in the wearing of hats too, etc.

When I ask you to explain, I don't mean "use explanations that I've
already shown to be inappropriate for legitimate reasons".

>
> <<Well, it's not magic. The "rational
> explanation", if you will, is that
> Alice is part of a mass/group, the mass/group
> that surrounds the
> building. If you have a different rational
> explanation, then please
> offer it.
>
> Calling it rational does not not make it
> intelligible or accurate. I have, as noted
> several time now, provided the alternative (well,
> not an alternative, since you haven't provided a
> the first yet, and also because I suspect yours
> will turn out to be about the same).

Alice is part of X
X surrounds the building

Graphite is a part of this pencil
This pencil is on the desk


This isn't a demonstrative example? It's not accurate? Intelligible?

>
> <<>
> >
> > <<If they said "the students surrounded the
> > rope", then you might have
> > an argument as to how it's meant. But if we say
> > "the group of students
> > surrounded the rope", then it's clear that we
> > mean the *group* (of
> > students), and not anything else.>>
> >
> > Not clear at all, since I don't see any group
> > there, just students.
>
> When I say "the *group* of students", you can't
> imagine a group?>>
>
> I can imagine all sorts of things, but I don't
> perceive one as necessarily there.

When I tell you "I saw a group in the street", you can't perceive that?

>
> <<> If you mean "the group of
> > students" to say, in different words, just what
> > "the students together" says -- that is,
> without
>
> No, "the students together" in your mind for some
> reason can't have
> the same meaning as it does for me and the
> dictionary (1. In or into a
> single group, mass, or place), it seems. We don't
> say the same things,
> because your variant excludes any possibility of
> "mass" in order to
> describe the relationship (it seems).>>
>
> Well, there are other definitions and some of
> them sound quite like what I mean. It does occur
> to me, howver, that your "group of students" is
> not "students together," but just "students"
> plural.

No. It's "a group".

There is no more.

Alice is part of the mass/group that surrounds the building.

That's all. Just the concept that together the students are something
that they are not on their own.

> building and a bunch of people � those were the


> things you just said were not the same.

You're asking me what the difference is between those. I'm asking you,
the difference between what?

No, I said that it's useless to care about the types, just like it's
useless to care about what every animal could be when thinking of
naming the concept of "animal".

> My point is that, in
> Lojbanology, "mass" gets used for a number of
> different concepts ("animal" presumably does not,
> for, while there are many kinds of animals, they
> are all animals under the same difeinition. That
> does not appear to be the case with masses (in
> the Lojban usage)).

A corporate mass is still a mass, and a reducible-to-same or
combineable-to-same mass is still a mass, just like a dog is an
animal, and so is a cat.

>
> <<Water, which can be combined or reduced into
> more water is a mass. I'm
> a mass. Just about everything is a mass. A crowd
> that makes noise is a
> mass, even if some people in it are quiet. Or
> even if all people in it
> are (relatively) quiet.>>
>

> This makes the term "mass" � already rather
> attentuated in Lojban � virtually useless.

How so

> That
> is why I suggested another term for what seemed
> to me a minute ago a rther specific meaning. I
> gather that that appearance was misleading and

> you reallywant something as muddled � indeed,
> more so � as Lojban usage. Water is a mass in the
> mass-noun sense � continuous, individualized only


> by portination, etc. You are a mass presumably
> in the sense that you have components that fit
> into a organizational scheme. Other things are
> masses in other ways (though usually in this
> latest one too). A crowd is a mass in the sense
> that it has components that are joined together
> to do some thing (make a noise, in this case).
> And so on (as noted, Lojban mass includes all
> these in various ways).

Everything is a mass. The utility of "mass" lies in being able to
refer to it by its component parts, or in describing what component
parts something is made of.

>
> <<If you want to taxonomize and label all of
> these different things that
> a mass can exhibit, or when certain masses stop
> being masses, or if
> you can combine certain masses to form a mass
> that is considered to be
> the same thing as the two masses were, go ahead.
> But it's still a
> mass.>>
>
> Only because "mass" has become so completely
> broad as to make no useful distinctions at all.

A mass made up of "those students" is different from a mass made up of
"these metals". I consider that a useful distinction. What kind of
distinction did you expect to have?

> Go ahead and use the term if you want, but don't
> be surpised if you get thoroughly misunderstood
> as a result. Or give the term a more specific
> meaning and thus help to make your point (there
> is a point here somewhere, isn't there?) clearly.
>
> <<> But in any case, I don't see
> > how this helps with the students: they do not
> > compose an organism or an organic whole, and
> > maybe not even an organization.
>
> If 1000 people together do not compose a "crowd",
> then what is a
> "crowd"? Just a way to refer to the 1000
> conceptualizations of people
> that you have "loaded" into your mind? Even if
> the crowd starts doing
> things that none of the people do on their own?>>
>
> Huh? 1000 people together would be a crowd
> usually, especially if they are doing something
> that needs the lot of them. "Crowd" is also a
> way to refer to 1000 people (or more or less) who
> are just milling about, perhaps with no common
> goals or activities at all (I don't generally

> refer to conceptualizations, just to people �


> certainly with the word "crowd" or "people").
> And I don't see what you use of "loaded" does
> here.

Think of a fork. Sense how there's an instance of a fork in your mind
now? Hold up your hand. Sense how there's a (rather specific) instance
of a hand in your mind now? That's what I mean by "loaded" - I have no
better word for it.

> concept of mass (whichever one you want � here I


> suppose the one that does what the pluralist says
> is done by things together) does not exist. I
> merely ask you to explain what it means to say
> that it does certain things. I have said what it
> means for the students together to surround a
> building; what does it mean for a group of
> students to surround a building? You were

It means that x1, which could presumably be called a crowd if you
wanted to refer to it directly, surrounds the building. The students
are parts of x1.

> claiming that only this latter locution is
> legitimate (or, at least, that it is the more
> accurate locution); my response is to say that,
> so afar as you have shown, it is not yet even an
> intelligible locution and to suggest that, when
> it is made intelligible, it will turn out to be
> the same as the mean just the same as the
> "together" locution.

What is your "together" locution?

> As you said in a different
> context, this is less an argument than a
> challenge, although the challenge is within the
> frame about whether group talk is intellible and
> more correct than the alternative. Occasionally
> you say something that sounds like saying that
> you recognize tht the two locutions say the same
> thing, but then you seem to go back to the

> position that only one of them is correct � an


> odd combination, so I suppose they are two
> separate points.

This is because my interpretation is

Alice is part of X
X surrounds the building

Sometimes, you seem to agree. Other times, you assert that there is no
X, and do this

Alice ?? ? ?? ??
?? ? ? surrounds the building

and then you offer explanations like this:

Alice ? ^ & ??
?# ? ? surrounds the building

or like this:

Alice
surrounds the building (yes, well she wears hats too...)

I need you to fill those places in with things that we can both agree exist.

Er, /how/ a mass surrounds a building? Well, a mass is an entity, so I
guess the answer is "in any way that would lead us to believe that
that entity is surrounding the building".

It's like asking for an explanation of Alice surrounds the (thing that
was a) pancake that she ate earlier. She just "does". "First order"
relationships don't really need explanations, I hope.

> The fact (if it is one) that each
> student is also a mass and the building one, too
> shed no useful light. If the point of this is to
> convince me there are masses, I never have denied
> that and so this chat is irrelevant. If the
> point is somehow to show how masses are involved
> in surrounding a building, it has so far failed,
> largely because no effort has been made to get
> behind masses to what is going on in mre neutral
> terms.

Ok, let's call the thing that the students are together a "srowak". A
srowak surrounds the building. The srowak is composed of students.
Does that make it better, or?

>
> <<The argument isn't really about this (I hope).
> It's about whether or
> not an entity with parts:students can exist. I
> say that it can, and
> frequently does. And I also say that this entity
> is the thing that
> surrounds the building.>>
>
> Now this is just confusing (confused?). No one
> denies that entities with parts exist; as you
> say, I am one. I take it that you think that
> "students" refers to such an entity rather than
> to several entities (the individual students, the
> putative parts of your entity).

Right

> I am not even
> denying that such an entity can exist; I am
> merely asking what it means to say that it
> surrounds a building.

I don't know. It's an entity, and it surrounds this other entity. It
surrounds in the same way that a former pancake can be surrounded, in
the same way that a ring surrounds something.

> As a pluralist, I don't
> have to acknowledge that it exists or that it
> surrounds a building, since I can account for
> students surrounding a building without it.

I've been asking you for an explanation of how you account for this,
you haven't yet provided one.

> From
> that point of view, I can wonder what it would
> mean to say that a mass is surrounding a building
> and ask you to explain in terms that I
> understand.

A fork exists as a fork only through us recognizing the relationship.
A crowd exists in the same way. What's the nature of "calamity"? A
"mass" is much more flexible. When something can be said to be a part
of something - one action being part of another larger action, an
event being part of..., the pull between two molecules being part
of... - that's the "composite entity - composite part" relationship.

> I have explained what "the students
> together surrounded the building" means in terms
> I assume you can understand, since the
> explanation contained no troublesome words like
> "together" or "set."

Where?

> I invite you to do likewise
> (or point out what about my explanation you don't
> understand so that I can readjust it to your
> apprehension). Then we can examine whether there
> is any reason to think that only your locution
> involving masses is correct or whether only the
> together version is correct or whether they are
> equally correectr and maybe even identitical
> behind the forms.

The "together" version is my version. "As part of a group/mass" is how
I use the word together.

>
> <<> Ok, then if it's not connected to the act of
> > "surrounding the
> > building" by way of a group, then how is it
> > connected? What is the
> > relation?>>
> >
> > Directly by each of them taking a place in a
> > pattern which constitutes surrounding the
> > building. You may call "taking a place"
> "forming
> > a group" but there is no necessity in doing so.
> >
>
> Ok, sure, that's another sensible way to think of
> it.
>
> [da poi sruri lo dinju] cu morna [la alis]
>
> x3 can even be "surrounding-the-building-wise".>>
>
> I am not sure Alice is appropriate for {morna2},
> she isn't a part of the pattern, after all, but
> occupies a place in that pattern, an x2. And, in
> general, what surrounds the building is not the
> pattern but things in that pattern.

A pattern isn't a schematic. Arrange some tiles in a certain way, and
it's not the places where the tiles are that form a pattern, but the
tiles being where they are that forms it.

Well, the crux is that I am not aware of any other explanation.

> nor does it constitute an explanation of
> the original (except, perhaps, for someone who
> heretofore spoke only in metaphysical
> periphrasis). But more to the point, that these
> two are equivalent is a hypothesis of mine (well,
> I know it holds for formal systems, the issue is
> whether it holds in normal language).

Good, because that's my hypothesis also (though perhaps in a different
sense), but I have to ask, how can they be equivalent if they do not
explain each other, if they aren't interchangeable?

> It needs
> to be demonstrated and to do that we need to know
> what "A group of students surrounded the
> building" means in the way that we know what "The
> students together surrounded the building" means.
> We can then compare. Clearly, if the existence
> of an entity, group of students, distinct from
> the students is essential to this notion, an
> irreducible factor, then they cannot be
> equivalent. But that will leave us with the
> problem of whether there is in this situation any
> such entity. A straightforward count of the
> factors involved in fifty students surrounding a

> building turns up 51 � the students and the


> building, not either 52 (an added group) nor 2 (a
> group and a building). What can convince us to
> change this count?
>

Are "the students together" the same as any of student 1, student 2, [...]?

