Re: tenses

[jorge@PHYAST.PITT.EDU]

Chris:
> You are arguing by repeated assertion now.

Well, it worked to convince And. :)

> You say "pu'o" means "start
> to.." in the pragmatic way it's used in English;

That's {co'a}. {pu'o} is "is about to...", it may never even start.

> others are saying it only
> applies to an event which actually ends up occuring, just like the facade
> and the church in And's example.

Yes, but the facade example is different. If you have a selbri
"x1 is a facade of x2", then clearly whatever you put in x1 and x2 has
to be something. This would be analogous to a selbri like "x1 is the
start of x2", in which case x1 and x2 have to be something.

But {co'a} is not a selbri by itself. {ko'a co'a broda ko'e} does not
mean "<something> is the start of <event ko'a broda ko'e>".

The selbri is {co'a broda} and it relates the two arguments {ko'a} and
{ko'e}. There is no logical necessity that {co'u broda} or {broda} or
any other selbri also relate those two arguments.

> I don't think we'll get beyond this point
> until someone starts talking about why their interpretation is more elegant
> or useful or consistent than another interpretation.

Well, there is nothing inconsistent with either interpretation, as far
as I can tell. I think mine is elegant and useful because it let's
you say things like "I start to eat an apple" without claiming that
I will eat the whole apple, or things like "I was about to leave when
the phone rang" without claiming that I did leave after all. How would
you say that with the "extensional" view? (Thank you pc for the terminology.)

        mi pu'o cliva ca le nu le fonxa co'a janbe
        I was about to leave when the phone started to ring.

That is useful and elegant. If you need to assert that the event
actually does happen you can use {baco'a}. For example:

        mi co'a cliva ba le nu le fonxa co'a janbe
        I started to leave after the phone started to ring.

> If you don't finish, you didn't really count to ten and {co'a} causes the
> bridi to refer to the counter's intentions rather than their real acts.

Not necessarily the intentions, but the event as understood by whoever
describes it. For example, in the following situation:

Mary: Please count to ten.
John: One, two...
Mary: Stop there!

I can say "John started to count to ten, but Mary stopped him before
he could finish". I don't really know whether John intended to finish
or not, but from the context I can reasonably say that he did start
to count to ten. Even if he stopped by himself, I could still say
that he started to count to ten but then for some reason didn't finish.
If he keeps counting after ten then I might use {za'o}, because in the
context the natural end was at ten.

> This could be perfectly consistent and workable, but *why*?  The event
> contours would strike me as being more logical if they did not affect the
> truth value of the selbri, except to indicate when something happened.

They don't affect the truth value of the selbri, but a selbri with {co'a}
is not the same selbri as one with {co'u}, etc. They state a different
relationship among the arguments.

>  I support an additional
> understanding that says you can't pu'o citka le plise without later taking a
> single bite.  The advantage of that rule is that you could always logically
> derive  {da broda de} from {da contour broda de}.

But that's a disadvantage! You are restricting the meaning of {pu'o citka}
to those situations before actual eating, rather than all the situations
of about to eat, which include those that for one reason or another don't
follow their natural course.

> We can't always make
> inferences in the language based on syntactic manipulation like that (for
> example, I don't think you can say anything about individuals from mass
> statements, or vice versa) but if we provide that property to the language
> whenever we reasonably can, it contributes to the claim of Lojban being a
> logical language.

But what's the point of being able to make logical inferences from claims
that we don't want to make? In fact, {pu'o} is probably at its most useful
when the expected event ends up not happening.

BTW, do you recognize these:

ritas: sei co'a xanka ti prane kanro .i le mikce pu pu'o catra ti
ritas:  ko ba'e tirna ti .i ba'e na ruble to sei pu'o klaku toi .i .ainai mi

I think that in both cases the event did not end up happening...
Usage supports my view!  :)

Jorge