Re: colin's grammar change proposal

[CJ FINE <cbmvax!uunet!cuvmb.cc.columbia.edu!C.J.Fine>]

>
> Unfortunately le pe mi pendo is ambiguous when generalized, because of the
> existence of indefinites (ci pendo, etc.)  This is a long-standing problem
> in increasing flexibility of the grammar of sumti, but JCB tried once
> or twice to get rid of indefinite sumti, and they just kept coming back in his
> and others usage being so natural.  Indefinites seem more important to me than
> fronted relative clauses.  I'm willing to exper8iment with the latter, but
> that experiment must be unambiguous, and must not use any new cmavo/selma'o
> or I will be opposed.
>
> lojbab
>
I don't understand this posting. I don't khow whether you are now
rejecting all of my suggestion, or part of it, or part of your
translation of it.
I can't believe that generalising
        LE <sumti> <selbri>
to
        LE <relative clauses> <selbri>
can introduce any ambiguity, and I think this is worth doing.

Your suggestion (apart from the vocative stuff, which I hadn't thought
of) seemed to try to combine my two ideas into one and then got stuck in
quantifiers. I am not entirely happy (as somebody else said) about
        le poi crino ku'o le pendo
- there seems to be one more descriptor than is comfortable, but I was
willing to go along with it for getting the generality of my idea in. If
your conclusion is that you can't get the full generality in without a
new cmavo, and you want to reject it for that reason, I accept that. But
if you are rejecting the whole idea, I am not happy.

If you are concerned about introducing an asymmetry, in that only
descriptions will be able to have fronted relatives, then I respond that
we already have that symmetry, as only descriptions are able to take the
highly anomalous fronted possessor.

                Colin