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

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

On 5/28/06, Jorge Llambías <jjllambias@gmail.com> wrote:

On 5/28/06, Maxim Katcharov <maxim.katcharov@gmail.com> wrote:
> On 5/27/06, Jorge Llambías <jjllambias@gmail.com> wrote:
> > Then I don't understand why the above was an absurd misrepresentation
> > of your position.
>
> Because
>
> 1 {le mu no tadni cu dasni lo mapku gi'e sruri le dinju}
>
> is not the correct way to say
>
> 2 "The fifty students wore hats and surrounded the building."

But I didn't say that your position implied that. I said that that's what
*my* view allows, i.e. the view that does not introduce an encompassing
entity allows to use a distributive and a non-distributive predicate with
the same referent. Your view disallows it, because under your proposed
interpretation of {le} the Lojban claim would entail that there was an entity
that both wore a hat and surrounded the building.


No, under my interpretation, (1) is not the correct way to say (2),
which means that at no point do I assert that you would have a single
entity of 50 students that wore a hat or what have you. My position is
that you can't treat the same thing in two different ways on a whim -
that is, you can't treat a mass of students as a mass, and then as
individuals.



> In fact, (2) is not the 'correct' way to say it either - (2) is a
> shortened form of
>
> 3 "The fifty students wore hats and together surrounded the building."

I think (2) is proper and correct English. (3) is also correct, and more
precise.


Yes, it is, because English allows you to omit certain words that many
other speakers are also prone to omitting. However, when translating
into Lojban, you need to use the most "proper" wording, otherwise
there's a large chance that a mistake will be made. In this case, a
mistake /has/ been made. (2) was translated verbatim into

{[L_ muno tadni] cu [dasni lo mapku] gi'e [sruri le dinju]}

which is incorrect, since it implies that the (each) same thing that
wore hats surrounded the building. This is then excused by saying that
{L_ muno tadni} could be taken either as a mass or individuals - which
is wrong. Even English doesn't have that ambiguity:

"The fifty students wore hats and surrounded the building."

It's not the case that "the 50 students" acts differently in each part
of the sentence, it's that the word "together" has been left out.
Another way to solve this problem should be found (not a hard task),
because this solution is not right.


>  {le muno tadni cu dasni lo mapku gi'e lu'o lego'i sruri le dinju}
>  "50 students wore hats and together they surrounded the building"
>
> (Perhaps {loi} is better than {lu'o})
> (I'm not perfectly sure about {lego'i}, maybe one of
> {vo'a/ko'a/la'edi'u/ra} is more correct)

That's ungrammatical. {gi'e} connects bridi-tails, which consist of
a selbri followed by any number of sumti. The structure is as follows:

<sumti> ... [(<selbri1> <sumti> ...) gi'e (<selbri2> <sumti> ...)]

The fronted sumti are common to the two selbri, the trailing sumti go
with each of the selbri. You can't introduce a sumti between {gi'e} and
<selbri2>.


Thanks, my understanding of gi'e is quite rudimentary.


In order to connect two full bridi, you need to use {i je} in afterthought
mode, or {ge ... gi ...} in forethough mode. So either:

 le mu no tadni cu dasni lo mapku i je lu'o ra sruri le dinju

or:

 ge le mu no tadni cu dasni lo mapku gi lu'o ra sruri le dinju


Sure, whatever works best.


> Of course, if you think that {lu'o lego'i} is a big headache, then
> perhaps you'll support something like:
>
>  {le muno tadni cu dasni lo mapku gi'eloi sruri le dinju}

That's ungrammatical too. {loi sruri} is a sumti and after {gi'e} you
need a selbri.


Indeed. It's a suggestion: have a word that facilitates what you'd
like. {loicu} seems a bit long. Since {lu'o} seems redundant, perhaps
its meaning could be altered?



I don't think the expression {lu'o le go'i} is a big headache. The big
headache is having to separate the sumti into two sumti every time
you need to combine distributive and non-distributive predicates.


If you needed to say something about a dog (that it's white) and about
some cats (that they run), you would say

 {lo mlatu cu bajra i lo gerku cu blabi}

and not

 {lo danlu cu bajra gi'e blabi}

Yes, the first requires more effort, but the second is incorrect. In
the same way, I argue that you cannot see {L_ muno tadni} as one thing
for {dasni...}, and then something entirely different for {sruri...} -
a mass of students is, after all, a different entity than each one of
the students is.