John E Clifford

unread,
Jun 7, 2006, 1:17:42 PM6/7/06
to lojba...@lojban.org
--- Maxim Katcharov <maxim.k...@gmail.com>
wrote:

> On 6/6/06, John E Clifford
> <cliff...@sbcglobal.net> wrote:
> > --- Maxim Katcharov
> <maxim.k...@gmail.com>
> > wrote:
> >

> > > On 6/5/06, John E Clifford


> > > <cliff...@sbcglobal.net> wrote:
> > > >
> > > >
> > > > --- Maxim Katcharov
> > > <maxim.k...@gmail.com>
> > > > wrote:
> > > >

> > > > > On 5/29/06, Jorge Llamb�as
> > > > > <jjlla...@gmail.com> wrote:
> > > > > > On 5/29/06, Maxim Katcharov
> > > > > <maxim.k...@gmail.com> wrote:
> > > > > > >
> > > > > > > Avoiding the word "mass"/"crowd"
> when
> > > you
> > > > > say "the students" does not
> > > > > > > mean that "the students" does not
> refer
> > > to
> > > > > a group of students. It
> > > > > > > does.
> > > > > >
> > > > > > That's the singularist view, yes. But
> it
> > > is
> > > > > not the only possible view.
> > > > >
> > > > > Ok, then please show an alternate view.
> > > You've
> > > > > flatly asserted that
> > > > > one exists, yet when I ask you to
> explain
> > > it, a
> > > > > vague two-word answer
> > > > > ("the students") with no explanation or
> > > > > demonstrative examples is all
> > > > > I get.
> > > >
> > > > On a pluralist view, reference is a
> relation,
> > > not
> > > > a function, so that a single term may
> refer
> > > > simultaneously to several things.
> > >
> > > Sure. In my singularist view, I too prefer
> to
> > > think of it as a
> > > relation. "run(dog, road)" seems silly to
> me.
> >
> > I don't see what this is illustrating or,
> > perhaps, just what is says about the way
> > reference is treated: function or relation.
>
> That the pluralist view is not the only one to
> think of things as a relation.

I wasn't talking about things in general but
about the device for giving the referet of an
expression. I don't think anyone ever doubted
that singularists as well as pluralists use
relations (the alternative was demonstrated to be
hopeless in the mid 17th century).

> >
> > > The question that I pose
> > > is: what is the nature of the relation
> between,
> > > say, Alice (one of the
> > > students that surrounds the building) and
> the
> > > surrounding of the
> > > building? The relation is crystal clear
> between
> > > Alice and the wearing
> > > of a hat, but the building-surroundment
> > > relation seems to be
> > > vaporizing as xorxes tries to nail it down.
> I
> > > suspect that this is
> > > because the true nature of this pluralist
> > > relationship is that of a
> > > mass - the relationship is that Alice is
> part
> > > of a mass/group that
> > > surrounds the building, and that there
> simply
> > > is no other sensible
> > > interpretation.


> >
> > Well, I suppose that Alice's relation
> surrounding
> > the building (when she is one of the students
> > surrounding the building)is "participation."
> I
>
> Participation in an event? xorxes already
> offered this. Consider "the
> students surround the students". What is Alice
> participating in?

Well, is Alice among the surrounding or the
surrounded? Those seem to be the two events in
which she could participate. In the one case she
is (more or less) on the outside looking in, in
the other on the inside looking out.

> > Or if I try to specify it in extension,
> > spelling out how she particpates (standing
> NEbyN
> > of the building at the same time as others
> are
> > standing at the other points of the compass,
> say)
> > you will relate that to being a member of the
> > group as well.
>
> Well, yes. This is the method of participation.
> For example, I can say
> "together the three men lifted the piano, by
> method of one man
> directing, and two men bearing".

This tells me what each does by way of
participating, but I still don't see anything
like a group here unless it is just the fact of

the perticpation being described in some


organized way. And that is just what a pluralist
would mean by "together," more or less.

> > To which I can only say
> > "Precisely" -- singularist and pluralist
> > languages are two different ways of stating
> the
> > same facts.
>
> Not quite. The pluralist view asserts that you
> don't introduce masses.
> Instead, there's a special "bunch-together" (or
> something - it hasn't
> exactly been elaborated upon) that supposedly
> handles the questions
> raised by the removal of "mass".

Well, you haven't introduced any masses yet
either (aside from assuring me that they are
there). Back to the students around the
building. Each student occupies a place wrt the
building and other students, roughly (let's say)
that if simultaneously each student joined hands

with thei neighgbor on each side the result would


be a closed loop and the footpad of the building
(and little else?) is entirely inside the loop.
The way I am reading the claim, I think it
requires that each student intends to be part of
surrounding the building, but there are other

readings which don't demand that. There are
problably more consitions but this seems to me to


be the essential one. The "together" of the
pluralist is just the fact that this pattern
requires all the students involved (which is
trivial) and perhaps that with many fewer
students similar patterns (that formed closed

loops arounf the building) are not possible --


certainly that no one student can form such a
pattern. Does the groupiness consist of anything
other than this? You've already said it is not a
thing over and above the students, so that the

students form a pattern seems to be the most
obvious next choice. But that, of course, means
that for reality, it just says what the pluralist
says but in differnt words. If it is something
else, that you need to say what and demonstrate
that it really is there. It seems that the
pluralist says "there are these students and they
form this pattern" and the singularist says
"there is this pattern and the students for it"
Why this stife there be/'twixt Tweedle-Dum and
Tweedle-Dee?

> > They are completely
> > intertranslatable in a one-one mechanical
> way.
> > You want a pluralist claim that is not
> > interpretable as a singularist one and there
> just
> > ain't any. This whole discussion is totally
> > vacuous.
>
> If pluralist "loi ci tadni" translated into "da
> poi gunma lo ci
> tadni", then everything would be just fine. But
> that's supposedly not
> what it translates to.

Well, I am not sure why not, but I'll trust
xorxes on this. A pluralist would read this
simply as saying "something which is lo ci tadni
taken collectively", that is "lo ci tadni taken
collectively" which looks essentially right, if
not in some detail. Not a thing, in a word, but
a way that some students are treated
linguistically.

> >
> >
> > > > A sentence
> > > > using this term will be true if those
> things
> > > are
> > > > in the extension of the predicate in the
> > > > appropriate way, either individually or
> > > together.
> > > > From this basis, a complete semantics
> can be
> > > > (has been) developed, which produces the
> > > > classical system with the "among"
> relation
> > > added.


> > >
> > > Elaborate? To me, "among" has implications
> of
> > > being "among a group such that".

Well, of course it would; you are a believing
singularist. For a pluralist, "x is among y"
just means that x is one of the ys.

> > And so it does -- when used by a singularist.


> > When used by a pluralist, it doesn't. But
> the
> > properties of "among" are the same for both.
>
> But in the pluralist view, there's still a
> group there, you just don't
> choose to acknowledge it, right?

Where? Go through the whole pluralist semantics
and nothing like a group turns up, just things,
one or several as the case may be. At the end of
it all, it is hard to say where the
unacknowledged group might be.

> >
>

=== message truncated ==

<<>
> > > In a totally parallel way, we can develop
a
> > > semantics with things and masses and the
> > usual
> > > definitions of truth and get the same
> > classical
> > > system with "among" added. What is said is
> > the
> > > same, the conditions for truth are totally
> > > intertranslatable, and so on.
> > >
> > > > I doubt that you'll be left anything to
> > explain
> > > > your position with
> > > > once you start explaining. The pluralist
> > view
> > > > relies on not looking
> > > > too deeply at what "the students" means,
> > > > because once you do you see
> > > > that it's either a mass, or the students
> > > > individually.
>
> This is merely metaphysical hubris: it's my
point
> of view, so the other must be defective in some
> way. Unfortunately, any way that the pluralist
> view is defective, the singularist is defective
> in an exactly matching fashion (in this case
> creating an entity that has no place in
reality).

What you just responded to wasn't so much an
argument as a challenge.
Fact is, explanations of how "bunch-together"
differs from "mass"
aren't really available. I attribute this to
there being no
explanation of "bunch-together" that is different
from "mass".>>

I attribute it to the fact that there is no
difference except verbiage. You seem to think
that the mass form the explanation is right and
the the other wrong, which is odd if they are the
same explanation. However, this is all empty,
since we have neither explanation at hand yet (I

have tried to suply one but I don't know whether
you will buy it).

<<> > > > >
> > > > > > Show me how and what "the students"
> > refers
> > > > to.
> > > > >
> > > > > In the pluralist view, it does not
refer
> > to
> > > > one thing. It refers to
> > > > > many things,
> > > > > i.e. the many students.


> > > >
> > > > Ok, then when I say "group of students",
I
> > too
> > > > am "referring to many things".>>

I agree, but you seem to think that you are
actually referring to one thing, the group. At
least you talk that way.

<<> > > > Avoiding the word "mass"/"crowd" when
you
> > say
> > > > "the students" does not
> > > > mean that "the students" does not refer
to
> > a
> > > > group of students. It
> > > > does.>>

Well if you mean by "a group of students" only
what a pluralist means by "the students", then
there is no problem, but you seem then to be
misled by what you say into insisting that there
is only (or also) one thing, the group.

<<> > > By you, yes. By xorxes, no -- it is all
> > about
> > > the pictures in your head.
> > >
> > > > "[The [many students]]" refers to a group
> > of
> > > > students.
> > > >
> > > > >
> > > > > > Additionally, I don't think that
Lojban
> > > > uses this mistaken concept of
> > > > > > "plural predication": it seems that
the
> > > > book that describes it has not
> > > > > > been published yet, and so Lojban
> > predates
> > > > it by about 20 years.
> > > > >
> > > > > That may be true. Is your argument then
> > that
> > > > conservatism requires
> > > > > that we stick with the singularist
view?
> > (CLL
> > > > does concede that pronouns
> > > > > at least can refer to "individuals" or
> > > > "masses" depending on context,
> > > > > so even there one can find, at least in
> > > > embryonic form, the pluralist view.)
> > > >
> > > > My argument here was that the burden of
> > proof
> > > > is on you to show that
> > > > a) this pluralist view exists and is
> > correct,
> > >
> > > Exists is easy; there is the book (and a
> > number
> > > of others going back to the late thirties).
> > Is
> > > correct doesn't arise if the alternative is
> > the
> > > usual singularist view, since they are the
> > same
> > > thing.
> > >
> > > > and b) that Lojban uses
> > > > this pluralist view. Until you do this,
you
> > > > should not attempt to use
> > > > this pluralist view in Lojban.
> > >
> > > We can't tell, of course, which one Lojban
> > uses
> > > because we can't get inside Lojban's head.
> > > Further, Lojban does not have devices for
> > > expressing some crucial distinction in the
> > > theory.
> >
> > Which distinctions?
>
> Primarily the difference between distributive
and


> collective predication. Even {loi} does not
> appear to be just collective predication -- it
> seems clearly to involve corporate and Urgoo
> cases as well. And there are cases which
cannot
> be dealt with using gadri.>>

Examples? I see no practical differences between
corporate masses and
regular masses, and I'm not familiar with Urgoo
cases at all. >>

Corporate masses (I don't much like that

terminology since it sugtgests more similarity


than I think justified)continue to be the same
even with a change of components; they also
inherit properties from their components
directly: if a component (acting as such) does or
is something, the corporation does or is, too.

Corporations also have properties in which some
components do not participate. I suppose there
are other charateristics but these are enough to
separate then from ordinary (collective
predication) masses. Urgoo is the stuff of which
some kind of thing is made: all dogs are chunks
of Dog, for example -- as are dog organs and the
mixture that results from a steamroller rolling
over a pack of dogs. This is an actual mass-noun

concept. So far as I can tell, Urgoo is like


corporations in some respects: it remains the
same even if its representations change, it

inherits propeties from its manifestations. It


differs in that it is homogenous, does not have
components, although the manifestations play a
somewhat similar role, but an Urgoo can exist

without any manifestations at all. I think these
two are enoguh different to justify some separate


consideration but both have been folded into the
muddle that is CLL mass.

<<Dividing
things into special classes of masses isn't very
useful. Yes, the
components of water are usually more water, and
yes, perhaps two
car-makers together have the potential to be seen
as a single car
maker, while two bottles together would not be a
single bottle.
However, while this serves to explain what things
can be seen as
masses and which can't, it doesn't say anything
about masses as they
are. A mass is just a mass - who cares if you can
split it up into
more of the same or not?

