Re: whether (was Re: ni, jei, perfectionism)

[Carl Burke <cburke@MITRE.ORG>]

At 04:31 PM 12/17/97 GMT+0, you wrote:
>Carl:
>> >> >> mi zanru le du'u melbi
>> >> >> I approve of the fact that (something is) beautiful.
>> >> >No. "I approve of the proposition that something is beautiful".
>> >> >"the fact that" is better rendered by "le nu".
>> >> So 'the fact that' is explicitly transient?
>> >No, but nor is nu.
>> I disagree.  Transience (used to be) the _essence_ of {nu}.
>
>What, then, would have been used for, say, modern {lo nu
>da zasti} - "the event of there being something that exists"?

That still sounds right, taking one meaning of 'modern' as the time
envelope which includes something that exists.  The bounds are left
unspecified, but there are probably contextual limits on the set of
objects that can satisfy {da} here.  Saying that something 'is' modern
(rather than referring to 'the modern era') might be {leka da zasti},
{leka lonu da zasti}, or something else entirely; I'm not sure about
translating that.  As a modifier, I'd rather translate it as some
variation on 'having the quality of being recent or soon'.

...
>> I disagree that {mi jmive} and {nu mi jmive} are equivalent,
>
>What would the difference be?

{mi jmive} is the statement that "I live/lived/will live";
{nu mi jmive} is the period of time during which "I live" is
true, treated as an entity; that's simplistic, but is pretty
much the way I understand it.  It feels like there is more to {nu}
than just the time interval and ability to sequence, but I'm not
sure how to describe it... I never really questioned what {nu}
meant to me until I had to explain why it was different from
your usage.  Certainly there is a sense that {nu} events are
not permanent, since they have a beginning and an end (even if,
as in {le nu munje}, the duration is infinite.)
...
After leaving this message half-written for a few hours, it seems
to me that {nu} really refers to the tense of the bridi; typical
usage restricts that to timelike aspects, but spacelike aspects
could probably fit as well if the context supported it.  Since {nu}
is the most general event abstractor, aspects such as achievement
are still implied within {nu}, but only as a part of the whole.
So, {nu} is saying that 'there's a bridi, and that bridi has a tense,
and I'm talking about that tense'.

>> although I do see that they both differ from {le du'u mi jmive}.
>> I can see that the predicate formed from {mi jmive} with {du'u} can
>> unify with both {mi jmive} and {nu mi jmive}; it could unify in some
>> sense with {ka mi jmive} or {ni mi jmive} as well, since those
>> are all aspects of the same base bridi, but the only thing that
>> really _matches_ {le du'u mi jmive} is {mi jmive}.
>
>I'm not clear what "matching" or "unifying with" mean. Can you
>explain again?

Hmmmmm... my usage was sloppy.  'Unification' as I meant it is a
pattern-matching algorithm used in artificial intelligence for
finding facts that match with rules, binding variables to specific
facts; it's something I'm aware of but not an expert in.  My intent
was to describe how I thought bridi in the environment could satisfy
the predication, without being rigorous about the process.

So, {le du'u mi jmive} is a predication, and the only fact/object
that makes it true is {mi jmive}.  Abstractions such as {nu mi jmive}
can satisfy that predication to some degree, in that an abstraction
suggests or implies the existence of the unabstracted bridi (but does
not state it, so there might not be an actual referent.)

>> Now that I think about it, how does the {du'u} abstraction differ
>> from the {brodX} gismu family, which construct predicates?
>> Shorthand?
>
>I forget.

You're forgiven.  :)  Maybe it's analogous to the relationship
between {nu} and {fasnu}.

...
>It's very useful to have an abstractor to talk about the
>realization of a bridi in space-time.

Hmmmmm... I think I can see your point, although I still see {nu}
more as a reference to the tense envelope of a bridi rather than
the instantiation of that bridi within the envelope; I usually
expect the unabstracted bridi to refer to the object itself.

Maybe I _am_ seeing a difference between {nu} and {lenu}, if {nu}
is applied to the bridi rather than a sumti; that {nu} would
declare that there was a tense envelope for the bridi, and thus
at least one actual instance.  Does the refgram allow that?

--
Carl Burke
cburke@mitre.org