Bunches
> > "-F" (and "F-") – distributive predication
> ("collective predication,"
> > including personal predication for
> individuals, is the norm),
> > "F*" – "participates in Fing."
>
> I suppose you would also need to distinguish
> "*F" from "F*",
> "participates in Fing" from "participates in
> being F-ed".
Good point. If the second place is collective
then there must be participation notion there as
well. So, like dash, star goes more with the
argument than the predicate (though sometimes
with the predicate, apparently). Notation needs
some work here.
> I think that would be the best way of marking
> distributivity. Marking
> it on the sumti is not very convenient because
> we sometimes want
> to use the same sumti filling a distributive
> place for one predicate
> and a collective one for another, or because it
> is convenient for
> anaphora to have as antecedent a sumti that is
> neutral with respect
> to distributivity.
Well, sometimes we also want a predicate taking
one argument collectively and the other
distributively at the same place. As I said, the
notation needs some work here, lathough these are
actualy problems for applications, not for the
pure system, which lacks non-sentential
conjunctions.
> Unfortunately, in Lojban this way of marking
> distributivity is complicated,
> because Lojban predicates don't normally have
> just two arguments, but
> they can have any number of arguments. So where
> does one put the
> mark?
With the argument generally - except in the cases
noted. that is, after all how Lojban does it
(with {lV} vs. {lVi}. But it lacks devices for
the exception (and for the predication internal a
description). I want to get the formalism right
before I worry too much about any application to
Lojban.
> Perhaps a practical solution would be to put
> the distributive mark
> as a selbri-tcita when it corresponds to the
> x1, and on the corresponding
> sumti in other cases. This is not the most
> elegant way of doing it, but
> because the x1 is the most likely sumti to be
> shared by more than one
> selbri, it would at least cover the most cases.
As noted, Lojban already covers most cases quite
well. The remaining ones are going to take some
novel device in any case. As far as I can see,
some such device is going to be needed whatever
happens with this notion (It appears to have to
be UI or one of the other few "happen anywhere" groups.)