This seems to be about Urgoo and corproations,
not about masses in what I think of as the
currently primary sense participants (or a group
of particpants) which are predicated of together
(which is predicated of directly).

<<>
> > > So the best thing to say is that Lojban
> > > ut nunc does not adhere to either view but
> > > sometimes does things that look like one,
> > > sometime like the other. The proposal,
> > stripped
> > > of its picture thinking, is just to make
> > Lojban
> > > adequate for the view(s) and so get rid of
a
> > > number of false starts and missteps that a
> > > previous state of ignorance forced on us.
> >
> > My position is that if there was a state of
> > ignorance before, it's
> > being solved now by inducing a confusion, and
> > then not thinking too
> > deeply so that one does not see the problems.
>
> Well, that is polemics, not testable claims.
Or,

Sure. My arguments lie elsewhere.>>

Where? Assertion is not argument.

<<> if testable, then false, since the theory is
> there before you (the one that is not easily
> documentable is the singularist one, actually).

It's difficult to document?>>

Yes, while there is plenty of stuff about set
theory and standard logic, there is nothing much
that deals directly with the issue of how sets
(C-set, L-sets to a lesser extent) solve the
problems of representing plurals. The standard
language does not have predicates that do the
trick directly and the various work-arounds seem
to be discussed very rarely.


<<> <<>
> > > >
> > > > > Then what surrounds the building?
Please
> > > give an explanation,
> > > > > hopefully a detailed one, as opposed to
a
> > > vague 2-word answer.
> > > >
> > > > I'm afraid nothing further I might add
will
> > > change your mind. Luckily
> > >
> > > Why are you using the word "further" here?
> The
> > > only thing you've done
> > > to change my mind is answer "the students"
> when
> > > I ask "what does 'the
> > > students' refer to?".
> >
> > But that is a completely adequate answer. If
> you
> > don't see that, then it is unlikely that
> anything
> > else will work either. I would be inclined
to
> > have said that none of this matters, but that
> is
> > not going to be a point that works 200 some
> > entries into the discussion.
>
> No, it's not an adequate answer. When you ask
me
> about my position,
> "what does 'together the students' refer to?",
or
> "what does 'the
> students (individually) refer to?", I can, and
> have, given answers
> that were much more elaborate than "together
the
> students" (aka
> "mass") and "the students (individually)" (aka
> "bunch-individually"),
> respectively (though the answers were still a
bit
> crude). I could
> potentially write pages of explanations of the
> differences between the
> two. Not so with "bunch-together". The best
that
> can be done with that
> one is to call it by different names.>>
>
> But, of course, "bunch" (outside its technical
> use) or "mass" or even "student individually"
> also adds nothing to the discussion: it is a
word
> with (to a pluralist certainly) has not
> significant content.

"Mass"/"together" expands to "x1 is a mass with
components x2". This
is an actual relation. I consider that as
significant in terms of
content as you can get.>>

But you offer no evidence that it applies here.
"Together" is a real situation as well and I have
offered an explanation of what it means in
different terms. What does "is a mass composed
of" mean in different, neutral, terms. Failing
that we are just talking by one another, since we
are using language radically differently.

<<> They, on the other hand,
> would find oit odd that you cannot understand


> such a straightforward English expression as
"the
> students" (especially since you seem to
> understand the mysterious "the mass of
students").

It's about as mysterious as "the building for
students" - that is, not
mysterious at all. "the students", on the other
hand, is ambiguous: it
can refer as in "the students wore hats" or "the
students (as a mass)
surrounded the building", and then, of course,
there's also "the
students (as a bunch-together) surrounded the
building", which nobody
has really explained or demonstrated as being
different from "as a
mass", though copious flat assertions of the sort
have been made.>>

But you, of course, have nowhere demonstrated
that "as a mass" is different from "together"
nor explained what it meant. You have asserted

it is superior, but that is just your say-so. On
the other hand, if you really believe, as you
seem to be saying here, that the two expressions
mean the same thing, what is the argument all
about?


<<> Note that, if you do write pages explaining
the
> differnce, the pluralist can take it, make a
few
> uniform changes and provide you with the
> explanation you want for the difference between
> "the students individually" and "the students
> together."

Please, do it then! Do it with the crude
paragraphs I've offered. What
are you arguing this with me for, when simply
demonstrating this would
solve everything?>>

Gladly. Pleas provide the explanation for the


mass-talk form. Note, this will require saying
it without assuming masses or giving a fairly
complete formal system for masses.

<<>
>
> <<>
> >
> > > > for you, and for anyone else who prefers
> the
> > > singularist view, nothing
> > > > in Lojban prevents you from putting that
> view
> > > into practice. If you are
> > > > consistent with your view you simply
won't
> > > apply a distributive and a
> > > > non-distributive predicate to the same
> sumti,
> > > you will always have
> > > > to split your bridis in two in such
cases.
> > > This may make some things
> > > > more cumbersome to express, and I see
> nothing
> > > gained by it, but it's
> > > > always doable.
> > >
> > > Please show (a) and (b) before attempting
to
> > > use your pluralist view
> > > in Lojban. Until you do, you should use the
> > > singularist view.>>
>
> Why? It is not privileged until shown to be so
> (which it cannot be, of course). Lojban is not
> consistently either -- assuming we could tell
> them apart.
>
> <<> As noted, Lojban's adherence to the
> singularist
> > view in detail is as open to exception as a
> > pluralist view -- Lojban can't express either
> one
> > in any thorough way.
> >
> > > > > This brings us right back to:
> > > > >
> > > > > 2) You can't use {lo danlu cu bajra
gi'e
> > > blabi} to refer to a white
> > > > > dog and running cats,
> > > >
> > > > Right, because the animals that are
running
> > > are not the same animals
> > > > that are white. In the case of the
> students,
> > > the people that are wearing the
> > > > hats are the same people that are
> surrounding
> > > the building. If they were
> > > > not the same people you could not use one
> > > sumti for both predications.
> > >
> > > You're switching the meaning of "the
> students"
> > > in mid-sentence. The
> > > thing that surrounds the building is one
> thing.
> >
> > I wonder if this is really defensible. If
you
> > ask someone how many things are surrounding
the
> > building, I expect that the answer "Fifty
> > students" will be more frequent than "One
group
> > of students." "A bunch of students" is also
> very
> > likely, but flat ambiguous, if you think
> > singularist and pluralist are really
different.
>
> If I ask someone what surrounds the building,
> they'll answer "a bunch
> of students" or "a group of students". "Bunch"
> will be used in the
> sense of "group", and not in the sense that
we've
> defined it for the
> purposes of this conversation.>>


>
> You will no doubt take it that way; how are you
> sure the speaker meant it that way or even that
> he can sense the difference?

Uh, because "bunch" doesn't have the definition
that we've assigned it
(for the sole purposes of this argument) in
common use. Bunch is
simply "group", with implications of the things
being close together -
"bunch of twigs", etc.>>

Well, it does seem to have that meaning in my
dialect. That is, when I say "a bunch of things"
I am not implying that there is anything other
than those things there (not even necessarily

close together). I presumably have some reason


for dealing with them together but that is
nothing "out there" called "bunch," it is just
how I am dealing with them.

<<If they said "the students surrounded the
rope", then you might have
an argument as to how it's meant. But if we say
"the group of students
surrounded the rope", then it's clear that we
mean the *group* (of
students), and not anything else.>>

Not clear at all, since I don't see any group

there, just students. If you mean "the group of


students" to say, in different words, just what
"the students together" says -- that is, without

any reference to some other thing than the
students -- then, we agree and the arguemnt is
just about which is the better way to talk. and
the answer would surely be that each does as well
as the other, up to personal preferences.

<<>
> << I will less frequently get the answer
> "fifty students", because it's seldom that
people
> miss the forest for
> the trees, or in this case the crowd for each
> student, and when I do
> receive that answer, it'll be in the sense of
> "fifty students
> together".>>
>
> Precisely (though I am not so sure about your
> statistics). They mean the 50 students
together,
> not something other than the students.

You seem to think that "introducing" the entity
of this student group
is as odd as introducing the entity of a baboon.
It's not. It's
already in context.

"The 50 students surrounded the building" and
"the group of 50
students surrounded the building" are synonymous
in meaning. It's just
that one of them uses the word "group", which
invokes a certain frame
in your mind that the omission of the word
wouldn't.>>

Then what the Hell is this argument about? One
person talks one way, the other the other, as
their taste leads them. And, of course, that is
just what the formalism says: whether you give a
pluralist or singularist interpretation to the
system, the logic is the same.

> Forst are
> just trees after all (with some exceptions like


> willow forests which are apparently just one
> tree). (I don't of course, really mean this. I
> am just pointing out how useless taking what
> someone says is in figuring out which of the
> identical sides they are on.

A forest is not the same thing as a set/"bunch"
of trees, just as a
human is not just a set/"bunch" of organs... just
as a crowd
surrounding a building is not just a set/"bunch"
of students.>>

And the difference is...? I suppose it is

somethign that hold them all together, a common
interest them. that is something about us


usually, although it is often helped by
propinquity and short-chain causation and the
like.

<<>
> <<>
> > > The thing(s) that wear
> > > hats are each something different. One
thing
> > > being composed of others
> > > does not mean that it is the same as each
> > > component part.
> >
> > And no one said it was.
> >
> > > I am composed of my organs. When I run, my
> > > organs do not run. My
> > > organs together (i.e. my body) runs.
> >
> > Even that is open to some question; bodies
tend
> > to be -- for purposes like running -- more
than
> > the sum of their parts (well, at least
> different
> > from).
>
> Right. They aren't a mathematical set, they're
a
> mass.


>
> More than that too, an organism. That is, the
> organs in an organization. Without the
> organization, the organs are just a pile of
> specimens.

That's what I mean when I say mass. I discussed
this earlier using the
example of a piece of graphite and a piece of
wood not quite being a
pencil. Search for the term "graphite" if you're
interested.>>

Ah, that was the point of that story. It was not
very clear to me at the time. Your use of the
term "mass" is adding yet another meaning to that
already overworked word; can we find another word

for you concept. But in any case, I don't see


how this helps with the students: they do not

compose and organism or an organic whole, and
maybe not even an organization. They each fall


into a place in a pattern which we are taking as
significant and by virtue of which say they are

togethre. Is it also by virtue of this that we


say they are a mass? If not, what is involved?
If so, why are we having this argument (or, more
accurately, what the Hell are we arguing about,

sinc we see to agree on everything except what
words to use and that is merely a matter os style


and not open to argumentation).

<<> I think you are making the case for a
> body being a corporate entity, not either a
msss
> or a set. We haven't got that developed yet,
but
> it does not seem to say anything useful here --
> unless you want to say that the students
> surrounding the building are also a
corporation.

Can you explain what you mean by "corporation"?
To me, it seems like a
needless taxonomy - how is a corporate mass
different from other
masses?

> I don't think that need be the case and, if it
> were, we would need a discussion different from
> the one so far.
>
> <<>
> > > >
> > > > > and so you can't use {[L_ muno tadni]
cu
> > > [dasni
> > > > > lo mapku] gi'e [sruri le dinju]} to
refer
> > > to a number of students and
> > > > > to a mass composed of students.
> > > >
> > > > Right, because the mass is not the
> students,
> > > so if you only allow singular
> > > > reference, you can refer either to the
one
> > > mass of students or to each
> > > > one student individually.
> >
> > But, of course, you can predicate of that
mass
> > distributively (or could if the language
> > allowed).
>
> What do you mean? Once you've made a mass, it's
> difficult to split it up.>>


>
> Set theory, which seems to be the model for
talk
> of masses,

A mass is a relationship, it need not have
anything to do with set
theory. x1 is a mass of composite parts x2.>>

Huh!? There is a relationship of composition that
defines a mass, but a mass is not a relationship
(notice, by the way, that {gunma} is not a mass

of the sort you descibed earlier). It may also be


that the fact that things stand in a certain
relationship to one another is what gets them
into the mass, but the mass is not that
relationship either.

