* Monday, 2011-08-15 at 19:54 -0300 - Jorge Llambías <jjllambias@gmail.com>:
> On Sun, Aug 14, 2011 at 1:56 PM, Martin Bays <mbays@sdf.org> wrote:
> >
> > I was using 'xorlo' as shorthand for 'the bpfk gadri
> > proposal'. But it seems that the definition there of {lo} *does*
> > interact with masses/groups, because its referents are explicitly
> > allowed to be groups.
>
> The referents of "lo" can be anything at all. The referents of "lo
> girzu" will be groups. The referents of "lo gerku" could be groups
> only if there are groups that are also dogs, which is not a matter to
> be decided by "lo". It's part of the semantics of "gerku". Can a group
> of things gerku, or can only the members of (certain) groups gerku?
> That depends on how "gerku" is defined.
Good.
> In any case, the BPFK definition of "lo" doesn't mention groups, does
> it?
Not the definition itself, but later on the wiki page it states that
individuals can be groups.
If the reader is unfortunate enough to interpret that as including
CLL-masses (where by "CLL-mass" I mean to indicate that the upwards
closure axiom applies, i.e. a mass brodas if any submass does), then
they end up with such oddities as {lo plise cu canko} being possible
(where {lo plise} is interpreted as the CLL-mass consisting of an apple
and a window, which does indeed both plise and canko).
> > My feeling is that the level-mixing ambiguity which allowing group
> > satisfaction of broda in {lo broda} would introduce - {lo besna} could
> > have its referents being neurons,
>
> Only if neurons can be a brain. The problem is that in English "to be"
> is sometimes used to mean "to constitute". If "lo so'i nirna cu besna"
> is true,
Yes, I meant for us to say for the sake of argument, ignoring
physiological reality, that a brain does consist just of a load of
neurons.
Or do you want to say that it *doesn't* follow that those neurons
collectively {besna}, because the semantics of besna are such that only
individual brains satisfy {besna}? And generally that plural predication
is reserved for a few special predicates like {sruri}? That might make
the ambiguities more manageable.
> then I don't see much of a problem in using "lo go'i", i.e.
> "lo besna", to refer to the same things that "lo so'i nirna" refers
> to.
But what are the referents of {lo besna}? Brains, or neurons? If you
leave it ambiguous, won't this cause confusion? e.g. how would you
translate "these brains are conscious" without being misunderstood as
claiming that each of their constituent neurons are?
> Whether or not "lo so'i nirna cu besna" can be said to be true is a
> matter of the semantics of "besna", not of "lo".
>
> > and generally {lo gunma be lo broda},
> > with {gunma} as per your definition below, could have its referents
> > brodaing - would be an ambiguity too far.
>
> In my understanding of "gunma", "lo gunma be lo broda cu broda" is not
> a tautology, although it is true in many cases. It depends on what
> "broda" is.
Right - but if you are allowing group satisfaction of brode in {lo
brode}, then {lo gunma be lo broda} could be understood as having some
brodas as referents - since the brodas do, collectively, {gunma be lo
broda}.
The brain-neuron level-mixing ambiguity was an example of this, taking
{nirna} for {broda}, and assuming that brains gunma neurons.
> > And more generally, would you drop CLL's upwards closure axiom for
> > "masses", such that you can't have {lo plise} having as referent a group
> > whose constituents are an apple and a badger?
>
> ("lo plise" has apples as referents, so when speaking carefully I
> wouldn't say it has a group as referent.) And of course, I wouldn't
> say that "lo plise jo'u lo takside cu plise" is true, and so I
> wouldn't use "lo plise" instead of "lo plise jo'u lo takside" to refer
> to the apple and the badger.
Good.
> > However, I don't think that it is so useful to have the 'constituent'
> > relation as an ordinary selbri. A magic cmavo might be better.
>
> I think that whether there is a magic cmavo or not, there must
> definitely be an "x1 constitute x2" relation as an ordinary selbri.
Which holds when x2 has referent a group, x1 has as referents the
constituents of that group, and the predication is read collectively on
the left? And never holds with a distributive reading on the left? Or
can, but only when x1 and x2 both have the same single referent?
> > Consider : we would always have {ko'a gunma ko'a}, where we interpret
> > collectively on both sides.
> "gunma" as a symmetric relationship? We already have "du" for that.
This assumed that "ko'a collectively broda" and "the group whose
constituents are the referents of ko'a brodas" are equivalent - which
below you don't accept.
So how do you see collective predication and groups-as-individuals
interacting?
> And we would still need some other predicate for the cases where we do
> need the group to be a new entity. For example, when we want to
> quantify over groups.
>
> > Similarly, the referents of {lo selgunma be lo sruri} could be the
> > individuals which as groups sruri, but they could also be the sruris
> > themselves - depending on whether {lo sruri} is taken to sumti
> > distributively or collectively.