> But I guess onwards with demonstrating that your (and McKay's)
> perception of masses is incorrect. I'll remind you of my objection to
> your conception of "mass":
>
> -quote-
> On 5/27/06, Jorge Llambías <jjllambias@gmail.com> wrote:
> > The non-singularist view holds that you can make a predication about
> > several things without having to distribute it or having to introduce
> > an encompassing single entity.
>
> How exactly is this sensible? X is a surrounder of the building, but
> only when you look at it as her being in the company of A..W? X the
> groupmate surrounds the building, but X does not? No, X the groupmate
> does not surround the building. A..X, together/as a group/mass
> surround the building.
>
> A quote (of supposedly McKay) that I found:
>  The fact that some individuals are surrounding a building does not
> automatically imply that some single individual (of any kind)
> surrounds the building.
>
> Yes, it does imply that. In this case, the mass of students is that
> entity. Just because a mass does not take on a distinct physical shape
> does not mean it's not there.
>
> Let's say that 10 soldiers then surrounded the students (together, and
> not in the sense that these 10 soldiers each hugged a student). We
> have a way to say this: "the squad surrounded the students". Yes, we
> do explicitly treat groups as entities. You can't simply say "well,
> look! No definite entity is mentioned explicitly, and I won't bother
> to see if an implicit entity exists. After all, it's ridiculous to
> think that some things seen together could just 'magically' be seen as
> a new entity, right? ...and therefore no entity exists".
>
> -endquote-

But in my view {le sonci cu sruri le tadni}, "the soldiers surrounded
the students", does not require any new entity either. The things


In my view, it doesn't require any "new" entity either. The entity is
already implied: the mass of soldiers. Calling it "new" makes it sound
like I'm bringing in something entirely foreign to account for this.

What I'd like to know is how you account for the building being
surrounded. What surrounds it? Each student does *not* surround it.
What surrounds it is that "mass" of students. It's a type of thing
that can be clearly recognized - we even have names for it: crowd,
mob, swarm.

You seem to have a belief that you can say that each student surrounds
the building, but only when seen in the company of other students. And
aha! You don't have to introduce some sort of strange and
other-worldly entity that clearly doesn't belong. How efficient!

But how exactly is this sensible? X is a surrounder of the building, but
only when you look at it as her being in the company of A..W? X the
groupmate surrounds the building, but X does not? No, X the groupmate
does not surround the building. A..X, together/as a group/mass
surround the building.

The crowd of students surrounds the building. The mass formed of
students surrounds the building.

(Even people who write books can be incorrect.)


being surrounded can be more than one just as much as the
surrounders. Any argument of the predicate can be distributive
or non-distributive.

> > > Ok, so your rule for using {lo} distributively/ambiguously is "if the
> > > outer is blank, it is ambiguous, otherwise, distributive", correct?
> >
> > For any sumti whatsoever, not just for {lo}, I take an outer quantifier
> > to be distributive, yes.
>
> I take this to mean that your{re loi ci nanmu} means something
> entirely different from "(only) 2 of the group of 3 men" ("...carried
> the piano"), yes?

In my view {re loi ci nanmu} means the same as {re lo ci nanmu}, because
the non-distributivity introduced by {loi} is then cancelled by the
distributivity
of the outer {re}. You'd have to say {loi re lo ci nanmu} to get a
non-distributive
"two of three".


This doesn't strike you as unnecessarily complex?


> Here is a simple demonstration:
>
>  your{lo mu tadni sruri lo pa grana}
>
> could mean either:
>
>  "together the five students surrounded the pole"
>  "the five students each surrounded the pole"
>
> So, yes, I want to call that ambiguous, and I'm perfectly correct in
> my usage of the word. I find your insistence that "ambiguous" refers
> only to gismu-like words with multiple meanings to be disagreeable,
> both with me and with the common dictionary definitions linked to
> earlier.

The marking of distributivity/non-distributivity properly belongs with
the selbri, not with the sumti. "Together" and "indiviually" in English
are adverbs, they modify the verb phrase, not the noun phrase, that's


Lojban uses cmavo that account for English verbs, nouns, tenses,
forms, adjectives, pronouns etc. - actually, I havn't checked if this
is true, but the point remains: you can't treat Lojban as English.
Noun phrases are not quite sumti, verbs are not quite selbri. It's not
enough to say "together is an adverb in English, so in Lojban it's
[...]".


why in English you can use the same noun phrase with a distributive
and a non-distributive predicate at the same time:

The fifty students (wore hats individually) and (surrounded the
building together)


"Individually" is a word for helping the English listener when that
listener might suspect that the word "together" has been omitted
(which it usually is, by the way). Perhaps you'll argue vice-versa?
Regardless, there is no superclass between "individually" or
"together". There are only shortcuts of the English language.


Two predicates, "...wore hats individually" and "...surrounded the building
together" are predicated of the same referents, the fifty students.

The adverbs "individually" and "together" can help make the predicate
more precise, but they are not obligatory. This is similar to the way you can
say: "The students walked quickly" and just "the students walked". The
adverb "quickly" makes the predicate more precise, it excludes the possibility
that they walked slowly, but does not make "the students walked" ambiguous.


