[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
Adam:
> la and. cusku di'e
>
> > IMO, the inner quantifier *ought* to be ro, though. That way,
> > e-gadri give us a way to do "a kind of": {le broda} = "each dog
> > of a certain kind, each of a certain kind of dog"
> >
> > I imagine that the inner quantifier was set to su'o on the grounds
> > that you couldn't have in-mind a referent of {le no broda}. But
> > if the in-mind thing is an intensionally-defined set, i.e. the
> > referent of {le'i}, then {le no} is not excluded
>
> I don't see this. If 'le broda' is 'ro cmima be le'i ro broda' and the
> cardinality of 'le'i broda' is 0, then any statement about 'le broda'
> is vacuously true in the same way that any statement about 'ro
> pavyseljirna xirma' is vacuously true. It tells us nothing about the
> intensionality of 'lo'i pavyseljirna xirma' to say 'roboi pyxy. cu
> blabi .ije roboi pyxy. naku blabi'
I'm not sure what you don't see, or where we disagree. lo'i refers to
an intensionally-defined set -- the set of all things that the
property expressed by the sumti tail description. All I'm saying is
(a) {le'i ro} likewise refers to an i-defined set (though with
the defined property glorked from context constrained by the
sumti tail description), and -- much less relevantly -- (b) I would
have preferred for the default inner quantifier to be ro rather than
su'o.
> In other words, it seems to me that you are equating 'le no broda'
> with 'le'ei no broda', which IMO is not the case. 'le no broda' always
> makes a vacuously true claim, whereas 'le'ei no broda' (though very
> general and non-specific) makes a non-vacuous claim
I don't understand lo'ei/le'ei in anything but a vague and unsatisfactory
way, so I can't comment. At any rate, I claim that the sorts of
claims made about {ro le no broda} have equivalent status (in terms
of when and whether they are vacuous) as claims made about
{ro lo no broda}.
--And.