>
> There's no way of knowing whether "lo selgunma be lo sruri" refers to
> people or groups of people, since we don't know whether "lo sruri" is
> a group of people that surround the building or a group of groups of
> people that surround the building. (It has to be a group of something
> in order to fill the x2 of "selgunma"
I was expecting non-groups to be able to fill the x1 of gunma, as long
as they did so collectively.
> ) But that's because so many different things can sruri.
>
> > And if you want arbitrary sumtis to be able to sumti collectively,
> > then getting a good theory of collectivity is necessary even to
> > understand {mi}, nevermind {loi}.
>
> "mi" is hardly ever plural, so let's use "do", but it's the same deal.
> I don't see any special problem with it, it's basically equivalent to
> "lo te cusku" (at least as far as collectivity goes). No need to
> invoke the "mass" morass for it.
I just meant that you are now allowing non-distributive readings of
{do broda}, so it's important to understand how these work.
(Contrast with CLL, where 6.6.4 and the paragraph above it quite
explicitly equates {do} with {ro do})
> > So allow me to recap how I would like to understand all this:
> >
> > The data in the interpretation of an ordinary sumti-6 is just a set of
> > individuals, its referents. When it sumtis, whether it does so
> > distributively or collectively is ambiguous.
>
> Yes, but... "when they sumti", please, not "when it sumtis". It's the
> referents, not the set, that do it.
Agreed.
> > {lu'a} and {lu'o} can be used to disambiguate (forming extraordinary
> > sumti-6).
>
> I pass on that one. As I said, I don't think this is something that
> makes sense to mark on the sumti.
Well... grammatically, {lu'o ko'a} is a sumti. So if you don't want
distributivity marking, it would have to actually have different
referents from {ko'a}. The only obvious interpretation remaining is that
it has as unique referent the group whose constituents are the referents
of {ko'a}. Similarly, {lu'a} would have to correspond to the inverse
operation - breaking a group into its constituents. But then {lu'a ko'a
broda} generally won't be a possible reading of {ko'a broda}.
> But according to the lore, in "lu'o lo broda cu brode", "lu'o" is
> supposed to say how the brodas brode, not how they broda.
>
> > Non-fractional quantification unambiguously quantifies over the
> > referents.
>
> Yes, but... not everyone agrees on what the referents of "loi broda"
> are. Are they brodas, or are they groups of brodas? That's the
> neverending discussion, and neither solution is really satisfactory.
>
> > The referents of {lo broda} are such that
> > ONE OF (currently unsettled)
> > (i) each referent satisfies broda, i.e. {ro lo broda cu broda} is
> > tautologous.
> > (ii) the referents either each satisfy broda, or they collectively do,
> > i.e. {ro lo broda cu broda .ija lu'o ri cu broda}
>
> Assuming that "lu'o" means "collectively" and not "lo gunma be". And
> also presumably assuming that "collectively" includes such things as
> "in pairs" and other distributions, not just "all together".
Ugh. I was hoping to avoid that.
> For example, if I want to say "my coworkers brought their children to
> work today", and suppose that some of my coworkers have children
> together. How would you do it with "lu'o", keeping in mind that there
> is no child that they all brought together?
Assuming {lu'o} works appropriately, why not
{ca le cabdei su'o kansa be mi gunka cu kansa su'o panzi be lu'o su'o ri}?
(The first {su'o} could be {lo} or {loi} or {le} or {lei}, if you prefer)
> > This constitutes the definition of {lo broda}, i.e. no further
> > information about the referents can be deduced (modulo usual assumptions
> > of contextual relevance).
>
> I think (i) is the way to go: "lo broda" = "zo'e noi ke'a broda", and
> not "zo'e noi ro ke'a broda".
>
> > A group is a kind of individual, so a possible referent of a sumti-6.
>
> Certainly, for example a referent of "lo girzu".
>
> > A group has as data a set of individuals - its constituents.
>
> OK.
If you agree that this is all the data in a group, then {girzu} is
maybe not a good word to use... how about {zilgri}, defined to kill the
x2 and x4 places of {girzu}?
> >Things "collectively broda" iff the group whose constituents are
> >those things brodas.
>
> I'm not sure this will always hold. Do we need it for something?
Elegance?
> > (I think John Cowan would want to disagree here, and say in particular
> > that the group with only one constituent should not be distinguished
> > from that individual. Is that right, John?
>
> (I think you mean John Clifford.)
(Oops, yes)
> > But this seems not to fit with your doi xorxes account of loi.)
>
> I don't have one account of "loi", I claim there are (at least) two
> accounts, and I favor neither. My preference is for forgetting that
> "loi" exists.
OK, I'm happy to ignore it for now.
> > The exact semantics of when a group brodas depends on broda. Perhaps the
> > x1 of sruri is upwards closed, but the x1 of plise certainly isn't.
>
> Yes, I agree with that.
Martin
Attachment:
pgpR4IIvqQGG4.pgp
Description: PGP signature