Message-Id: <9107011441.AA06755@relay1.UU.NET>
From: cbmvax!uunet!ctr.columbia.edu!shoulson
Date: Mon Jul  1 11:30:50 1991
To: nsn@ee.mu.oz.au
Cc: lojban-list@snark.thyrsus.com, nsn@ee.mu.oz.au
In-Reply-To: nsn@ee.mu.OZ.AU's message of Mon, 01 Jul 91 14:57:04 +1000 <9107010457.AA16596@munagin.ee.mu.OZ.AU>
Subject: BAI nightmare
Status: RO

Nick makes a good point, bringing up a cmavo noi I'm frankly surprised that
ke'a wasn't invented earlier, namely co'e.

There are a million times when one needs to talk about doing or being
something unspecified, where co'e fills the bill just right.  I recall
seeing in earlier messages on the list sentences which used gasnu to fill
this need.  In some sense, maybe that's okay, but I think an explicit
elliptical selbri is *really* important, and I'm glad lojbab put one in.
There are some weirdnesses inherent in it, though.  What's its place
structure?  Likely it doesn't matter.

mi troci le nu mi co'e le vorme
I try the event (I do-something the door)

Clearly, this is scads better than mi troci le vorme, but we're a little
unclear on the relationship of the abstracted event.  Not what the
relationship is, that we're supposed to be unclear on.  But *how* the
relationship is is also fuzzy.  See, you can't really make general
statements about the places of a lojban selbri.  Oh, sure, *often* the x1
place is the "actor" and the x2 place is the "object" and other places
serve other roles, but that's not always true, and is certainly not the
lojbanic way of seeing it.  There's nothing necessarily "active" about the
x1 place or anything of the sort.  We seem to imply by "mi co'e le vorme"
that some predicate is going on in which I do something to the door, but
that's dangerous.  It's also not safe to say that there's a predicate which
fits in place of co'e such that "mi" is the x1 and "le vorme" is the x2,
since that may not be the case.  For all I know, the selbri I mean takes
"le vorme" in the third place and "mi" in the second.  I suppose that's not
so bad, as "co'e" can be assumed to have some arbitrary string of SE
conversions which make everything all right.  And anyway, even if there is
*no* selbri which indicates the relationship we using "co'e" to ellipsize,
that's okay, because the speaker either knows what we mean, doesn't care,
or can ask for a tedious, painstaking explanation.

Anyway, I still like co'e.  A lot.  Now, what was that other new elliptic?
Oh, do'e, for unspec. BAI.  This is good too, though I haven't found myself
thinking in terms of it, as I do co'e.  whatever.

As to the real import of Nick's message (or at least, that part of it which
didn't go over my head), I don't really know what to make of tu'a.  I was
kinda lost in lojbab's introduction of it, and I think co'e makes a much
better plan.  And then the use of kau (after looking it up...) semms to
fill some other bills very well.  I really don't know about tu'a.

Rest of Nick's letter will have to wait until I can understand it.

~mark