From Pycyn@aol.com Tue Jan 25 02:17:09 2000
X-Digest-Num: 344
Message-ID: <44114.344.1841.959273825@eGroups.com>
Date: Tue, 25 Jan 2000 05:17:09 EST
From: Pycyn@aol.com
Subject: Re: Subjunctives

I can't check with the Book right now, but my memory of how the tense and 
aspects were intended to work is as follows.
1) tense and aspect are separated (and from mood and the like as well, all of 
which are jumbled together in English and, historically at least, in most 
familiar languages)
2) tense is based on axis and vector reference, though not restricted to the 
four axes implicit in most natural systems (though never fully realized in 
any).
3) The same cmavo are used for retro vector and past axis, for simultaneous 
vector and present axis, and for pro vector and future axis. The differences 
are positional and/or determined by context. Thus, pu might be either past 
vector to the current axis or establishing a new axis prior to the current 
one. Where the difference is too important to be left to context to decide, 
capu would indicate the vector (brief glance back without chaning the focus 
of the narrative) and puca the new axis (or maybe it is the other way 'round 
-- I ermember this got argued and I forget which one, this seems most natural 
to me at the moment).
4) Only axes are points, so that punctile clauses, like ca..., must apply to 
axes, making bapu ca... unambiguous "event before the future event indicated 
by ca..."
Of course, with a clearly future event, maybe even capu would work.
5) Aspects carry temporal implications but are not completely temporal. 
Thus, the perfective of an event does entail that the event occurred in the 
past (though even this can be doubted, since some maintain that the 
inchoative does not entail that the event takes place in the future) but the 
converse does not quite follow, for not all past events still throw their 
aspectual shadows into the present (effects from causes, continued existence 
of participants, ... -- the list varies in some unclear ways) as the 
perfective seems to indicate. 
6) So, in Lojban, past axis, retro vector and perfective aspect are all 
slightly different and in different ways. But in English they tend to fall 
together, certainly away from present tense, and thus sorting out which one 
is meant by a given Englsh sentence is not subject to clear rules, except 
that one must think what one means to say, both in the given sentence and in 
those around it.

The perfective seems to involve relevance conditions, which are one of the 
hairiest problems in possible world games (of which tense is one in the logic 
business). For the contrary to fact cases being discussed, the best course 
is to say every world exactly like this one except for the condition named in 
the protasis (if....) and whatever is required by that change. So, clearly, 
changing the world by having me possess a million just requires that I also 
be shifted into the class of rich folk (I think -- a million just ain't what 
it was anymore) and maybe nothing or very little else. Or does it: can I 
have a million and still be a retired professor from a really cheap 
university? Don't have to have had some source for that million and if so 
what? So maybe the worlds can vary on the ways I got the million. But if 
they vary too much, I come to doubt that this is still me they are talking 
about. And, if I start to vary too much, does this not affect others around 
me (wives and childen, etc., at least and students and colleagues and....). 
Where does it end? Cut it off too soon and the world so little changed is 
greatly changed (a retired professor gets a mill out of the blue); let it run 
too far and it no longer seems to apply to me (or to be about this sort of 
world at all). 
Probably all that the original really means is that anyone with a million is 
rich, perhaps with the added wish that I were one such. (And, of course, if 
I were as rich as Rothschild, I'd be richer than Rothschild.)
Talk of possible worlds really brings up a point about my favorite (and 
everybody else's least favorite) change, restricted quantification. As 
Xorxes points out, "for every possible world w, if I have a million in w, 
then I am rich in w" could be true just because there is no possible world in 
which I have a million -- hardly an improvement on the material reading in 
this world. On the other hand "in every possible world in which I have a 
million, w, I am rich in w" looks only at the worlds in which I have a 
muillion -- and says that there are some. Clearly the latter is much closer 
to what is wanted, though even it may not be quite right (Lojban has the 
means to do this, but does not use it for this purpose). 
pc