<<> does indeed not allow direct
> predication of members through predication of
the
> set. It does it indirectly, by inclusion and
> perhaps some other relations. This is, of
> course, just an accident of the way the
language
> of set theory was set up. As Lesniewski and
now
> pluralist logic show, the set up could be
> different, making these distributive
predications
> easier (and more obvious). That is, splitting a
> mass (in this sense) is merely a function of
> language and can be changed with a change of
> language.
>
> <<>
> > > > But if you allow plural reference, then
it
> is
> > > the very same students who
> > > > wear the hats and surround the building.
In
> > > this case, the two predicates
> > > > are predicated of the _same_ referents,
and
> > > so you can use one sumti to
> > > > refer to them.
> > >
> > > What surrounds the building?
> > > (The students.)
> > > Does each student surround the building?
> > > (No.)
> > > Then what is it that surrounds the
building?
> > > (The students.)
> > > So you mean the students together?
> > > (No, the students.)
> >
> > Yes, the students together, not that is
> anything
> > other than the students; it is just a way
that
> > are predicated of
> > >
> > > I'm not being dense when I ask you these: I
> > > understand your position
> > > perfectly. You think that saying "the
> students"
> > > frees you from
> > > implying that they're a group. I recognize
> > > this, and I assert that
> > > it's incorrect. Avoiding the word
> > > "mass"/"crowd" when you say "the
> > > students" does not mean that "the students"
> > > does not refer to a group
> > > of students..
> >
> > Well, actually it does. At least it allows
it.
> >
>
> It does does not refer? What do you mean?>>
>
> I mean that using "the students" rather than
"the
> group of students" does mean that the "the
> students" does not refer to a group.

Ok, then if it's not connected to the act of
"surrounding the
building" by way of a group, then how is it
connected? What is the
relation?>>

Directly by each of them taking a place in a
pattern which constitutes surrounding the
building. You may call "taking a place" "forming
a group" but there is no necessity in doing so.

<<> Of course,
> you can mean that equally well using "the group
> of students," but it is harder to see. And, by
> parity of reasoning (since the two are formally
> identical) "the students" does refer to a
group,
> if you want to go that way, although it is
clearer
> if you say "the group of students."
>

What are formally identical? Thinking of them as
a group and not
thinking of them as a group?>>

Well, thinking of them as a group and thinking of
them as acting together.

To unsubscribe from this list, send mail to lojban-li...@lojban.org

Jorge Llambías

unread,
Jun 10, 2006, 10:06:40 AM6/10/06
to lojba...@lojban.org
On 6/10/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > >
> > > lu'o ro lo tadni cu sruri lo dinju
>
> If your grammar says that an outer {ro} on a {lo} marks it for
> distributivity, why is (1) exempt from this rule?

Because the presence of {lu'o} means {ro} is no longer the outermost


marker. The distributive/non-distributive marker marks a slot, and no
matter how many {lo}s and {loi}s and {lu'o}s and quantifiers are contained
inside the sumti expression, the one that determines whether the slot is
distributive or not is the outermost. An outermost quantifier is distributive,
an outermost mass-marker is non-distributive. It's that simple.

In addition to quantifiers, logical connectives are also distributive and


work just like quantifiers, and {joi} works like {loi} and {lu'o}. The neutral
connective, not marked for distributivity and which corresponds to {lo}
is {jo'u}. So if we have two people, Alice and Betty:

la .alis .e la betis = ro le re prenu
la .alis .a la betis = su'o le re prenu
la .alis .onai la betis = pa le re prenu
la .alis na.enai la betis = no le re prenu
la .alis na.anai la betis = su'epa le re prenu
la .alis .o la betis = ro ja no le re prenu

(The remaining logical connectives are not symmetric, and therefore don't
have a corresponding quantifier.)

la .alis joi la betis = lei re prenu

la .alis jo'u la betis = le re prenu

> > {lo rokci joi la alis cu sruri le dinju}.


> >
> > But apparently under your current interpretation, from that it follows that
> > {loi rokci cu sruri le dinju} and also that {lu'o la alis cu sruri le dinju}
> > (= {lai alis cu sruri le dinju}?). But neither of those follow at all, the
> > way I understand it.
>
> As I've said, this isn't English, you don't need the same pragmatics
> and verbatim translations.

I'm not talking about pragmatics here, I'm talking about what follows


logically from an expression. For all cases of ko'a and ko'e and broda,
independently of their meanings, under your interpretation you have
that from:

(1) ko'a joi ko'e broda

you can deduce:

(2) lu'o ko'a broda

Just from knowing that ko'a and ko'e do X together, you can deduce


that ko'a is part of a group that does X. There's no pragmatics

involved there. But for me (and also the way Lojban has always


been, as far as I can tell) {lu'o ko'a} is not "some group that has the
referents of {ko'a} as components, possibly among other components",
it means, in singularist terms, "a group that consists of the referents of
{ko'a}, no more and no less" or in pluralist terms it means that the
referents of {ko'a} do something together.

> Now, I could say {lu'o la alis cu sruri lo dinju}, but usually I


> wouldn't. I'd say {lo rokci joi la alis cu sruri lo dinju}. However,
> the former would still be true - Alice is a part of the
> surrounder/surroundment of the building.

Under your reinterpretation of {lu'o}, that's correct. Under the usual


interpretation, that's not correct. This is independent of whether
you take the singularist or the pluralist road. The singularist and
pluralist roads take you both to the same final place, but this new
spin that you want to put on {lu'o} changes it to something else.

> I'd still like to have that explanation of distributivity that I've


> been asking for.
>
> 1) {lu'o la tadni cu sruri lo dinju}
> 2) {loi tadni cu sruri lo dinju}
> 3) {la tadni cu sruri lo dinju}
> 4) {la tadni cu dasni lo mapku}
>
> What is Alice's relationship to each relationship?

(I assume you mean {le} or {lo} rather than {la}.)

In all cases, Alice is one of the referents of the sumti that appears in x1.

There's no more to it than that. The answer to your question "If Alice and


Betty do something together, what am I saying that Alice is doing by
herself?" is "I'm not saying anything about Alice more than that she is
one of the two people doing something together."

mu'o mi'e xorxes


Alex Martini

unread,
Jun 9, 2006, 6:57:29 PM6/9/06
to loj...@yahoogroups.com

On Jun 9, 2006, at 11:13 AM, Jorge Llamb�as wrote:

> On 6/9/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
>> For
>> example, I think that the following is grammatical under your rules,
>> but I find the interpretation infeasible (or awkward), since it uses
>> both "markers":
>>
>> lu'o ro lo tadni cu sruri lo dinju
>
> For me that's equivalent to {lu'o lo ro lo tadni} and it marks the
> slot
> that the sumti fills as non-distributive. It is only the outermost
> marker
> that concerns the distributivity of the slot.
>

Could someone explain the use of {lo} twice here? I've never seen
such a thing
before, and would have expected it to be ungrammatical.

ki'e mu'o mi'e .aleks.

John E Clifford

unread,
Jun 5, 2006, 10:40:14 AM6/5/06
to lojba...@lojban.org

--- Jorge Llamb�as <jjlla...@gmail.com> wrote:

> On 6/3/06, John E Clifford
> <cliff...@sbcglobal.net> wrote:
> > As you may recall, my suggestion mirrors the
> > English for "individually" and "collectively"
> (or
> > "together"), attaching as convenient to sumti
> or
> > predicate place (so, I suppose that UI is
> about
> > the only selmaho that will work -- unless we
> > invent a new one).
>
> In English those adverbs normally indicate how
> the predicate
> applies to the subject. For example:
>
> The men carried the pianos together.
>
> would normally mean that each piano was carried
> by all the men,
> not that each man carried all the pianos at
> once. Is that what you
> have in mind, that when the predicate is
> tagged, it indicates how
> it applies to the x1?

Yeah, though not restricted to x1, of course.



> > The point of using {lu'o} and
> > the like is that they would have no use in
> the
> > mildly revised system and so could be used
> for
> > something else -- in this case something
> related,
> > even.
>
> There has been resistance to changing utterly
> useless words, so
> I don't expect a proposal to change something
> not totally useless
> like {lu'o} to succeed. I wouldn't especially
> oppose it, but I know
> others would.

Yes, the accumulated detritus in Lojban is
considerable. I don't really expect that my
suggestion would be accepted (either the one for
{lu'o} or the general idea), but I do have to
offer the opportunity to straighten things up s
bit. And there is always LoCCan.

Nice point; thanks. So, some usage might go by
the way (none in the corpus I keep in searchable
form, but it is now a rather small portion of
what is available). I always rather expected that
the predication mode markers would have to be sui generis.

Jorge Llambías

unread,
Jun 9, 2006, 8:06:22 PM6/9/06
to lojba...@lojban.org
On 6/9/06, Alex Martini <ale...@umich.edu> wrote:
>
> Could someone explain the use of {lo} twice here? I've never seen
> such a thing
> before, and would have expected it to be ungrammatical.

Multiple gadri make more sense with specific {le} than with {lo}.
For example:

so mlatu cu nenri le zdani
"There are nine cats inside the house."

mi viska mu le so mlatu
"I saw five of the nine cats."

ci le mu le so mlatu cu blabi
"Three of the five of the nine cats were white."

And it can be done indefinitely. It's not really very useful, but it's
grammatical.

mu'o mi'e xorxes

Jorge Llambías

unread,
Jun 7, 2006, 7:37:36 PM6/7/06
to lojba...@lojban.org
On 6/7/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> On 6/7/06, Adam D. Lopresto <ad...@pubcrawler.org> wrote:
>
> > mass/team/aggregate/whole, together composed of componets x2, considered
> > jointly." It seems pretty clearly that x1 is a mass, and that x2 is a plural
> > reference.
>
> No, it isn't a plural reference. It's a quite strictly a singular
> reference when it's the expanded form.

If a group is composed of Alice, Betty and Carrie jointly, it does not follow
that the group is composed of Alice jointly. What could that mean?
{gunma} does not mean "x1 has x2 as a member", that's {selcmi}. {gunma}
means "x1 consists of x2".

For example:

le kamni cu gunma la alis jo'u la betis jo'u la karis
"The committee consists of Alice, Betty and Carrie."

It does not follow that the committee consists of Alice.

le kamni cu selcmi la alis .e la betis .e la karis
"The committee has Alice, Betty and Carrie as members."

It does follow that the committee has Alice as member.

The x2 of gunma is (normally) non-distributive and the x2 of selcmi
is (normally) distributive.

I say "normally" because the x2 of {gunma} can be distributive in
another way:

le kamni cu gunma lo nanmu .e ba bo lo ninmu
The committee consisted of men and (later) of women.

In this case, it does follow that the committee consisted of men, and
also that it (at a later time) consisted of women. But it is not distributive
for the men or for the women. The commitee never did consist of each of
the men nor of each of the women.

mu'o mi'e xorxes

Maxim Katcharov

unread,
Jun 8, 2006, 5:25:54 AM6/8/06
to lojba...@lojban.org
On 6/7/06, John E Clifford <cliff...@sbcglobal.net> wrote:
> --- Maxim Katcharov <maxim.k...@gmail.com>
> wrote:
>
> > On 6/6/06, John E Clifford
> > <cliff...@sbcglobal.net> wrote:
> > > Well, I suppose that Alice's relation
> > surrounding
> > > the building (when she is one of the students
> > > surrounding the building)is "participation."
> > I
> >
> > Participation in an event? xorxes already
> > offered this. Consider "the
> > students surround the students". What is Alice
> > participating in?
>
> Well, is Alice among the surrounding or the
> surrounded? Those seem to be the two events in
> which she could participate. In the one case she
> is (more or less) on the outside looking in, in
> the other on the inside looking out.

Sure, I guess. I don't think that this helps much in terms of


explaining it, though. She participates in the wearing of hats too,
after all.

