* Sunday, 2011-10-30 at 11:08 -0300 - Jorge Llambías <jjllambias@gmail.com>:
> On Sun, Oct 30, 2011 at 1:47 AM, Martin Bays <mbays@sdf.org> wrote:
> >
> > I suppose the point is that effectively forcing *distributivity* is
> > possible, in this way. But I don't know what could force
> > non-distributivity, unless we actually have ambiguous predicates.
>
> We do have predicates that are non-commital about distributivity. In
> "ko'a bevri ko'e", and assuming both ko'a and ko'e have more than one
> referent each, we don't know (and possibly don't care) how the loads
> were distributed. If we say "the men carried the tools to the shed" we
> may not know which of the men carried which of the tools, whether some
> tool was carried by more than one man, whether some man carried more
> than one tool, and so on.
Right, but that isn't actual ambiguity of the predicate, just laxness in
its truth conditions. If we wanted using {lVi} to give more information,
it would have to be by claiming or denying bevri for some subbunches of
ko'a and/or ko'e. I don't see any possibilities which would do anything
useful and sensible.
> >> She could have said: ".ei no roi ku su'o da poi cinfo zo'u do klama lo
> >> jibni be da".
> >
> > With that being clearer just because it would be perverse to
> > existentially quantify over a singleton set, indicating that the
> > intended domain probably has multiple things which cinfo?
>
> Right.
>
> > But they could
> > just as well be multiple kinds of lion rather than individual lions?
>
> Yes, but that would still be enough to forbid Moople from getting
> close to any kind of lion, even if he accepted the existence of kinds
> for a minute.
>
> > I think she ought to be able to be even clearer.
>
> She should say "su'o do" and "su'o jibni" for extra safety.
>
> >> (I first wrote "ko" instead of "do", but that suggests the scope of
> >> the imperative is within the scope of "no roi", which seems wrong.
> >
> > Hmm... I'd have thought that the imperativeness of {ko}, like the
> > questioniness of {ma} and perhaps the constancy of {lo}, beats ordinary
> > quantifiers. Any reason to have it otherwise?
>
> It just sounds wrong to me. Fortunately in the case of "ko", I can
> always replace it with ".ei ... do" and place ".ei" at the right
> level.
And {.ei} is affected by scope?
{.ei mi klama su'o zarci} -> I have to go to a market
{mi klama su'o zarci vau .ei} -> There's a market I have to go to
{su'o zarci mi .ei se klama} -> There's a market I have to go to
?
Sounds good, though I don't see immediately what the scope rules should
be (e.g. would it make a difference in the third sentence had I written
{su'o zarci .ei [...]}? How about {su'o zarci ku .ei [...]}? Or do these
in any case indicate that it's the market which should be coming to
me? And is there a subtle difference between the second and third
sentences?).
> >> In any case the moral of the story was that there are sound
> >> evolutionary reasons for the kind approach, since obviously we are all
> >> descended from Cless and he is the one that got the right meaning. :)
> >
> > Somehow this doesn't seem to be the moral I've taken. Maybe Moople had
> > a cousin with individuating parents?
>
> But what language do they speak? Not English, because in English we
> can say "lions are dangerous, don't go near them".
Yes, but we mean something about individual lions when we say it
- namely that they tend to be dangerous.
Attachment:
pgp3sGsbGdkIg.pgp
Description: PGP signature