This is all based on something unproven: please show that there is a
superclass above 10 students, and the collective entity formed of 10
students.


An ambiguous phrase is something like "time flies like an arrow" which
can be parsed in (at least) two different ways. But nevermind, keep
using ambiguous in your sense if it makes you happy. In your sense, my
{lo} is ambiguous, yes. I just don't think that says something very useful.

In Lojban, for whatever reasons (some good ones and some bad ones),
the distributivity of a place of a predicate ended up being marked on the
sumti rather than on the selbri.


Yes, {loi X} would mean {lo gunma lo X}. It's on the sumti because
that's where it should be. In English, it's "on the verb" because we
want to refer to X as a noun phrase, even though it's not the
'highest' noun phrase, because this is convenient in so many cases:

 "Together the students surrounded the building and wore hats."

We'd like "the students" to be the noun phrase of "wore hats", and so
that's the way English chooses to handle it - by making "together"
transient, that is, it only applies to the referent in that single
sumti-place, and is thereafter discarded.


But it is still useful to have a neutral form
of the sumti, so that you can combine distributive and non-distributive
predication without having to replicate the sumti.


Use {lu'o} (or whatever) after a {gi'e} in the same transient manner
in which English occasionally uses "together". There are many other
solutions.

I think it deserves mention that I don't see it as a "neutral form" at
all, since I don't think that such a thing exists, aside from as an
ambiguous structure in your version of Lojban.



> > If the speaker says {lo ro cribe poi zvati le dalpanka}, the very act of saying
> > those words brings bears not in the zoo into the universe of discourse
>
> Sure, but just because you 'brought a whole bunch of bears into the
> discourse', doesn't mean that your{lo ro cribe} will refer to all of
> them later. For example, we might talk about all bears in the forest,
> and then I might ask you your"did you see all the bears when you went
> to the zoo?" question:
>
> xu do pu viska lo ro cribe ca lo nu do vitke le dalpanka
> Did you see all bears when you visited the zoo?
>
> The your{lo ro cribe} would refer not to "all bears that have ever
> been brought up" (i.e. bears in the forest), but clearly to a
> contextually *sensible* group of bears, that is, the bears in the zoo.

That's right.


So, putting your 'individually'/'together' aside for now, you have your

A:  {lo ro cribe} "all bears of a contextually sensible set of bears"

which is equivalent to

B:  {ro lo cribe} "all bears of a contextually sensible set of bears"

since they would both mean the same thing in

 {xu do pu viska XXX ca lo nu do vitke le dalpanka}

putting your 'individually'/'together' aside, of course. We are right
to put it aside, because we should decide what the inner and the outer
mean before dealing with things like distributivity. Now, you have a
third ambiguous interpretation to fit in among {lo} (individually) and
{loi} (together). You choose to use one of the above to implement it.

There are two problems with this:

1) You're basing this addition on {lo ro cribe} meaning "all of a
group of contextually sensible set of bears" - effectively the same
thing as {ro lo cribe}. But really, you don't have two options, one of
which you can use for this "ambiguously together/individually". I
don't know what inner {ro} means in your version (I suspect it has no
meaning), but {lo ro cribe} does not mean "all of a group of
contextually sensible set of bears", because that's butchery.

Let's treat the outer and inner as a special fraction, for
demonstrative purposes. {pa lo re} would be [1/2], {su'o lo so} would
be [?/9], {ro lo ci} would be [*/3], {lo ci} would be [?/3], {ro lo}
would be [*/?]. This fraction has special properties, for example:
"[1/3] bears swim" means that the other 2 bears do not; etc.

Now, you propose that inner {ro} means "all of a contextually sensible
group of...", which is [*/?]. This would mean that {ci lo ro} is
[3/[*/?]] (which math tells us is the same as [3/?]). Why, exactly, do
we have a deviant three-part fraction like [3/*/?], especially when
[3/?] does the job well enough?

The interpretation that {lo ro} means "all of a contextually sensible
group of ..." is wrong. "All of a contextually sensible group of ..."
is *{ro lo}*, no need to create a strange and synonymous way to say
it. Ah, (you say,) but {ro lo} is ambiguous/distributive/whatever, and
that's different! No: first come up with a coherent and systematic
system of inner and outer quantifiers, and *then* worry about
distributivity/{lo} vs. {loi} vs. {lu'o}.

2) You can't use {lo danlu cu bajra gi'e blabi} to refer to a white
dog and running cats, 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. A mass of students is, whether it's
convenient or not, a different entity than what each one of the
students is. There is no way to refer both to "mass composed of X" and
"X" at the same time (there is no superclass).