>

It's equally about things involved and what they're involved in. But


in the end, it's the thing that the students compose that does the
surrounding, and not the students themselves. I don't care which one
of

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

lu'o le tadni cu sruri lo dinju

expands to, and I see the difference between the two.

Perhaps you could make this disinction between "distributivity" and


"non-distributivity" in a way that (usefully!) explains what relation
Alice has to the surroundment of the building? The way I see it,

[da poi sruri lo dinju] cu gunma [la alis]
"alice is part of that which surrounds the building" or
"alice is part of the surroundment of the building"

I think that that's perfectly reasonable.

>


> > > Or if I try to specify it in extension,
> > > spelling out how she particpates (standing
> > NEbyN
> > > of the building at the same time as others
> > are
> > > standing at the other points of the compass,
> > say)
> > > you will relate that to being a member of the
> > > group as well.
> >
> > Well, yes. This is the method of participation.
> > For example, I can say
> > "together the three men lifted the piano, by
> > method of one man
> > directing, and two men bearing".
>
> This tells me what each does by way of
> participating, but I still don't see anything
> like a group here unless it is just the fact of
> the perticpation being described in some
> organized way. And that is just what a pluralist
> would mean by "together," more or less.

What lifts the piano? The three men, right? What relation does Avery


have to the lifting of the piano then, if he doesn't lift it himself,
and he's not part of the mass that lifts it? (This is the same thing
as with Alice, and the things that should be noted there should be
noted here.)

There is no real explanation - none that I can think of, and none that


you've provided (correct me if I'm wrong). Yes, I can see how,
intuitively, one may think of it that way, but there are a lot of
things that we sense intuitively that are wrong. In order to
understand it, I need an explanation. This isn't an axiom we're
talking about here. You should be able to explain it. (By it, I mean
my question about Alice.)

>
>

"part" implies being part of something to me. Does it not to you?

> There are


> problably more consitions but this seems to me to
> be the essential one. The "together" of the
> pluralist is just the fact that this pattern
> requires all the students involved (which is
> trivial) and perhaps that with many fewer
> students similar patterns (that formed closed
> loops arounf the building) are not possible --
> certainly that no one student can form such a
> pattern. Does the groupiness consist of anything
> other than this? You've already said it is not a
> thing over and above the students, so that the

You have the two "over and above"s confused. One refers to a thing


that something is by nature of being what it is (a dog is an animal,

since dogs are animals by nature). The other refers to things that are


aggeregates of other things. "Forest" being over and above "trees" is
different from "animal" being over and above "dog". You can say
"animal" when you have a dog in mind, but you can't say "forest" (or
"grove") unless you have a forest (or grove) in mind - which you
usually do, assuming that there are more than 6-10 trees.

It's not very often that someone gives you a specific answer when more


than X things are involved (X being perhaps 10ish). "What's going on
there?" "Some kids are carrying a bunch of chairs to the garden". But
"some kids" is clearly some sort of special plural predication, since
you don't mention the words "mass" or "group"! No, it isn't. The
average human will think of, say, a group of 20 kids massively, and
won't actually summon-to-mind 20 instances of "kid". "Some kids" in
this case refers to a mass of kids.

> students form a pattern seems to be the most


> obvious next choice. But that, of course, means
> that for reality, it just says what the pluralist
> says but in differnt words. If it is something
> else, that you need to say what and demonstrate
> that it really is there. It seems that the
> pluralist says "there are these students and they
> form this pattern" and the singularist says
> "there is this pattern and the students for it"
> Why this stife there be/'twixt Tweedle-Dum and
> Tweedle-Dee?

My position is that you need an at least implicit group/mass, so that


you can expand (i.e. explain using more axiomic terms) things like
"lu'o" or "loi".

You seem to be contrary to this.

> > > >


> > > > Elaborate? To me, "among" has implications
> > of
> > > > being "among a group such that".
>
> Well, of course it would; you are a believing
> singularist. For a pluralist, "x is among y"
> just means that x is one of the ys.

x is a referent of ys. Yes. But even for a pluralist, Alice is also a


referent of/among "the students (wearing hats)". Again, this doesn't
say anything of the difference between distributive/non-distributive.

>


> > > And so it does -- when used by a singularist.
> > > When used by a pluralist, it doesn't. But
> > the
> > > properties of "among" are the same for both.
> >
> > But in the pluralist view, there's still a
> > group there, you just don't
> > choose to acknowledge it, right?
>
> Where? Go through the whole pluralist semantics
> and nothing like a group turns up, just things,
> one or several as the case may be. At the end of
> it all, it is hard to say where the
> unacknowledged group might be.

Ok, then use these pluralist semantics to (usefully!) explain the


relationship between Alice and "surrounds a building", as opposed to
Alice and "wear hats".

> What you just responded to wasn't so much an


> argument as a challenge.
> Fact is, explanations of how "bunch-together"
> differs from "mass"
> aren't really available. I attribute this to
> there being no
> explanation of "bunch-together" that is different
> from "mass".>>
>
> I attribute it to the fact that there is no
> difference except verbiage. You seem to think
> that the mass form the explanation is right and
> the the other wrong, which is odd if they are the
> same explanation. However, this is all empty,
> since we have neither explanation at hand yet (I
> have tried to suply one but I don't know whether
> you will buy it).

They're not the same explanation. One says that there is no mass:

Go through the whole pluralist semantics


and nothing like a group turns up, just things,
one or several as the case may be.

Right? So, no, not the same.

> > > > > Ok, then when I say "group of students",


> I
> > > too
> > > > > am "referring to many things".>>
>
> I agree, but you seem to think that you are
> actually referring to one thing, the group. At
> least you talk that way.

Yes. That, or I'm saying "the students are [part of a group such that


that group surrounds the building]". Doesn't matter which, but both
involve "group"/"mass".

> > collective predication. Even {loi} does not


> > appear to be just collective predication -- it
> > seems clearly to involve corporate and Urgoo
> > cases as well. And there are cases which
> cannot
> > be dealt with using gadri.>>
>
> Examples? I see no practical differences between
> corporate masses and
> regular masses, and I'm not familiar with Urgoo
> cases at all. >>
>
> Corporate masses (I don't much like that
> terminology since it sugtgests more similarity
> than I think justified)continue to be the same
> even with a change of components; they also
> inherit properties from their components
> directly: if a component (acting as such) does or
> is something, the corporation does or is, too.

This seems to have more to do with the details of when a person stops


calling something a mass. I don't think that we need to categorize
every mass into a certain type of mass in order to use them.

As for doing something, and having the corporation do it too, this too
I think has to do with unimportant details. It's not a fixed rule. To
give an example (that's perhaps more similar than justified), if a
salesperson makes a sale, the corporation makes a sale, but if that
salesperson gets the flu, the corporation doesn't have the flu. It's
not perfecly certain how each thing works out.

These things are details, and are not critical to the concept of a mass.

> Corporations also have properties in which some


> components do not participate. I suppose there
> are other charateristics but these are enough to
> separate then from ordinary (collective
> predication) masses. Urgoo is the stuff of which
> some kind of thing is made: all dogs are chunks
> of Dog, for example -- as are dog organs and the
> mixture that results from a steamroller rolling
> over a pack of dogs. This is an actual mass-noun
> concept.

Well, if your conception of "dog" extends to that, then sure. For me,


something stops being a dog when it gets rolled over - it becomes
"paste formed from a dog corpse". I'd still say "that dog has been

squished", or "we'll bury the dog", but it would be in the sense of


{lo pu gerku}, and not {lo ca gerku}.

> So far as I can tell, Urgoo is like


> corporations in some respects: it remains the
> same even if its representations change, it
> inherits propeties from its manifestations. It
> differs in that it is homogenous, does not have
> components, although the manifestations play a
> somewhat similar role, but an Urgoo can exist
> without any manifestations at all.

This is like an ideal mental form of something, that all things that


are it are composed of: {loi ro gerku}, or something of the sort.

> I think these


> two are enoguh different to justify some separate
> consideration but both have been folded into the
> muddle that is CLL mass.

I don't think that the distinction of corporate vs. non-corporate


entities needs to be made on such a raw level.

>


> "Mass"/"together" expands to "x1 is a mass with
> components x2". This
> is an actual relation. I consider that as
> significant in terms of
> content as you can get.>>
>
> But you offer no evidence that it applies here.
> "Together" is a real situation as well and I have
> offered an explanation of what it means in
> different terms. What does "is a mass composed
> of" mean in different, neutral, terms. Failing
> that we are just talking by one another, since we
> are using language radically differently.

The evidence is a sensible explanation of what "the students surround
the building" means: "the students are part of a group that surrounds
the building".

Is that wrong? How is it wrong?

How is "the students surround the building" different from "the group


of students surrounds the building"? Actually different, and not in
terms of English frames or English pragmatics.

>


> <<> They, on the other hand,
> > would find oit odd that you cannot understand
> > such a straightforward English expression as
> "the
> > students" (especially since you seem to
> > understand the mysterious "the mass of
> students").
>
> It's about as mysterious as "the building for
> students" - that is, not
> mysterious at all. "the students", on the other
> hand, is ambiguous: it
> can refer as in "the students wore hats" or "the
> students (as a mass)
> surrounded the building", and then, of course,
> there's also "the
> students (as a bunch-together) surrounded the
> building", which nobody
> has really explained or demonstrated as being
> different from "as a
> mass", though copious flat assertions of the sort
> have been made.>>
>
> But you, of course, have nowhere demonstrated
> that "as a mass" is different from "together"
> nor explained what it meant. You have asserted

I've explained many times what it meant.

"together the students surround the building" :

X is a mass, and each student is a component part of that mass

X surrounds the building

the students are part of a mass such that surrounds the building

I don't think that I need to prove that such a thing as "a
relationship between certain things" exists.

> it is superior, but that is just your say-so. On


> the other hand, if you really believe, as you
> seem to be saying here, that the two expressions
> mean the same thing, what is the argument all
> about?
>
>
> <<> Note that, if you do write pages explaining
> the
> > differnce, the pluralist can take it, make a
> few
> > uniform changes and provide you with the
> > explanation you want for the difference between
> > "the students individually" and "the students
> > together."
>
> Please, do it then! Do it with the crude
> paragraphs I've offered. What
> are you arguing this with me for, when simply
> demonstrating this would
> solve everything?>>
>
> Gladly. Pleas provide the explanation for the
> mass-talk form. Note, this will require saying
> it without assuming masses or giving a fairly
> complete formal system for masses.

A mass is a relationship like any other. Do you deny that such a


relationship ("x1 is a mass/aggregate/composition of x2 / x2 is a
component part of x1") exists? Do you deny that such a thing as a
(predicate) relationship exists?

I'm using perfectly established structures, in English or otherwise,


to do the explaining of how Alice relates. Predicate relationships,
and the idea of "x1 is a composite of component parts x2". Both are
established in both our minds, right?


You did say "Gladly". Could you now do this?

> > You will no doubt take it that way; how are you


> > sure the speaker meant it that way or even that
> > he can sense the difference?
>
> Uh, because "bunch" doesn't have the definition
> that we've assigned it
> (for the sole purposes of this argument) in
> common use. Bunch is
> simply "group", with implications of the things
> being close together -
> "bunch of twigs", etc.>>
>
> Well, it does seem to have that meaning in my
> dialect. That is, when I say "a bunch of things"
> I am not implying that there is anything other
> than those things there (not even necessarily
> close together).

As I've said, you make it seem like I'm bringing in the concept of a


baboon to explain away this thing.

You really aren't bringing in anything new, because your mind probably


has "mass of 20 students" 'loaded' (though of course, the pragmatic
implications of "mass" and all the 'framing' that "mass" entails are
not loaded), because humans don't usually like to 'load' each of 20
students when they don't have to. "Mass of students, 20 component
parts" is good enough for most people.

> I presumably have some reason


> for dealing with them together but that is
> nothing "out there" called "bunch," it is just
> how I am dealing with them.

So there isn't anything out there? Or just nothing out there called


bunch? Because if there isn't anything out there, I can't imagine you
explaining what Alice's relationship is. With humor, I imagine
something like:

Alice is ? ? ?


