[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
On Sun, Dec 15, 2002 at 09:27:43PM -0000, And Rosta wrote:
> Jordan:
> > On Sun, Dec 15, 2002 at 02:27:56PM -0000, And Rosta wrote:
> > > Jordan:
[...]
> > > > I disagree. o-gadri are out, regardless of the membership of lo'i
> > > > stedu be mi, because I am talking about a specific thing.
> > >
> > > Certainly if you intend to refer to a +specific thing, i.e. something
> > > involving le'i, then o-gadri are out. But one can speak of someone's
> > > head without referring to it
> >
> > We're talking about refering to it. One ought not to make claims
> > about their head without refering to it. Claims without refering
> > to it should be limited to claims of its existence
>
> I think we need to concretize which expressions we agree count as
> referring, and which are permissible in talking about someone's head.
>
> As things stand, I cannot see why one ought not to make claims about
> someone's head without using a +specific or LA gadri or ko'a-series
> KOhA, which is how I would tend to interpret the term 'referring
> expression'.
Because stuff like
lo stedu be mi cu cmalymau lo'e plini
suggests I have multiple heads.
This viewing
da stedu mi
as a claim about me, not about my head.
> > > > lei makes
> > > > no sense, because it implies there is more than one,
> > >
> > > As you know, I think the implication that there is more than one is
> > > either less with lei than with le, or else equal
> >
> > I don't see why you think that
>
> The reason for them being equal is if le is distributive and lei
> is collective; the distinction is pertinent only to groups of more
> than one.
But this is precisely why I think collectives imply there is more
than one.
> The reason for lei implicating plurality more weakly is that lei
> is itself a reference to a single individual (tho not necessarily
> a single broda), while le quantifies over a set of broda. When
> I say "the reason", I mean "a reason", because this isn't a
> matter of right and wrong.
But lei references the members of the set also, it just doesn't
quantify across each individual. Since in the case of only 1 member
of the set treating it as an individual is the same as treating it
as a mass, and since individuals are more fundamental than masses[1],
using a mass gadri is odd.
[1] In cases where individuals don't really make sense for a
particular thing, such as water/sand/etc, the brivla are defined
so that individual references actually reference masses already (le
djacu is a mass, and lei djacu is a mass of masses), which keeps
individuals more fundamental.
> > > > and furthermore
> > > > that collectively it can be (conceptually) infinitely subdivided
> > > > le'i refers to something other than the head. le'e is obviously
> > > > wrong, and implies that there is more than one head.
> > >
> > > le'e is definitely not wrong, since it refers to a single head,
> > > or, if you prefer, *the* single head. Some but not all opinions
> > > on the generic gadri (but not the Prototype view) imply that the
> > > membership is nonsingleton, but I can't see how the implication
> > > would be stronger than with other gadri
> >
> > The whole point with le'e/lo'e is taking a representative, imaginary
> > instance of the whole set.
>
> As you know, we don't agree on this. I don't agree that the instance
> need be representative or imaginary. It can be real, but it must be
> the only member. If under the operative worldview it cannot be true
> that there is only one real member, then the one member must be
> imaginary. Even then, it needn't be representative if there is a
> sensical way to reduce a manymembered set to a single nonrepresentative
> member, though it may in fact be the case that representativeness is
> inherent in the process of singularization.
*shrug* I think it needs to be representative. If it isn't, the
gadri is far too open in terms of meaning (you probably want a
tanru---mi tanxe nitcu, instead of this mi nitcu lo'e tanxe stuff).
> > If there's only one member of le'i
> > broda, then "le broda" is the representative member by definition
>
> As is lei.
But lei has the problems discussed above. We're talking about le'e now.
[...]
> > > > "le" is the correct gadri
> > >
> > > None of the gadri are incorrect. It is a matter of which one has
> > > the least unfortunate implications. I won't repeat all the arguments
> > > all over again, because they're in earlier messages and on the wiki
> > > Personally, for {stedu be mi} I would use {loi'e}, or {lo'e} if
> > > that is equivalent to {loi'e}
> >
> > But you (by your own admission with those stats on the wiki) use
> > lo'e for basically everything. It makes no sense to me. Esp
> > since we already ruled out o-gadri
>
> I take "lo'e broda" to mean "the one and only broda", with the
> "one and only" bit presupposed rather than asserted. In the light
> of that, it will be obvious why it gets used so often.
I understand that. What I don't understand is *why* you take it
to mean that.
> I'm not sure what you mean by "we already ruled out o-gadri". Any
> gadri (apart from lV'i) yields truth when talking of a singleton
> category. The disagreement is about which gadri most strongly
> implicates that the category is not singleton. If we had gadri
> that actually encoded singletonness (in paradigmatic contrast
> with the other gadri) without any element of genericity,
> then we wouldn't have to be having this debate.
Lojban goes out of its way to try to not have any difference based
on number of referents (there's no plural, etc). Do you just want
things to depend on plurality? I think there's nothing wrong with
using "le" for this, as is intended. In cases where the number is
important we have the inner quantifiers, or we can explicitly state
the outter quantifier, and such things.
--
Jordan DeLong - fracture@hidden.email
lu zo'o loi censa bakni cu terzba le zaltapla poi xagrai li'u
sei la mark. tuen. cusku
Attachment:
binOy01N8E6JK.bin
Description: application/ygp-stripped