Re: quantifiers on sumti - late response

[jorge@PHYAST.PITT.EDU]

And:
> Re lohi/lehi: Is this right?
>
>    lui ro broda = lohi broda
>    lui le ro broda
>    lui suo broda = "at least one set of broda"??
>    lui le suo broda = lehi broda ?????
>    re lui suo broda = "2 sets of broda" ???

I'd say yes, but note that {lu'i le ro broda} is almost the same as
{lu'i le su'o broda} = {lu'i le broda}.  The "almost" is because only
the first could refer to the empty set.

> I was putting forward the idea that every sumti in a certain way involves
> universal quantification. That's not the standard view, of course.

No, but also I don't think it would buy us much.  It would make saying
some things easier, but others more difficult.

> I suppose I ought to be careful to avoid {le remna cu citka lo plise}
> and instead say {ro le remna cu citka lo plise}. Ah - but that's what
> {le remna cu citka lo plise} means anyway, isn't it.

Yes, they both mean the same, no need to be careful there.

> If the same apples
> got eaten by each of the people, I'd have to say {fe lo plise le remna
> cu citka}.

Yes.  (Except for the place numbering convention.  I'm not sure what the
convention is, I think what you have is right, but I would take {le
remna} to be in the x3 of {citka}.)

> {lo remna cu citka lo plise} should be - how do I say it? -
> "some people each ate some apples"

Should?  What it says is that "at least one human x, and at least one
apple y are such that x eats y".  You are claiming that at least one
event takes place.

> {ro lo remna cu citka lo plise}?

That's "each and every human eats an apple" (not nec. the same apple)

> {ro lo suo lo remna cu citka lo plise}?

Yes, I think that's the one you want.

> > Consider:
> >   Three of my sisters have a pair of blue eyes.
> > What you really mean is:
> >        ro lo ci le mi mensi cu se kanla lo blanu remei
> Or {ro lo ci lo mi mensi cu se kanla lo blanu remei}?

Yes, I think.

> I of course see the logic behind this. But I find it mildly
> counterintuitive. If I want to say "some books are blue"
> I can say either {ro lo suho lo cukta cu blanu} or just
> {lo cukta cu blanu}, while if I want to say "some books have
> a colour" I need to say {ro lo suo lo cukta cu se skari}, or
> something like that.

What's wrong with {lo cukta cu se skari}?  Unfilled places are very
accomodating.

> And "People went" would be {ro lo suo
> lo remna cu klama}, unless I wish to assert they each went
> to the same place, from the same place, via the same place,
> etc. Is this right?

No, because the default filler is not {da} but {zo'e}, which has lots of
wonderful properties.  (Very lo'e-like now that I think of it, great
that the vowels agree!)

> > But you never encounter {lo'e remna}.
> I think it's more a question of how we perceive the world than how the
> world is. Suppose it is the case that {mi penmi da poi remna}. How
> can I describe that state of affairs to you? I can do it
> quantificationally, with {mi penmi lo remna}. Or I can select a
> specific individual possibly from among many, with {mi penmi le remna}.
> Or I can choose not to perceive that there are different remna;
> I can choose to perceive there to be only one remna. Then I'd say
> {mi penmi loe remna}. It's non-specific, because it makes no sense
> for you to ask "Which one", since I'm saying I only recognize there
> to be a single remna.

I agree completely.  As long as you don't go on to say that the head of
whom you met is the one an only head of {lo'e remna}.

> If I had to choose a gadri for {stedu be xorxes}, I could choose,
> on the one hand, {suo lo}, {ro lo}, or {le}, and the choice would
> be arbitrary, since they all come to the same thing. Or, on the
> other hand, I could choose {loe}, which I find especially appropriate
> for selecting from singleton categories.

Well, I find {le} made to fit for that, but I agree you can use {lo'e}.

But the logical implications you can derive from the statement with
{lo'e} are very different.

> At any rate, I do think {loe remna} can refer to a particular
> individual.

I don't, not in the same way that {le pa broda} refers unquestionably to
an individual.  We may be using "refer" differently, though.

> Is it also false to conclude from
>   mi penmi lo mamta be la xorxes
>   le se penmi be mi se stedu pa da
> that
>   lo mamta be la xorxes se stedu pa da
> ?

No, that seems sound. But {lo'e} is not like {lo}.

And notice that your conclusion is much weaker than what you really can
conlude.  You are concluding that "at least one of my mothers has a
single head", but you could in fact conclude that "the mother (the one
you have in mind, the one you met) has a single head".

> > > But I would like to know how to say loe remna has exactly one head,
> > > two arms, two legs.
> > It doesn't, in my opinion, because {lo'e remna} is not a sumti with
> > a fixed referent, to which you can give properties, much like
> > {pa remna} is not a sumti with a fixed referent to which you can give
> > properties.
>
> Well, you *can* say that pa remna has one head, two arms, two legs.

You can say anything you want, but if you say that you are saying
something that is not true, since more than exactly one human have that
number of members.  (Unless you want to argue that you mean at a certain
time, in a certain place, but I don't think that's fair.)

> > Ok, but make it {lo'e plise renomei}, which is slightly less
> > ambiguous.
> Why is it slightly less ambiguous? I thought Lojban eschewed
> ambiguity altogether.

Except for the tanruic variety.  {lo plise renomei} is a 20-some of type
apple, which can't be much more than a mass of 20 apples, I think.  But
{lo renomei plise} is at least one apple of type twenty-some.  It can be
any number of them as long as they are of the twenty-some variety,
whatever that is.

> > The properties of {loi remna} are of the same type as those of
> > {lo remna}.
> The properties of {loi remna} are a superset of the properties of {lo
> remna}.

Not really, they are just of the same type.  (Unlike sets, which have
weird properties like cardinality, and being included in other sets, and
such.)

For example: ta grake li ki'o
             That weighs one kilogram.

             ti grake li reki'o
             This weighs two kilograms.

             ti joi ta grake li ciki'o
             This and that together weigh three kilograms.

Both the mass and the individuals have the same type of property, weight
(actually mass, but who cares) but it is not the case that the property
of each component is a property of the mass.  Only some properties are
directly inherited by the mass.

> More accurately, the properties of {luo suo re remna} are a
> superset of the properties of each of {suo re remna}. The properties
> of {luo pa remna} are the same as the properties of {pa remna}.

The set of properties of the mass and the set of properties of the
components have non-empty intersection, but neither is a superset of the
other.

The properties of the set, on the other hand, are different.  Sets don't
have mass or weight.  It would be nonsense to say {ti ce ta grake da}

> The whole challenge of {loe} is that in some ways it is similar to
> {luo ro lo remna}, and in other ways similar to {lo pa remna}.

Yes, but I think it is most similar to {zo'e}, only more informative.
The important point is that the statement is not about the lo'e-sumti,
but about the other sumti.  The lo'e-sumti only complements the selbri.

> Pau loe nu smuni joi gerna certu kei se jalge loe nu gehi snada gi
> tolsnada vau loe nu tadni loe tarci saske?

i nagenai i zo'o mi na tadni lo'e tarci saske i zoi gy phyast gy sinxa
zoi gy Physics & Astronomy gy i mi cmima la'e gy Physics gy i a'o snada
ca lo djedi

co'o mi'e xorxes