? surrounds the building.
where "?" stands for "magic happens here".

Well, it's not magic. The "rational explanation", if you will, is that


Alice is part of a mass/group, the mass/group that surrounds the
building. If you have a different rational explanation, then please
offer it.

>
>


> <<If they said "the students surrounded the
> rope", then you might have
> an argument as to how it's meant. But if we say
> "the group of students
> surrounded the rope", then it's clear that we
> mean the *group* (of
> students), and not anything else.>>
>
> Not clear at all, since I don't see any group
> there, just students.

When I say "the *group* of students", you can't imagine a group?

> If you mean "the group of


> students" to say, in different words, just what
> "the students together" says -- that is, without

No, "the students together" in your mind for some reason can't have


the same meaning as it does for me and the dictionary (1. In or into a
single group, mass, or place), it seems. We don't say the same things,
because your variant excludes any possibility of "mass" in order to
describe the relationship (it seems).

> "The 50 students surrounded the building" and


> "the group of 50
> students surrounded the building" are synonymous
> in meaning. It's just
> that one of them uses the word "group", which
> invokes a certain frame
> in your mind that the omission of the word
> wouldn't.>>
>
> Then what the Hell is this argument about? One
> person talks one way, the other the other, as
> their taste leads them. And, of course, that is
> just what the formalism says: whether you give a
> pluralist or singularist interpretation to the
> system, the logic is the same.

Ok, then you should have no problem telling me:

In "the students surround the building", Alice is part of the
mass/group that surrounds the building. Of course, we don't have the
same imagery invoked in our mind as is typically associated with
English "mass" or "group", but yes, it's a mass/group regardless. So

when I say "the students surrounded the building", I mean that Alice


is part of a mass such that surrounds the building. Same goes for
Bryce, Carol, David, etc.

Right?

>


> > Forst are
> > just trees after all (with some exceptions like
> > willow forests which are apparently just one
> > tree). (I don't of course, really mean this. I
> > am just pointing out how useless taking what
> > someone says is in figuring out which of the
> > identical sides they are on.
>
> A forest is not the same thing as a set/"bunch"
> of trees, just as a
> human is not just a set/"bunch" of organs... just
> as a crowd
> surrounding a building is not just a set/"bunch"
> of students.>>
>
> And the difference is...? I suppose it is
> somethign that hold them all together, a common
> interest them. that is something about us
> usually, although it is often helped by
> propinquity and short-chain causation and the
> like.

The difference between what? A tree and a forest? A person and a crowd?

> > More than that too, an organism. That is, the


> > organs in an organization. Without the
> > organization, the organs are just a pile of
> > specimens.
>
> That's what I mean when I say mass. I discussed
> this earlier using the
> example of a piece of graphite and a piece of
> wood not quite being a
> pencil. Search for the term "graphite" if you're
> interested.>>
>
> Ah, that was the point of that story. It was not
> very clear to me at the time. Your use of the
> term "mass" is adding yet another meaning to that
> already overworked word; can we find another word
> for you concept.

It's not an overworked concept in the same way that "animal" is not an


overworked concept, because there are so many types of animals.

Water, which can be combined or reduced into more water is a mass. I'm


a mass. Just about everything is a mass. A crowd that makes noise is a
mass, even if some people in it are quiet. Or even if all people in it
are (relatively) quiet.

If you want to taxonomize and label all of these different things that


a mass can exhibit, or when certain masses stop being masses, or if
you can combine certain masses to form a mass that is considered to be
the same thing as the two masses were, go ahead. But it's still a
mass.

> But in any case, I don't see


> how this helps with the students: they do not
> compose and organism or an organic whole, and
> maybe not even an organization.

If 1000 people together do not compose a "crowd", then what is a


"crowd"? Just a way to refer to the 1000 conceptualizations of people
that you have "loaded" into your mind? Even if the crowd starts doing
things that none of the people do on their own?

> They each fall


> into a place in a pattern which we are taking as
> significant and by virtue of which say they are
> togethre. Is it also by virtue of this that we
> say they are a mass? If not, what is involved?
> If so, why are we having this argument (or, more
> accurately, what the Hell are we arguing about,
> sinc we see to agree on everything except what
> words to use and that is merely a matter os style
> and not open to argumentation).

Well, at first you seemed to deny that the concept of mass was used in


plural predication, but now you seem to deny that the concept of mass
(or group) exists at all. So that's what we seem to be arguing about.

> > Set theory, which seems to be the model for


> talk
> > of masses,
>
> A mass is a relationship, it need not have
> anything to do with set
> theory. x1 is a mass of composite parts x2.>>
>
> Huh!? There is a relationship of composition that
> defines a mass, but a mass is not a relationship
> (notice, by the way, that {gunma} is not a mass
> of the sort you descibed earlier).

Is not mass of what sort?

> It may also be


> that the fact that things stand in a certain
> relationship to one another is what gets them
> into the mass, but the mass is not that
> relationship either.

All (?) things (physical things, especially) are masses. Maybe the


strings of string theory aren't a mass, but everything else, we've
found it to be a mass. Tiny things, arranged pencil-wise, form a
pencil. Can everything be broken down into something else? Is
everything composed of something else? Yes.

The argument isn't really about this (I hope). It's about whether or


not an entity with parts:students can exist. I say that it can, and
frequently does. And I also say that this entity is the thing that
surrounds the building.

> Ok, then if it's not connected to the act of


> "surrounding the
> building" by way of a group, then how is it
> connected? What is the
> relation?>>
>
> Directly by each of them taking a place in a
> pattern which constitutes surrounding the
> building. You may call "taking a place" "forming
> a group" but there is no necessity in doing so.
>

Ok, sure, that's another sensible way to think of it.

[da poi sruri lo dinju] cu morna [la alis]

x3 can even be "surrounding-the-building-wise".

>


> <<> Of course,
> > you can mean that equally well using "the group
> > of students," but it is harder to see. And, by
> > parity of reasoning (since the two are formally
> > identical) "the students" does refer to a
> group,
> > if you want to go that way, although it is
> clearer
> > if you say "the group of students."
> >
>
> What are formally identical? Thinking of them as
> a group and not
> thinking of them as a group?>>
>
> Well, thinking of them as a group and thinking of
> them as acting together.
>

Sure. So

[da poi sruri lo dinju] cu gunma [la alis]

is a correct/complete way to express your pluralist "lo tadni cu sruri
lo dinju", right?

Jorge Llambías

unread,
Jun 11, 2006, 9:50:50 AM6/11/06
to loj...@yahoogroups.com
On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> On 6/10/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > An outermost quantifier is distributive,
> > an outermost mass-marker is non-distributive. It's that simple.
>
> No doubt it's a simple rule, but the question is if it's the proper
> way to describe how we communicate.

It's a proper way to describe how I understand Lojban works.


That's all I can offer you.

> > la .alis .e la betis = ro le re prenu


> > la .alis joi la betis = lei re prenu

> > la .alis jo'u la betis = le re prenu
>

> Are you offering these definitions as suggestions, or
> as explanations of how it really is?

As explanations of how I understand them to work in Lojban.

> So you would disagree that


>
> loi tadni cu sruri lo dinju
>
> expands to
>
> [da poi sruri lo dinju] cu gunma [lo tadni]
>
> which you had previously described as a legitimate interpretation,
> yes?

Using your place structure for {gunma} with distributive x2, I


would disagree, yes. I agreed when I thought we were using
{gunma} with the same meaning.

> {loi tadni cu sruri lo skori} and {ro lo tadni cu


> sruri lo skori} are different in some way, yes?

Yes.

> I'm asking you to explain the difference.

I've already already done that. (And besides you understand
the difference perfectly well.)

> Alice is part of X


> X surrounds the building
>
> Alice wears a hat
>
> Is the difference not apparent there?

Yes. In the pluralist view one would say instead:

Alice is one of X
X surround the building

In the singularist approach, X is a single thing, distinct from the


things that surround the building and consisting of them, while
in the plural approach X are many things, just those very things
that surround the building.

mu'o mi'e xorxes

Maxim Katcharov

unread,
Jun 8, 2006, 6:31:14 AM6/8/06
to lojba...@lojban.org
On 6/7/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 6/7/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > On 6/7/06, Adam D. Lopresto <ad...@pubcrawler.org> wrote:
> >
> > > mass/team/aggregate/whole, together composed of componets x2, considered
> > > jointly." It seems pretty clearly that x1 is a mass, and that x2 is a plural
> > > reference.
> >
> > No, it isn't a plural reference. It's a quite strictly a singular
> > reference when it's the expanded form.
>
> If a group is composed of Alice, Betty and Carrie jointly, it does not follow
> that the group is composed of Alice jointly. What could that mean?

It doesn't mean anything, I don't think that it's the right way to say
it. Well, no - if that means that the group is composed of Alice, with
her seen as a component part, then that seems fine. If X is a house
with ...'innards' A B C (considered as things inside; considered
insidely), then the container has inside it Alice (considered
insidely).

(Note is that the components of a group are 'joint'/together by the
very nature of them being in a group - "jointly" is pretty much
redundant, in the same way that mass / group / aggregate / whole /
etc. are redundant.)

To be quite first-orderish about this:

Alice is a component part of X
Betty is a component part of X
Carrie is a component part of X

X is a "whole" (a mass/aggregate/composite entity/etc.), such that its
parts are considered jointly/together/etc. (in the same way graphite
and wood are only a pencil when they're considered jointly).

> {gunma} does not mean "x1 has x2 as a member", that's {selcmi}.

I don't know if member is the right word, even for cmima, because it
has implications of something like a club, staying the same even when
the members change radically.

The thing with cmima is that it mentions "set", but I suppose that
it's not in the sense of a mathematical set, so it might be better,
yes.

There are several other relationships that are quite close:

cmima: x1 is a member/element of set x2; x1 belongs to group x2; x1 is
amid/among/amongst group x2
gunma: x1 is a mass/team/aggregate/whole, together composed of
components x2, considered jointly


pagbu: x1 is a part/component/piece/portion/segment of x2 [where x2 is
a whole/mass];

What's the difference between all of these? My take on it is that


{cmima} concerns things seen more seperately - a squadron of planes,
family of bears, [...]. {gunma} would concern things that look like
they're quite close together - a pencil, a book, a car, a body.
{pagbu} would be the word to use when you don't care to detail if you
see them dispersed or visibly combined.

But it doesn't really matter to me which exact interpretations are


given: all of these have the same format - there's one aggregate, and

component parts of it. (Or possibley two aggregates - perhaps the
students can be of two different groups/masses and we want to
predicate things about both of them - but this is very very rarely the
case.)

> {gunma} means "x1 consists of x2".
>

I'd disagree with that. It's a "mass/team/aggregate/whole, together


composed of components...". "Consists" is a special term, it has very
specific pragmatics attached to it.

> For example:


>
> le kamni cu gunma la alis jo'u la betis jo'u la karis
> "The committee consists of Alice, Betty and Carrie."
>
> It does not follow that the committee consists of Alice.

"Consists" implies "consists only of", "includes" implies "includes
but is not limited to". If you want to translate using the full and
proper implications of "consists", you'd add {po'o}, in which case I'd
agree with you - if the committee consists of po'o Alice, Betty and
Carrie, it does not follow that the committee consists of po'o Alice.

"Consists" is not the proper word to use. "includes" might be it, but
it's best to go with at *least* all the gloss terms given: "is a
mass/team/aggregate/whole, together composed of components..."

>
> le kamni cu selcmi la alis .e la betis .e la karis
> "The committee has Alice, Betty and Carrie as members."
>
> It does follow that the committee has Alice as member.
>
> The x2 of gunma is (normally) non-distributive and the x2 of selcmi

So... non-distributivity would expand to that group expansion that I
gave, but with {po'o} thrown in? I don't think that that's right.

