Re: [jboske] gadri

[Robert LeChevalier <lojbab@lojban.org>]

At 04:01 PM 12/27/02 +0000, Jorge Llambias wrote:
> la nitcion cusku di'e
> >I note in passing that the gismu definitions do treat sets as n-tuples,
> >just as Bob's addled recollection leads me to believe. (No, I don't
> >like what he did. Because I'm stuck trying to clear up the mess. And I
> >don't give a fuck if the mess is originally James Brown's.) Why else
> >use sets in the gismu definitions where he did?

Lojbab attempted to distinguish sets from collectives (and I won't try to 
figure out what an n-tuple is in this context - I would have presumed it to 
be a set comprised of n members which are ordered, but that makes no sense 
in the context).

> I don't think the mess is JCB's. JCB always used "set" for
> collectives, and when the logicians tried to restrict "set"
> to mathematical sets he vetoed them. In Loglan it is "leu"
> (called "set") that carry the log, and as far as I know
> it has always been and remained so.

It hasn't. That was JCB revisionism in the two articles on masses from 
1995 posted on the www.loglan.org website. From TL 4/3 (November 1980), in 
which pc summarized (with JCB concurrence) the additions to the language 
from 1975-80 in what was stated to be an official supplement to Loglan I:

"
(Lua and Lue) Each forms a designation of a set or a class as such rather 
than of their members, as is the case with le and lea [lea in lojban is ro 
lo]. Lue is analogous to le in that it is "intentional". It designates a 
particular set which the speaker "has in mind" by his mentioning one or 
more of the properties shared, or apparently shared, by its members. Thus 
lue mrenu [lojban le'i nanmu] means 'the set composed of men, or apparent 
men, which I have in mind'. lu'a [lojban lo'i], in contrast, is like 
lea. It designates the set composed of all who have the described property 
or properties. Lua mrenu designates the set composed of all (past, present, 
and future) men.

To be able to designate sets rather than members of sets is to permit one 
to say things about their size, their inclusion in or exclusion from other 
sets, and in other ways to treat them mathematically, logically, and in 
science, taxonomically. Set designation is not a regularly handled feature 
of natural language. But in Loglan we require this device to rescue us 
from the referential ambiguities that plague most natural language 
discussion of these objects by virtue of the ad hoc arrangements that have 
been adopted ...
"

Clearly the word was added to refer to mathematical sets (and I believe it 
was added by pc, but that would take more digging to verify). 4th edition 
Loglan 1 confirms this in its examples, though less clearly:
http://www.loglan.org/Loglan1/chap4.html#sec4.20
(that section also defines JCB's intent for loe[lo'e] which has been 
consistent through the years)


http://www.loglan.org/Loglan1/chap4.html#sec4.9
incidentally, discusses the understanding of masses in TLI Loglan prior to 
the revisionism of 1995 (which if I recall correctly led to the resignation 
of Randall Holmes, the TLI logician, who could not accept the conflation of 
mathematical sets and their referents, and the rejection of the 
mathematical set descriptor as being important).

That text clearly shows that JCB had in mind that masses encompassed 
substances, "-kind", and the Trobriand Islander concept, though he backed 
away from the latter in 1995.

Note also that JCB and pc apparently understood the "in mind" aspect of le 
(and le'e) to be +intentional rather than +specific; I'm not sure when le 
became -opaque, but suspect that it came from a discussion involving these 
same parties (Cowan, And, and Jorge) in 1994 including
http://balance.wiw.org/~jkominek/lojban/9411/msg00073.html
I suggest people review the discussions of that era in the archives, 
especially looking for "opaque" and "specific" to see where this debate got 
started (if you also look for "Iain" you will find where the abstraction 
place in sisku came from), and perhaps find that we resolved it already.

> The problem came when Lojban instituted the new "mathematical
> set" category, and then displaced the collective function to
> "mass". The collective function, being much more frequent than
> the substance function, eventually came to be the canon for
> {lei}, and the description of "mass" became the messy conflated
> thing.

Reviewing the history, I think that the collective function has been 
largely ignored or conflated with [ro] lei, which for JCB was merely his 
equivalent of loi with a restriction or a quantifier. Lojban made an 
attempt to cover it with girzu (which is explicitly distinguished from the 
mathematical set which is in x3) as well as jo'u. I don't mind breaking it 
out in additional cmavo, since we did a partial job with the joi/jo'u 
distinction.

> >Bob says the gismu for collectives might be kampu, The x2 of kampu
> >is... a set! Ergo, Bob thinks of sets as collectives/tuples. Thanks a
> >bunch.
>
> And x1 of kampu is a property, so it is not clear what it has
> to do with collectives. From the definition I would have guessed
> that x1 is a property that each member has, rather than an
> emergent property of the group as a whole.

I think you are partially right - I think kampu could refer to either 
individual or emergent properties (and I could be wrong, but I think that 
ce'u doesn't work properly in a kampu x1, but that is another issue)

I think girzu is the gismu for collective, since it clearly distinguishes 
the group (x1) from the set (x3). I'll admit to fuzziness at the moment as 
to the distinction between x2 and x4 of girzu. The history is that there 
was one place that conflated a defining property and membership (which were 
accidentally numbered x3 instead of x2). Someone noticed this at the same 
time that we were also dealing with sumti-raising in place structures, and 
I decided that this sort of conflation was not acceptable, so we expanded 
the one place into 3 places, a defining property, a complete set 
membership, or a relation (du'u) that the group shared in common giving 
each a separate place. The wording was probably fouled up at some point 
because we forgot this. Whether this form of breakout was the Right Thing 
To Do (and whether the three separate places still make sense given the 
emergence of ce'u), it appears that x1 is the collective and definitely is 
NOT the set (which is x3), and one of either x2 or x4 is supposed to be the 
defining property of the set, and the other, the emergent property of the 
group. The clumsiness of the wording certainly allows us to Make It So 
when the byfy acts.

lojbab

--
lojbab lojbab@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org