* Sunday, 2011-11-06 at 19:42 -0300 - Jorge Llambías <jjllambias@gmail.com>:
> On Sun, Nov 6, 2011 at 7:11 PM, Martin Bays <mbays@sdf.org> wrote:
> > * Sunday, 2011-11-06 at 18:58 -0300 - Jorge Llambías <jjllambias@gmail.com>:
> >>
> >> You could extend that method to all predicates. So instead of:
> >>
> >> lo prenu cu klama lo zarci lo zdani lo dargu lo karce
> >>
> >> you can say:
> >>
> >> lo prekla cu klama lo zacyselkla lo zdaterkla lo dagvelkla lo karcyxelkla
> >
> > Are you mocking me? zo'o
>
> Only on the surface. :)
>
> > There's no problem with prenu klamaing; there is a problem with ranmapku
> > klesiing.
>
> But maybe someone else does have a problem with prenu klamaing, this
> method would solve it.
>
> The difficulty arises when you want to apply the method to "lo prenu
> cu klama gi'e tavla", or, in your case, to "lo ranmapku cu klesi lo
> mapku gi'e cmamau lo bolmapku". How do I know whether "lo ranmaplei
> ka'e cmalu"?
Yep, this is why the difference between using your "anything goes"
domains and this more regimented domain is more than those three
characters...
I wouldn't want to say that berets being small implies that lo ranmaplei
is/are small - because the same logic would give that lo ranmaplei cu
ranmapku, which is what leads to evil.
So lo ranmaplei would be more like another way of looking at lo ka
ranmapku.
This is partly why I'm not all that averse to allowing {lo mapku} get
the kind 'berets', or something like it, which does mapku. It's just
that we then have to complicate the logic to ensure that {su'o mapku}
can't get it.
(just as was done with plural semantics)
> >> >> >> Why is beret - hat - garment artificial?
> >> >> > But they're all on the same level, no?
> >> >> As kinds? A beret is a kind of hat, and a hat is a kind of garment.
> >> >
> >> > I'm only seeing two levels there - one containing berets and hats, and
> >> > one containing kinds of hat and kinds of garment.
> >>
> >> One level contains the garment, next level down the hat, next level
> >> down the beret (and a fourth level I hadn't mentioned because it won't
> >> enter into a klesi relationship contains some particular individual
> >> beret.) So "klesi" relates two things from different levels.
> >
> > Oh, no, we're not using the same definition of 'level'.
> >
> > I was reserving that for the problematic case, where going up a level
> > corresponding to passing from AE to EA. I believe this corresponds to
> > (your version of) {mupli}.
> >
> > I meant to ask for a natural example where we go up two mupli-levels
> > from mundanes.
>
> lo vi ranmapku cu mupli lo ranmapku noi mupli lo mapku noi mupli lo taxfu
>
> But I admit I became lost as to what we are after here.
Assuming I'm guessing correctly what your various {lo} phrases are meant
to refer to, that's only two levels in the sense I mean 'level'.
"Some hats here are examples of Berets, which is an example* of Hats,
which is an example* of Garments"?
You could use {klesi} for 'example*', so there's no level-crossing.
Beret is a subclass of Hat, not an instance.
If I'm not guessing correctly, and you really did mean
"Some hats here are examples of Berets, which is an example of
Kinds-of-hat, which is an example of Kinds-of-Kinds-of-garment", then
I dispute the naturality of the example!
Attachment:
pgpbm2PfUGiUg.pgp
Description: PGP signature