John E Clifford

unread,
Jun 5, 2006, 3:24:19 PM6/5/06
to lojba...@lojban.org
Very nice. But, since the terms might be
differently predicated on the two brivla, we
would have to place the markers on the predicate
not the term. This seems to mean that these

markers would themselves be strung out like terms
with the predicate, to be properly correllated
with the terms in the term list, probably
something like (using d and c for distrib and
collect)
ko'a ko'e ko'i ko'o ko'u broda d c c d c * brode
c d d c d (* because I suspect the conjunction
at this point will change since we are connective
bridi tails not selbri). Not yet a very comely
system, though it would be rare that we would
need to show all the markers. Still, it does
circumvent the problem you not (and, no doubt
creates other for the parser or interpreter).

--- Jorge Llamb�as <jjlla...@gmail.com> wrote:

> On 6/5/06, John E Clifford


> <cliff...@sbcglobal.net> wrote:
> >
> > > > As you may recall, my suggestion mirrors
> the
> > > > English for "individually" and
> "collectively" (or
> > > > "together"), attaching as convenient to
> sumti or
> > > > predicate place (so, I suppose that UI is
> about
> > > > the only selmaho that will work -- unless
> we
> > > > invent a new one).
>

> Here is a mathematical proof that it cannot be
> done with a
> single word. Consider the sentence:
>
> ko'a ko'e ko'i ko'o ko'u broda gi'e brode
>
> It consists of eight words, so there are
> exactly nine places where
> a word meaning "together" could be added.
>
> But there are ten argument places it could be
> modifying:
> each of the five slots of broda and each of the
> five slots of brode.
>
> Therefore, a single word, no matter how
> strategically well placed, could
> not be used to do the job of unambiguously
> marking any sumti slot
> as collective.

Jorge Llambías

unread,
Jun 8, 2006, 11:32:53 AM6/8/06
to lojba...@lojban.org
On 6/8/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> On 6/8/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
>
> > The building is surrounded by students and professors together. Then:
> >
> > loi tadni cu sruri le dinju
> > Students surround the building.
> >
> > is false.
>
> No, true.

Are you sure? Is "Alice surrounds the building" true? Is "Alice and Betty


surround the building" true? Suppose Alice and Betty are the only students
in the group, and the rest of the people surrounding the building are
professors,
is "Students surround the building" true?

> > {gunma} does mean "consists of". But even if you think {gunma} means


> > something else, the predicate that you need to make your expansion
> > of {loi} is one that means "x1 consists of x2", not one that means
> > "x1 has component x2"
>
> Ah. No, I don't need it to make my expansion.

I think you do.

mu'o mi'e xorxes

Jorge Llambías

unread,
Jun 8, 2006, 9:43:23 AM6/8/06
to lojba...@lojban.org
On 6/8/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:

> My take on it is that
> {cmima} concerns things seen more seperately - a squadron of planes,
> family of bears, [...]. {gunma} would concern things that look like
> they're quite close together - a pencil, a book, a car, a body.
> {pagbu} would be the word to use when you don't care to detail if you
> see them dispersed or visibly combined.

If your take on {gunma} is that the x2 is distributive, then you cannot


expand {loi tadni} as {lo gunma be lo tadni}. Consider this case:

The building is surrounded by students and professors together. Then:

loi tadni cu sruri le dinju
Students surround the building.

is false.

lo gunma be lo tadni cu sruri le dinju


A group that has students as components surrounds the building.

would be true.

da poi sruri le dinju cu gunma lo tadni


Something which surrounds the building has students as components.

would also be true.

So you must either take the x2 of gunma to be non-distributive, or you


need a different expansion for {loi tadni}.

> But it doesn't really matter to me which exact interpretations are
> given: all of these have the same format - there's one aggregate, and
> component parts of it.

{pagbu} and {cmima} are both like that, yes. If A and B and C are parts of D,


then A by itself is a part of D, and B by itself is a part of D, and C by itself
is a part of D. If A and B and C are members of D, then A by itself is a
member of D, and B by itself is a member of D, and C by itself is a member
of D. That's what we mean by saying that the x1 of {pagbu} and the x1 of
{cmima} are distributive.

But {se gunma] is different: if A and B and C conform D, then A by itself
does not conform D, B by itself does not conform D and C by itself does
not conform D. Only together, jointly, do A, B and C conform D.

{gunma} means "x1 consists of x2".
{se gunma} means "x1 conform x2".


> I'd disagree with that. It's a "mass/team/aggregate/whole, together
> composed of components...". "Consists" is a special term, it has very
> specific pragmatics attached to it.

{gunma} does mean "consists of". But even if you think {gunma} means


something else, the predicate that you need to make your expansion
of {loi} is one that means "x1 consists of x2", not one that means
"x1 has component x2"

mu'o mi'e xorxes

Jorge Llambías

unread,
Jun 9, 2006, 8:49:20 AM6/9/06
to lojba...@lojban.org
On 6/8/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> On 6/8/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > Suppose Alice and Betty are the only students
> > in the group, and the rest of the people surrounding the building are
> > professors, is "Students surround the building" true?
>
> Yes, it becomes much more apparent (to me, at least) once written out
> in Lojban. It's not {la alis cu sruri lo dinju}, it's {lu'o la alis cu
> sruri dinju}.

My understanding of our disagreement so far had been as follows:

(1) loi tadni cu sruri le dinju
(2) lu'o lo tadni cu sruri le dinju
(3) lo tadni cu sruri le dinju

We both agreed (I thought) in which situations (1) would be true and
in which situations it would be false. We both agreed that {loi} could
be decomposed as {lu'o lo}, so that (2) says the same as (1). We
disagreed on whether the {lu'o} component could be optionally omitted.
I held that {lu'o} was optional, and so could be dropped when context
made it unnecessary, as in the example of students surrounding a
building, and you held that {lu'o} could not be dropped because if it
was dropped the predicate {sruri} had to be necessarily taken as
distributive.

But now the disagreement seems to occur at an earlier step.
We no longer seem to agree about the meaning of (1). For example,
if there is a chain of rocks around the building, with a gap filled by
Alice, who is a student, you would say that (1) is true and I would
say it is false. For me it is false because it is not the case that
students surround the building in that case. For you it is true because
it is the case that a group of things that includes at least one student
surrounds the building.

Do we at least agree on what we disagree about?

Jorge Llambías

unread,
Jul 10, 2006, 9:27:54 PM7/10/06
to loj...@yahoogroups.com
On 7/10/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
>
> What's a plural variable?

A variable that can have more than one referent at the same time.

> I point out that again, you're not offering any sort of explanation of
> 'non-distributive', you're simply assuming that it exists.

All right then, we'll just have to leave it at that. I'm afraid my pedagogical
resources are exhausted.

Maxim Katcharov

unread,
Jul 10, 2006, 9:34:07 PM7/10/06
to lojba...@lojban.org
On 7/10/06, John E Clifford <cliff...@sbcglobal.net> wrote:
> Boy I dropped this one too soon. This is all very clear and thus easily answered. Suppose in the
> group surrounding the building there are some non-students and suppose further that, if all the
> students were in their same places and these non-students were absent (and not replaced by
> others), then the building would not be surrounded. In this case is {loi tadni cu sruri le
> dinju}true? NO! Since the building was not surrounded but for the non-students, the students did
> not surround the building *though they participated in the surrounding and may even have been the
> main force in it -- and might even have been able to surround the building by moving to new
> arrangements).
>
> So, is {loi tadni cu sruri le dinju} equivalent to {da poi sruri le dinju cu gunma lo tadni}?
> Assuming that {gunma2) need not be a complete list of the members (whatever) of the mass (that {se
> gunma} means "is a member of mass..." not "are all the members of mass...") then NO! again, ssince
> the latter allows that others might be essentially involved. (I skip over xorxes' worries about
> admitting that there are such things as masses; they make no detectable difference). The {loi}
> form is apparently equivalent to {da poi srur le dinju cu gunma lo tadni po'o] however (much as I
> hate sing {po'o}).

Yes. My first June 11th message addressed this:

<
Sure. I'd be perfectly happy to say that

loi tadni cu sruri lo dinju

expands to

[da poi sruri lo dinju] cu gunma [[lo tadni] po'o]

for the purposes of this discussion, since really, the discussion
isn't much affected by it.
>

This is beside the point of this discussion.

In this discussion, I ask you to provide an explanation of how your
plural predication /could/ work, using ideas that we all agree on.
I've done it for my explanation of how things "could" work using
gunma/mass/one-thing-that-is, and I'd like you to do the same for
yours. Everything I've used to explain myself we can all agree on.

Or perhaps you could tell me why you think that it what you offer is
sufficient? It's as if we're two primitive people looking at a car. I
say that it must be water and fire that moves it ("I hear water noises
when we push it around, and smoke comes out the back when it moves"),
and you say that it works through "the force of Dutablibi". That's
appealing, but it would be a lot better if you told me how this all
happens using things that we both agree on.

Maxim Katcharov

unread,
Jul 10, 2006, 9:12:48 PM7/10/06
to lojba...@lojban.org
On 7/10/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 7/10/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > Has there been any progress on a response? My last message was both
> > short and lucid; I imagine that the responding silence indicates
> > concession, though I doubt that this is the case.
>
> It only indicates that I didn't see us getting anywhere. But let's pick up where
> we left if you want.

>
>
> > On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > > On 6/11/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > > > On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > > > > The surest way to show that I'm a fool for asking this 30th time
> > > > > is to point me to an explanation that I haven't rightly shown to
> > > > > be unexplanatory.
> > > >
> > > > I don't think you are a fool.
> > > >
> > > > Explanations are pointless at this point, because definitions cannot
> > > > be right or wrong. We are now working with different definitions.
> > >
> > > I don't think that this is the issue.
>
> I think it is. In particular, we seem to have different understandings
> of what "distributivity" means. Do you accept that some relationships
> can be non-distributive in one ore more of their arguments? If you don't
> agree, if you can only accept distributive relationships, then no matter
> how much we argue, we won't get anywhere.

This isn't a matter of definition, but a matter of a difference in
understanding how this 'distributivity' works. Simple predication is
understood by everyone, simple predication is how I understand your
plural-predication to work at its core. You offer a different
explanation, but because you don't show how it /could/ work, much less
that it /is/ the way that it works, I have severe doubts about your
position.

>
>
> > > For example, both of us understand and agree with a 1to1 relationship:
> > >
> > > Alice is inside the school
>

> I wouldn't call that a 1to1 relationship. That for me is a predication.
> The relationship in that predication is "...is inside of..." and it is not
> necessarily 1to1. For example, there may be other people inside the
> school, and Alice may be inside some other thing (a classroom, for
> example) when she is inside the school. But that's probably irrelevant
> to what we were discussing.

The predicate of "Alice is inside the school" is "is inside the
school". By referring to it as a relationship (which is what our sort
of predication is, I think), I sought to avoid confusion. What I was
getting at by predication/relationship was in fact "...is inside
(of)...".

In this example, the relationship /is/ 1to1. There is one Alice, and
one school. There may be other things that Alice is within, but I
don't think that 1to1 should imply that I mean exclusively 1to1,
simply that in /that/ sentence, the relationship between two things is
being specified.

>
> > > and both of us understand and agree with the basic plurally
> > > predicative relationship
> > >
> > > the 26 students are inside the school >>
> > > Alice is inside the school
> > > Bryce is inside the school
> > > [...]
> > > Zoe is inside the school
>

> We agree that "...is inside of..." can be distributive, yes.


>
> > > and both of us understand and agree with the "mass" relationship
> > >
> > > the graphite and the wood are component parts of the pencil >>
> > > the graphite is a component part of the pencil
> > > the wood is a component part of the pencil
>

> There is nothing new there. "...is a component part of..." is just as
> distributive as "...is inside of...". Nothing especially "mass" about it.
> It is not an example of a non-distributive relation. There is no
> reason to call "...is a component part of..." a mass relationship.

I'm not talking about the distributivity of those examples. I'm
talking about the semantic meaning of "is a component of".

