[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >


[Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: [engelang] Xorban: Properties




On Wed, Oct 24, 2012 at 5:21 PM, selpa'i <seladwa@hidden.email> wrote:
 

Am 24.10.2012 23:23, schrieb John E. Clifford:


> On Oct 24, 2012, at 3:46 PM, "Mike S." <maikxlx@gmail.com
> <mailto:maikxlx@gmail.com>> wrote:
>>
>> le je li fgri mplike li glri kmeki plkeka'a
>> "The fire-instantiated property of hot things is pleasant to me."
>> "The fire's hotness is pleasant to me."
>>
>> That's verbose, but that does seem to be the right logical form. In
>> Lojban:
>>
>> .i lo se mupli be lo fagri be'o je ka glare cu pluka mi
>
> Or more simply lo ka lo fagri cu glare cu pluka mi. (apparently you
> can't do that in X).

You can't do that in modern Lojban either. A ka without a ce'u is not
well-formed by today's standards (at least according to most people).

That would be "lo nu lo fagri cu glare cu pluka mi". I don't see how a
property is necessary in this particular case.

I think I made a suggestion about properties indirectly in some thread,
and I think the idea was to have a variable that works somewhat like
ce'u, then you can just do something like "glre'u" (I know e'u has been
used as "someone"), which means pretty much "hotness". It's the abstract
idea of something being hot. "e'u" wouldn't be bound by anything in the
sentence, it would stay vacant all the time to allow for those general
"-ness/dom" expressions.
I think you can usually use normal events for things like the above
sentence, and in the less common cases where you can't, you might just
say something equivalent to "the fire's hotness" ~ "la fgra le'u glre'u
psake'u" (how to say this using the new p- operator?), and now I see the
problem with this approach; there is only one e'u variable, and you
might want more than one ce'u in a sentence. Clearly, this idea isn't
entirely sound..., but I hope I could get it across. I wonder how
terrible "le glre'u" would be instead of "le'u glre'u", that way we'd at
least have infinitely many ce'u to work with, but the binding looks bad.
I'm really not convinved that properties need a new ka-like abstraction
to be added. Gua\spi also has only a single abstractor for everything
(plus one for direct speech).

mu'o mi'e la selpa'i

Having gotten your input, and that of another over at the Lojban group, the modern/CLL usage of L "ka" is a clearer to me now.  I think we already have a facile way of importing this "ka", if we want to do so.  Observing that X fV corresponds to L nu and L ka is equivalent to L nu with one or more places filled with "ce'u", all we need is a way of marking a given variable as being "ce'u"-like.  We already have a set of possible forms to do the trick: fV + kV + kV + ...

le fe la pra li mlti prmaki
lo nu lo prenu cu prami lo mlatu
"e is the proposition that people love cats"

le feka li mlti prmaki
lo ka prami lo mlatu
"e is the property of (one) loving cats"

le fekaki prmaki
lo ka ce'u prami ce'u
"e is the relation of (one) loving (another)"

In effect, the kV of fVkV binds its V as a sort of metavariable.

I am not sure we *need* this right at the moment, but it seems somewhat doubtful to me that we will find a better use for fV + kV.  (It's interesting that this was already proposed once earlier, but the first proposal evolved into X xV.)

--
co ma'a mke

Xorban blog: Xorban.wordpress.com
My LL blog: Loglang.wordpress.com