Message-Id: <m0qTdIQ-000023C@xiron.pc.helsinki.fi>
Date:         Thu, 28 Jul 1994 17:46:24 -0400
Reply-To:     Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Subject:      Message from Colin FIne
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 2972
Lines: 85

The following message by Colin Fine was rejected by the listserv because
it contained a *quoted* "Sender:" header.  Don't do that, folks.  The
software overdoes it in an attempt to avoid mail loops.

Colin's text starts here:

Jorge says (to Randall):

> If you want literally that, then: (cmima lu'i <argument>) is just that.>
>
> cmima mim cmi      member           x1 is a member/element of set x2;
>
> lu'i      LAhE     the set composed of
>
> I still don't see a problem with {lu'i le nanmu} for "lea me le mrenu".
>
> I don't think {lu'i le nanmu} is a set with one element "the men", but
> rather the set that has "all the men" as elements.
>

Randall answers:

> I don't think so.  Your phrase simply means "set containing the man",
> and will have plural reference to sets (not necessarily the same)
> containing each man who is desginated by "the man" if "the man" has
> multiple reference.
>

Jorge comes back:

> I don't know how "le" works in Loglan, but in Lojban its default
> quantifier is "ro" = "all". Then {lu'i le nanmu} means the same
> thing as {lu'i ro le nanmu} and cannot be a set with a single man
> as an element if {le nanmu} refers to three men. It means "the set
> of all the men I have in mind". I hope John Cowan is reading and
> will correct me if I'm talking nonsense.
>
> Jorge
>

I think Jorge is right, but that explanation is not sufficient.

In my understanding, Randall is right that 'le' (in either language,
and also 'lo' and 'la' in Lojban) is distributive, making the claim
independently of all the (subjectively in the case of 'le') designated
entities.
Therefore, Jorge's last point is not germane, or at least insufficient:
 it doesn't matter whether
the quantifier is 'ro' or anything else, because on the face of it the
term is still distributive.

I believe however, that there is special semantics in the conversion
operators such as "lu'i". They 'construct' a new multiplicity out of the
objects designated by the contained sumti, and they are opaque to
distribution - hence the quantifier becomes relevant.

So

        [ro] le nanmu cu xagji = [each of] the men are/is hungry

but le'i (approximately TLI 'lea' I think) constructs a set:

so      lu'i le nanmu cu se cimei
        (*lea mrenu ga nu tera, but I don't know if -ra has that place
                structure so I've starred it)

does mean "the set of (all) the designated men is a set of cardinal 3"

I thought at first that "le'i nanmu" meant the same thing, but now I think
it means "the designated sets of men" - I'm not sure, though.

In the same way,
        "lu'a le klesi" means "[all] the members of the designated classes",
distributively, and
        "lu'o le nanmu" means "the mass of [all] the designated men"
(which has properties of some of the men, properties of all of the men,
and properties of the mass)

Colin Fine


--
John Cowan              sharing account <lojbab@access.digex.netfor now
                e'osai ko sarji la lojban.