Re: [lojban] An approach to attitudinals

[pycyn@aol.com]

In a message dated 6/10/2001 9:02:04 AM Central Daylight Time,
richardt@flash.net writes:

.  But anyway, if you want to say exactly what you mean, use a bridi.
> Attitudinals are for expressing how you feel, not for saying what you
mean.

If I'm understanding you correctly, you're saying that:

.ui ko'a snada
.a'o ko'a snada

both mean the same thing, but tell you different ways that I feel about it.
 If {a'o} altered the truth of sentence two, I would be using it to say
what I meant, and not just how I feel.  Assuming
that's what you're saying I agree with you completely.  

However, others have already made the claim that sentence 2 does change the
meaning of the sentence (and thus they _do_ use attitudinals to say what
they mean, at least in their view).  I was
suggesting adding a suffix as a way of keeping both meanings, such that we
could express both without the listener either:

1) having to guess
2) having to memorize which attitudinals are interacting and which are not.

Barring that, I'd be happiest if there were a list of which attitudinals do
and do not interact with the (truth value/meaning) of sentence.  Then, when
someone misunderstands me, I can point to the
list and say, "Nope, attitudinal number 206 doesn't do that.  Study harder."

The attachment idea would be useful if most attitudinals had the double
function Lojbab ascribes to hope.  I don't think they do, though the hope
case sets a pattern that could be repeated in a number of other cases as
well. So, until the hope case is dealt with, we'll keep your suggestion in
mind.
As for the lsit, we're working on it.  Originally it was supposed to be
pretty simple.  Starting from afew words thart sounded a lot like English
exclamations (Whee, oy, ow,...) the list from each initial letter was
expanded with others that were like the core one in their logics.  But, of
course, too many turned up for one category and too few for another, so some
lists were mixed.  And too many emotions and the like came to light, so we
started using derivatives of gismu.  And now there really is no quick way
other than memory to tell which does what (association with a gismu often
helps a bit). And now we are also finding that some items we thoiugh were
comfortably placed are in fact being used for other things.  So the whole
needs some (unofficial because of the freeze) rethinking.
      Just by the way, I think your first examples read as follows: {ui} is
purely a react to his succeeding, which he has done (or not -- in which case
the whole is false).  {a'o}, as xorxes says, presupposes that he has not
succeeded (or, at least that the speaker does not know he has) and the whole
is not truth-valued at all (precative function, not informative), related to
imperatives, perhaps, but without an addressee and without a specific action
being called for.  I don't see it as claiming that he has succeeded and
therefore being hopeful that something else (unspecified but what the speaker
takes to be good) will happen.