[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lojban] Masses



Definitely not for Lojban-beginners.  There is a formal system (actually a 
family of them) which are open to a variety of interpretations: mereology (the 
part-whole relation, Lesniewski), the calculus of individuals (Leonard, Goodman, 
Quine) and plural reference (and quantification) (McKay et al).  At the systems 
core is a relation , # (not what generally used, but machine don't offer the 
usual) which is reflexive (a # a), transitive (a#b & b#c => a#c) and 
antisymmetric (a#b & b#a => a=b) and is that which holds between (depending on 
interp) members and L-sets, parts and individuals, components and plurals.  
Three other theses are also useful to know (using {} in a set theoretical sort 
of way for the moment)  {a}=a and {a{bc}} = {{ab)c) and {ab}={ba}.  Talking 
about sets makes this all very confusing, since these wholes don't behave like 
sets in almost any way (# seems to be a confusion of membership and inclusion, 
for example).  Thinking of individuals works better, if you think of an 
individual as something that can have a name or be a value of a variable.  But 
once you get that far, the last step is just to drop the {}s and take the list 
of members as itself a perfectly good individual, just no longer called a set or 
even an individual.  So, that is the theory that underlies much of xorlo (though 
xorlo was in play before its relation to this theory was noticed).

xorlo is one of the later steps in Logjam for working with a cluster of concepts 
of which one -- the most often perplexing one -- was mass. Once xorlo was in 
place and set on a firm theoretical ground, we no longer needed the various 
notions of mass, since we could reproduce the various desired effects using 
xorlo and some related notions.  But now to mention mass is to bring back to 
mind (and maybe even to discussion) all the old confusions.  The fact that CLL 
was still back in the confusion times doesn't help at all, since some folk have 
now grown up with the mess.  Hopefully that is being suppressed over at 
Lojban-beginner, but apparently not in time for you. What xorlo did eventually 
was to remove the distinction between a mass (in one useful sense of that 
expression) and the set from which it grew and move the different properties of 
sets over to different ways the one set (sorry, the set habit dies hard) relates 
to the property in question, collectively (the old mass sense) or 
distributively.  Since the distributive sense can always be expressed by the use 
of external quantifiers, there was no need for a special form for that. So the 
unquantified form could be used for the collective sense.  However, it was 
thought (and practice shows this was a correct thought), that there was also 
needed a neutral expression, unmarked for either collective or distributive.  
This role was given to simple forms, without external quantifiers.  Then the 
somewhat shakier thought came along that we didn't really need a separate form 
for collective use, since the use of the simple form was generally clear and, 
when not, the matter was easily rectified.  So there is not unequivocal 
collective form.  Sets have the properties of their members distributively and 
have proeprties of their own collectively.

1.  Yes, but it is clearer as {ro lo jenmi cu sanci}
2.  Yes (when they do in fact form an army) , clearer would be {lo sanci cu mi'e 
lo jenmi} (I am woreking blind at the moment so I may have the central relation 
wrong)
3  Absolute individuals, I guess (things such that they are their only member), 
though even that may not work with some predicates.  BTW, I think (not reliable) 
that {je} means the predicates are logically separable and so terms can have a 
collective sense with one predicate and a distributive sense with the other.  I 
think {gi'e} requires the same sense for both and should be contradictory for 
any X.
4.  Does it have to be some expression that actually mentions cats and dogs? 
Those get mess in themselves, but assuming you have done that, it will do, 
whatever it is, since the object is neither either collectively or 
distributively..  {su'o X cu gerku je mlatu} would be true, however.  (The claim 
about distributively is open to challenge).
5.  Your solution works fine, except you only need {lo sanci} at the end.
6. (still blind, so no memory of a good word for powerful) {lo jenmi cu 
[powerful]} maybe adding "but no part of it is by itself" though that is 
probably not true.


Prolix, as usual.


----- Original Message ----
From: Felipe Gonçalves Assis <felipeg.assis@gmail.com>
To: lojban@googlegroups.com
Sent: Sat, May 14, 2011 9:08:00 PM
Subject: Re: [lojban] Masses

Thanks for your clarifications John,

I must say, however, that I am only highly disturbed. Before I proceed with any
discussion, I must ask for help (move this to [lojban-beginners] if
appropriate).

I have been believing that a mass (whatever you call it) was just as much
a logical object as anything else, of a somewhat vague nature, but
uniquely associated to a specific set, and as such, to a specific cardinality.
This is clearly wrong. Where the CLL introduces lVi descriptors I read

"A mass has the properties of each individual which composes it, and may have
other properties of its own as well. This can lead to apparent contradictions."

which basically denies my suggestion that masses should not inherit the
properties of its individuals.

The CLL was my first and primary source for learning lojban, so I can only
conclude that this interpretation of masses came to my mind because I felt
the need to express such an idea.

My big question now is how should I correct my lojban writings and thoughts
to the actual definition and community usage of descriptors. For now, I would
appreciate if someone answered the following more specific questions

1. Is {lo jenmi cu sonci} usually true?

2. Is {lo sonci cu jenmi} usually true?

3. For what kind of X can I guarantee {X cu broda naje broda} to be false?

4. What kind of X can refer to a group of three dogs and two cats as a
single object
for which {X cu gerku je mlatu} is false?

