[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
xorxes:
> la and cusku di'e
>
> > > Do you want {lu'i lo'i broda} = {lo'i broda}?
> >
> >I don't really care, so far. I hadn't seen a need for lu'i until
> >you pointed out that "lu'i su'o broda" can replace "lau'i broda"
>
> I want lu'i lo'i broda to be the set that has lo'i broda as its
> only member rather than the set that has the members of lo'i
> broda as its members
I understand. Hmm. I think I am undecided, though I wanted "lu'ilo'i"
to be equivalent to "lo'i" for the sake of a macrogadri paradigm
unbiased by length.
BTW, this gives us *three* possibilities for the meaning of
"lu'i ri", where ri already refers to a set.
1. The set containing ri as its sole member.
2. The set ri.
3. The underlying set of ri's antecedent.
> > > Do you want the arguments of LAhE to always be sets, or do you
> > > want it to be irrelevant what gadri is inside other than
> > > the o/e/a distinction?
> >
> >In one sense, there would be no other gadri but lo'i/le'i/la'i,
> >except as abbreviations
> >
> >But that doesn't really answer your question. I suppose my
> >current answer is that whatever the answer, it needs to be
> >consistent with reducing non-set gadri to LAhE+set gadri
>
> I don't see you can make these compatible:
>
> 1- lu'o lo'i broda = loi broda
> 2- lu'o ro lo broda = loi broda
> 3- lu'o da poi selcmi be ro broda = loi broda
>
> 2 and 3 are incompatible. I want 2. You want 1 and 3
Yes. But suppose the rule is that LAhe(x) = the LAhE of
the constituents of x, and if x has no constituents then
x is a constituent of itself. 'Constituent' = member
of set and member of collective. It seems to me, though
the haze of my woolly thinking, that this rule will cover
1, 2 and 3.
But "lu'o ro lo selcmi" would yield the collective of the
members of each selcmi, not the collective of the selcmi.
I don't know whether you could live with that.
--And.