From @YaleVM.YCC.YALE.EDU:LOJBAN@CUVMB.BITNET  Tue Mar 16 13:03:21 1993
Received: from YALEVM.YCC.YALE.EDU by MINERVA.CIS.YALE.EDU via SMTP; Tue, 16 Mar 1993 08:19:14 -0500
Received: from CUVMB.CC.COLUMBIA.EDU by YaleVM.YCC.Yale.Edu (IBM VM SMTP V2R2)    with BSMTP id 1352; Tue, 16 Mar 93 08:18:07 EST
Received: from CUVMB.BITNET by CUVMB.CC.COLUMBIA.EDU (Mailer R2.07) with BSMTP  id 9556; Tue, 16 Mar 93 08:19:14 EST
Date:         Tue, 16 Mar 1993 13:03:21 GMT
Reply-To: C.J.Fine@BRADFORD.AC.UK
Sender: Lojban list <LOJBAN%CUVMB.bitnet@YaleVM.YCC.YALE.EDU>
From: C.J.Fine@BRADFORD.AC.UK
Subject:      TECH:" more on *mo'u
To: Erik Rauch <erikr@MINERVA.CIS.YALE.EDU>
Status: OR
Message-ID: <5YzXFPU-0DD.A.a-.R30kLB@chain.digitalkingdom.org>

I've thought some more about Bob's "I in English and you in Lojban discuss".
I agree that this is something we can't satisfactorily do at present, but
I am adamant that the solution should not bring the "bau" into the x1.

Consider a case where the language is a tergismu:

        mi tavla do ti la lojban .ije dy tavla do ti la gliban

        I talk to you about this in Lojban, and D talks to you about this in
        English.

This clearly contracts to

        nu'i mi fo la lojban nu'u .e dy. fo la gliban cu tabla do ti

        I in Lojban and D in English talk to you about this.

But this is a logical connection, and says nothing about whether we do it
jointly.

The approximate non-logical equivalent is

        mi tavla do ti la lojban .ijo'u dy. tavla do ti la gliban

        ('approximate' because the meaning of non-logical sentence
        connectives is not entirely clear, and they cannot be
        mechanically transformed into non-logical sumti or bridi
        connectives like logical ones)

This strongly suggests that what we need is, as already proposed,
non-logically connected termsets:

        *nu'i mi fo la lojban nu'u jo'u dy. fo la gliban. cu tavla do ti

and this will go straight over to the casnu example:

        *nu'i mi bau la lojban nu'u jo'u do bau la gliban. cu casnu

(Note that there is another way that probably achieves this, using fa'u:

        mi ce do casnu bau la lojban fa'u la gliban.

but I'm not certain this is extensible).

I am actually beginning to wonder whether the neat collection of what in
Loglan are several different selma'o into a single BAI is giving us more
problems than it solved. It is clear that in practice most BAI fall into
separate classes, those typically used as sumti tcita, and those
typically used as selbri tcita; and I have previously pointed out other
inconsistencies among them. I wonder if the present discussion is not
showing that "mau" is sufficiently different from the others that it should
be treated differently. In a way, this is endorsing Greg's original suggestion,
but I am firm that neither "mau" nor "bau" should be treated as a connective,
but for very different reasons.

        Colin