5. How can I unambiguously say "The army is composed of a thousand soldiers",
as opposed to "a thousand divisions" or "a thousand limbs"?
Perhaps {lo jenmi cu ki'omei lo'i sonci}?

6. How can I unambiguously say "The army is powerful", as opposed to "it has
a powerful soldier" or "it has a powerful division"?

I'd also like to receive pointers about L-sets.

Thanks in advance for any attention provided.
I also hope this is enlightening to other lojbanists.

mu'o
mi'e .asiz.


On 14 May 2011 20:33, John E Clifford <kali9putra@yahoo.com> wrote:
> Yeah, that is a rub.  Officially (insofar as xorlo is official -- which is
> probably more so than just about anything else), {lo} terms are neutral 
between
> the distributive and collective senses (the word "mass" generates a whole 
>nother
> set of problems from Logjam history).  Using it as both in a single sentence
> seems wrong when you think about it, but perfectly natural in use: "The 
>students
> wore green headbands and surrounded the building.". "The girls dressed like 
>Lady
> Gaga but were a group of five."  There is a temptation to take [lo] terms
> without external quantifiers as representing collective use, since those with
> quantifiers are always distributive, but, convenient as that would be (to 
avoid
> questions about who carried the piano, say), there is greater convenience (it 
>is
> said) in the present system, which allows for the double use and also for 
cases
> where we just don't know where we don't know (or it doesn't matter) how the
> group pulled it off.
> But that, of course, is separate from the issue about {-mei}.  L-sets are a
> little hard to get used to thought of as sets, but one of the rules about them
> is that (abc)=((ab)c)=(a(bc))=(b(ac))=((abc)), so, there is no particular
> problem in the same things constituting a pamei, a remei and a cimei. it's all
> in how you (mentally or not) group them.  And so, of course, is the matter of
> what other predicates apply: taken one by one in the freest form the combined
> armies are soldiers, in another form, they are armies.  And another axiom is
> (a)=a, so that. like it or not, your broda is also a pamei.  Personally, I 
>would
> as soon take the language of {lo} and {-mei} as basic and not try to expand it
> out in some string of quantifiers, other nameoids, {noi} or {poi} or whatever.
> Each of these definitions seems to lack some of the simplicity and clarity of
> the original language and never seem to fit conveniently into other contexts.
>
>
> ----- Original Message ----
> From: Felipe Gonçalves Assis <felipeg.assis@gmail.com>
> To: lojban@googlegroups.com
> Sent: Sat, May 14, 2011 5:30:26 PM
> Subject: [lojban] Masses
>
> coi rodo
>
> From the talk <[lojban] "lo no">:
> 2011/5/14 Jorge Llambías <jjllambias@gmail.com>:
>> I agree with the gist of tijlan's post, but I'd like to add some 
observations.
>>
>> On Sat, May 14, 2011 at 7:56 AM, tijlan <jbotijlan@gmail.com> wrote:
>>> lo ci gerku = da poi gerku je cimei
>>> lo no gerku = da poi gerku je nomei
>>
>> lo ci gerku = zo'e noi gerku gi'e cimei
>> lo no gerku = zo'e noi gerku gi'e nomei
>>
>
> These interpretations look wrong to me. The individuals are {gerku},
> while only the mass composed of them is a {cimei}. This distinction
> is essential, so that, e.g., the referents of
> {da poi jenmi je so'imei},
> {zo'e noi jenmi gi'e so'imei}, and
> {lo jenmi je so'imei}
> can all be unambiguously understood to be an army (or armies) of
> many soldiers, instead of a lot of armies, regardless of semantic
> nuances between these expressions.
>
> Were we to accept that a mass of broda can always be described as
> broda in any of these ways, then we would have to accept that a
> mass of two armies, each one composed of a thousand men is a
> {jenmi gi'e solci gi'e remei gi'e ki'omei}. In particular, we must accept
> {remei je ki'omei} to be no contradiction.
>
> More importantly: How would I clarify that my broda is just a broda,
> and not any conceivable mass (of masses of masses... ) of broda?
> Perhaps by saying something like {broda gi'e gunma noda poi broda}?
>
> In summary, systematically assigning to a mass the properties of its
> individuals, while not logically problematic, is highly confusing, and
> requires heavy work when a common disambiguation is called for.
>
> I would say that the meaning of {lo PA broda cu brode} is closer to either
> {PA da broda .i da poi broda cu brode} (CLL), or
> {PA lo broda cu brode} (xorlo only), or
> {zo'e noi gunma PA da poi broda cu brode} (xorlo only).
>
> What do you think?
>
> mu'o
> mi'e .asiz.
>
> --
> You received this message because you are subscribed to the Google Groups
> "lojban" group.
> To post to this group, send email to lojban@googlegroups.com.
> To unsubscribe from this group, send email to
> lojban+unsubscribe@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/lojban?hl=en.
>
> --
> You received this message because you are subscribed to the Google Groups 
>"lojban" group.
> To post to this group, send email to lojban@googlegroups.com.
> To unsubscribe from this group, send email to 
>lojban+unsubscribe@googlegroups.com.
> For more options, visit this group at 
>http://groups.google.com/group/lojban?hl=en.
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to 
lojban+unsubscribe@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/lojban?hl=en.

-- 
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.