There is a reason to call "...is a component part of..." a mass
relationship. That's what a "mass" relationship is. We're not talking
about a "glob" or "amount" when we say "mass". We're talking about a
certain thing that is composed of aggregate things. Or what does
"mass" mean to you?

>
>
> > > I don't see your position as equally sensible. If I were to say "ok,
> > > there's no mass involved", I would have nothing like this to rely on.
> > > I would have
> > >
> > > Alice is a referent of X
> > > X surround the building < axiomic explanation of this is needed
>

> As written, that doesn't make sense. If Alice is a referent of X, then
> X has to be something like a word. But words don't surround buildings,
> so X could not surround the building.

This is the closest I can come to a sensible explanation of your view.
This is the form of explanation that I'm looking for. So far, you've
offered "it just is - in 'the students surround the building', isn't
it clear that they do it non-distributively?", and that isn't an
explanation. In that case it is clear to /me/ that they are component
parts of a single mental entity that surrounds the building - and I've
shown how the mechanics that we both agree on can allow for this (but
not that this /is/ how things work).

>
> What we want to say is
>
> Alice is one of X
> X surround the building.

What do you mean by "one of"? That she is one of the students?

>
> where "X" is a plural variable, i.e. a variable with (possibly) more than one

What's a plural variable?

> referent, and where "...surround..." is a relation that can be non-distributive

I point out that again, you're not offering any sort of explanation of
'non-distributive', you're simply assuming that it exists.

> with respect to its first argument (also with respect to the second, but that's
> not relevant in this example).


>
>
> > > > For me {loi tadni} means "students", just like {lo tadni}, and
> > > > the mass gadri in addition indicates that whatever is predicated
> > > > of the students is predicated non-distributively. For you it means
> > > > "a group of things that includes students among its members",
> > > > which is something quite different.
> > >
> > > If you'd like, it can be "an entity composed of students", it doesn't
> > > really matter. And yes, this would be quite different, because it
> > > treats the students together as a different entity than each of the
> > > students themselves.
>

> Quite. You only accept referring to single entities, either to the single
> entity that consists of all the students, or to each of the students at
> a time. You won't admit a variable that has several referents, the students
> themselves not a new entity consisting of the students, which is what
> we need for plural predication.
>

You say this as if there was something to admit to. You haven't
provided an explanation, and my understanding works perfectly well
without needing a special "plural referent", so I don't need to fall
back on "hmm... well, this is the only option, so I guess it makes
sense".

John E Clifford

unread,
Jul 10, 2006, 5:44:22 PM7/10/06
to lojba...@lojban.org
--- Jorge Llamb�as <jjlla...@gmail.com> wrote:

> On 7/10/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > Has there been any progress on a response? My last message was both
> > short and lucid; I imagine that the responding silence indicates
> > concession, though I doubt that this is the case.
>
> It only indicates that I didn't see us getting anywhere. But let's pick up where
> we left if you want.
>
>
> > On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > > On 6/11/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > > > On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > > > > The surest way to show that I'm a fool for asking this 30th time
> > > > > is to point me to an explanation that I haven't rightly shown to
> > > > > be unexplanatory.
> > > >
> > > > I don't think you are a fool.
> > > >
> > > > Explanations are pointless at this point, because definitions cannot
> > > > be right or wrong. We are now working with different definitions.
> > >
> > > I don't think that this is the issue.
>
> I think it is. In particular, we seem to have different understandings
> of what "distributivity" means. Do you accept that some relationships
> can be non-distributive in one ore more of their arguments? If you don't
> agree, if you can only accept distributive relationships, then no matter
> how much we argue, we won't get anywhere.
>
>

> > > For example, both of us understand and agree with a 1to1 relationship:
> > >
> > > Alice is inside the school
>
> I wouldn't call that a 1to1 relationship. That for me is a predication.
> The relationship in that predication is "...is inside of..." and it is not
> necessarily 1to1. For example, there may be other people inside the
> school, and Alice may be inside some other thing (a classroom, for
> example) when she is inside the school. But that's probably irrelevant
> to what we were discussing.
>

> > > and both of us understand and agree with the basic plurally
> > > predicative relationship
> > >
> > > the 26 students are inside the school >>
> > > Alice is inside the school
> > > Bryce is inside the school
> > > [...]
> > > Zoe is inside the school
>
> We agree that "...is inside of..." can be distributive, yes.
>
> > > and both of us understand and agree with the "mass" relationship
> > >
> > > the graphite and the wood are component parts of the pencil >>
> > > the graphite is a component part of the pencil
> > > the wood is a component part of the pencil
>
> There is nothing new there. "...is a component part of..." is just as
> distributive as "...is inside of...". Nothing especially "mass" about it.
> It is not an example of a non-distributive relation. There is no
> reason to call "...is a component part of..." a mass relationship.
>
>

> > > I don't see your position as equally sensible. If I were to say "ok,
> > > there's no mass involved", I would have nothing like this to rely on.
> > > I would have
> > >
> > > Alice is a referent of X
> > > X surround the building < axiomic explanation of this is needed
>
> As written, that doesn't make sense. If Alice is a referent of X, then
> X has to be something like a word. But words don't surround buildings,
> so X could not surround the building.
>

> What we want to say is
>
> Alice is one of X
> X surround the building.
>

> where "X" is a plural variable, i.e. a variable with (possibly) more than one

> referent, and where "...surround..." is a relation that can be non-distributive

> with respect to its first argument (also with respect to the second, but that's
> not relevant in this example).
>
>
> > > > For me {loi tadni} means "students", just like {lo tadni}, and
> > > > the mass gadri in addition indicates that whatever is predicated
> > > > of the students is predicated non-distributively. For you it means
> > > > "a group of things that includes students among its members",
> > > > which is something quite different.
> > >
> > > If you'd like, it can be "an entity composed of students", it doesn't
> > > really matter. And yes, this would be quite different, because it
> > > treats the students together as a different entity than each of the
> > > students themselves.
>
> Quite. You only accept referring to single entities, either to the single
> entity that consists of all the students, or to each of the students at
> a time. You won't admit a variable that has several referents, the students
> themselves not a new entity consisting of the students, which is what
> we need for plural predication.
>

> mu'o mi'e xorxes

And it seems he does not allow even these single entities the full range of properties -- or he
doesn't translate well between the two equivalent languages "Alice is one of X" = "Alice is a
member of X" and so on.

Maxim Katcharov

unread,
Jul 10, 2006, 3:10:47 PM7/10/06
to lojba...@lojban.org
Has there been any progress on a response? My last message was both
short and lucid; I imagine that the responding silence indicates
concession, though I doubt that this is the case.

On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:

Maxim Katcharov

unread,
Jul 10, 2006, 9:45:08 PM7/10/06
to lojba...@lojban.org
On 7/10/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> On 7/10/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> >
> > What's a plural variable?
>
> A variable that can have more than one referent at the same time.
>
> > I point out that again, you're not offering any sort of explanation of
> > 'non-distributive', you're simply assuming that it exists.
>
> All right then, we'll just have to leave it at that. I'm afraid my pedagogical
> resources are exhausted.
>

I don't think that you've explained anything, though you've certainly
repeated "plural predication exists" many, many times. I take your
inability to explain your position as a sign that your position isn't
based on something explicable. If you do think that you've offered an
explanation of plural predication, let me know where it is. I would
hate to have your exchanges end and have missed it.

John E Clifford

unread,
Jul 10, 2006, 5:36:54 PM7/10/06
to lojba...@lojban.org
Boy I dropped this one too soon. This is all very clear and thus easily answered. Suppose in the
group surrounding the building there are some non-students and suppose further that, if all the
students were in their same places and these non-students were absent (and not replaced by
others), then the building would not be surrounded. In this case is {loi tadni cu sruri le
dinju}true? NO! Since the building was not surrounded but for the non-students, the students did
not surround the building *though they participated in the surrounding and may even have been the
main force in it -- and might even have been able to surround the building by moving to new
arrangements).

So, is {loi tadni cu sruri le dinju} equivalent to {da poi sruri le dinju cu gunma lo tadni}?
Assuming that {gunma2) need not be a complete list of the members (whatever) of the mass (that {se
gunma} means "is a member of mass..." not "are all the members of mass...") then NO! again, ssince
the latter allows that others might be essentially involved. (I skip over xorxes' worries about
admitting that there are such things as masses; they make no detectable difference). The {loi}
form is apparently equivalent to {da poi srur le dinju cu gunma lo tadni po'o] however (much as I
hate sing {po'o}).

Jorge Llambías

unread,
Jul 10, 2006, 4:42:14 PM7/10/06
to lojba...@lojban.org
On 7/10/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> Has there been any progress on a response? My last message was both
> short and lucid; I imagine that the responding silence indicates
> concession, though I doubt that this is the case.

It only indicates that I didn't see us getting anywhere. But let's pick up where


we left if you want.

> On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > On 6/11/06, Jorge Llamb�as <jjlla...@gmail.com> wrote:
> > > On 6/11/06, Maxim Katcharov <maxim.k...@gmail.com> wrote:
> > > > The surest way to show that I'm a fool for asking this 30th time
> > > > is to point me to an explanation that I haven't rightly shown to
> > > > be unexplanatory.
> > >
> > > I don't think you are a fool.
> > >
> > > Explanations are pointless at this point, because definitions cannot
> > > be right or wrong. We are now working with different definitions.
> >
> > I don't think that this is the issue.

I think it is. In particular, we seem to have different understandings


of what "distributivity" means. Do you accept that some relationships
can be non-distributive in one ore more of their arguments? If you don't
agree, if you can only accept distributive relationships, then no matter
how much we argue, we won't get anywhere.

> > For example, both of us understand and agree with a 1to1 relationship:
> >
> > Alice is inside the school

I wouldn't call that a 1to1 relationship. That for me is a predication.


The relationship in that predication is "...is inside of..." and it is not
necessarily 1to1. For example, there may be other people inside the
school, and Alice may be inside some other thing (a classroom, for
example) when she is inside the school. But that's probably irrelevant
to what we were discussing.

> > and both of us understand and agree with the basic plurally


> > predicative relationship
> >
> > the 26 students are inside the school >>
> > Alice is inside the school
> > Bryce is inside the school
> > [...]
> > Zoe is inside the school

We agree that "...is inside of..." can be distributive, yes.

> > and both of us understand and agree with the "mass" relationship


> >
> > the graphite and the wood are component parts of the pencil >>
> > the graphite is a component part of the pencil
> > the wood is a component part of the pencil

There is nothing new there. "...is a component part of..." is just as


distributive as "...is inside of...". Nothing especially "mass" about it.
It is not an example of a non-distributive relation. There is no
reason to call "...is a component part of..." a mass relationship.

> > I don't see your position as equally sensible. If I were to say "ok,
> > there's no mass involved", I would have nothing like this to rely on.
> > I would have
> >
> > Alice is a referent of X
> > X surround the building < axiomic explanation of this is needed

As written, that doesn't make sense. If Alice is a referent of X, then


X has to be something like a word. But words don't surround buildings,
so X could not surround the building.

What we want to say is

Alice is one of X
X surround the building.

where "X" is a plural variable, i.e. a variable with (possibly) more than one
referent, and where "...surround..." is a relation that can be non-distributive
with respect to its first argument (also with respect to the second, but that's
not relevant in this example).

> > > For me {loi tadni} means "students", just like {lo tadni}, and
> > > the mass gadri in addition indicates that whatever is predicated
> > > of the students is predicated non-distributively. For you it means
> > > "a group of things that includes students among its members",
> > > which is something quite different.
> >
> > If you'd like, it can be "an entity composed of students", it doesn't
> > really matter. And yes, this would be quite different, because it
> > treats the students together as a different entity than each of the
> > students themselves.

Quite. You only accept referring to single entities, either to the single


entity that consists of all the students, or to each of the students at
a time. You won't admit a variable that has several referents, the students
themselves not a new entity consisting of the students, which is what
we need for plural predication.

mu'o mi'e xorxes

Reply all
Reply to author
Forward
0 new messages