From pycyn@aol.com Tue Sep 10 09:06:09 2002 Return-Path: X-Sender: Pycyn@aol.com X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-8_1_0_1); 10 Sep 2002 16:06:09 -0000 Received: (qmail 76884 invoked from network); 10 Sep 2002 16:06:09 -0000 Received: from unknown (66.218.66.218) by m5.grp.scd.yahoo.com with QMQP; 10 Sep 2002 16:06:09 -0000 Received: from unknown (HELO imo-d10.mx.aol.com) (205.188.157.42) by mta3.grp.scd.yahoo.com with SMTP; 10 Sep 2002 16:06:08 -0000 Received: from Pycyn@aol.com by imo-d10.mx.aol.com (mail_out_v34.11.) id r.182.e1fa1b9 (4529) for ; Tue, 10 Sep 2002 12:06:02 -0400 (EDT) Message-ID: <182.e1fa1b9.2aaf7269@aol.com> Date: Tue, 10 Sep 2002 12:06:01 EDT Subject: Re: [lojban] Re: I like chocolate To: lojban@yahoogroups.com MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="part1_182.e1fa1b9.2aaf7269_boundary" X-Mailer: AOL 7.0 for Windows US sub 10509 From: pycyn@aol.com X-Yahoo-Group-Post: member; u=2455001 X-Yahoo-Profile: kaliputra --part1_182.e1fa1b9.2aaf7269_boundary Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit In a message dated 8/22/2002 3:53:52 PM Central Daylight Time, jjllambias@hotmail.com writes: << > > > lo du'u la djil sipna kei naku se djuno la djak > > > =/= That Jill is asleep is not known by Jack. > > > The Lojban is true and the English false. > > >> > >I agree they are not equal, because the English us a translation of {lo > >du'u > >la djil sipna kei na se djuno la djak}, inversion and denial of {la djak > >djuno lo du'u la djil sipna}. > > That's arguable, because negation in English does not > normally have such scope: "Some things are not known by Jack" > is not an inversion and denial of "Jack knows some things". > My suspicion is that "that Jill is asleep" is a singular > term in English, not an existentially quantified one. > >> Well, if "that Jill is asleep" is a singular term in English (which seems plausible), then the scope of the negation is irrelevant, but the problem is about Lojban, where it is relevant. And the question then is which meaning is involved if "that" clauses are quantified. The straightforward quantified case does not seem to me to be very informative, since the English rules for quantifiers and negations are rather different from the Lojban. The crucial thing seems to me to be that, if Jack does not know that Jill is asleep, then he does not know any proposition that amounts to that claim, not just that he does not know some particular ones of that sort -- but does know others. That, of course, relies on my understanding of what {du'u la djil sipna} means, and you may (and, apparently, do) disagree with that. However, it is a predicate, so intended to apply to many things (even if it may not live up to intentions) and I see no reason why it does not do so here. << >there >is no quantifier place, quantifiers are optional and so we can just leave >them off. Can we really? Can we use {lo broda} and NOT mean {su'o lo broda}? >> Yes, but it takes special circumstances, because the *convention* is to take {lo broda} as {su'o lo broda}. The *grammar* does not require quantifiers (what it means to say that they are optional). And, no, I cannot whip up a plausible case on the spur of the moment -- and, more importantly, in the limited context of an e-mail message. << >So {lo'i nu broda} sometimes is the set of type {nu broda}s and >in other contexts it is the set of token {nu broda}s? Or is >it always a set that includes both type and token {nu broda}s? >(My position is that it is always a set of token {nu broda}s >only, independently of context.) > >> >An interesting position, but how would you support it from the text? Or in >practice? I'm not sure about supporting it from the text, there may be support for any position in it, but I'm not an interpreter of the true doctrine, I create my own. :) >> Not really a smiley matter, since it tends to confuse the issue of how to do it in Lojban rather than Xorban. Returning to that (the nominally specified topic here): << As for practice, you know how I do it: I use {lo'e} for the types and {lo/le} for the tokens. ({le'e} is also a type, obviously, but I'm not sure yet how it would/could/should work in my scheme.) >> I am not sure that this is how I would describe what you do, but that is largely because I have yet to get a clear grip on (or a clear statement from you about) what your usage amounts to. This descriptions seems different from others you have presented, but that may be just the differences in the predicates involved. If I read what you say literally (which is probably grossly unfair -- but no more so than expecting someone to work out what the metaphor is wihtout some serious help) then lo'e nu broda is a/the (?) type of which all members of lo'i nu broda are tokens and lo/le nu broda is/are particular token(s) of that or some type of lo'i nu broda. This leaves the question of how to refer to the members of lo'i nu broda itself. If, as it seems you have it, they are just the tokens themselves, then they are not nu broda, since that is abstract and the token events are concrete -- which goes against {lo} anyhow, if not {le} (which might, I suppose, be systematically used for specific concrete token events). In *your* scheme I suppose that {le'e} is to {lo'e} as {le} is to {lo} -- though that still leaves the impression that there is more than one type under the heading {nu broda}, so I probably have that wrong. << >I think that lo'i nu broda is a set of abstracts always and that we can use >reference to these abstracts to refer to occassions tht full under them. But why not do the same for dogs and chocolates then? Why are they priviledged (or underpriviledged rather) with respect to events? Isn't one of the claims of Lojban that they are treated the same? >> I am not sure I follow this at all. Dogs and chocolates are different from events in that they are concrete and events are abbstract: or rather {gerku} and {cakla} doe not claim on their faces to be abstract (and so, by default, are assumed to be concrete, since good Gricean manners would require noting such a significant feature) while {nu broda}, etc. do have this on their faces. I am unclear here what the privilege involved is. The whole point of my argument is just that like things be treated alike: classes are classes of things satisfying the predicate defining the class, so members of classes of concreta are concrete, those of abstracta are abstract. I suspect a deep (too deep for me to dig out so far) confusion here: maybe between concrete and abstract v. token and type, maybe between classes and other abstract entities, maybe something else entirely . And I am not sure which of us is confused -- or whether both of us in antagonistic ways. << >(how do you build types out of tokens unless >you somehow have a criterion for their being "the same." i.e. have a type >in >utero at least.) The type (the intension) is in the meaning of the word, of course. The question is how you access it. Not with lo/le, I say. lo/le give you the extension, what you quantify over, the members of the class one by one. >> Or maybe between type and intension and meaning (sense)? I don't get this at all. Quantification does not enter ere at all in any interesting way. Remember, in Lojban, all entities are on the same level in the grammar, so it is as easy to quantify over the members of a set of types or intensions (not necessarily the same thing, by the way) as of a set of dogs. One hopes that the type of a thing is not in the meaning of the word for that thing, because that will lead to an infinite regress, circling forever around to get out the type, given only tokens. Unless you wan to fall back on the bootstrapping empirical procedure of science and leave a forever incompletely specified type -- which makes talk of types pretty senseless against other locutions. --part1_182.e1fa1b9.2aaf7269_boundary Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: 7bit In a message dated 8/22/2002 3:53:52 PM Central Daylight Time, jjllambias@hotmail.com writes:

<<
> >      lo du'u la djil sipna kei naku se djuno la djak
> > =/=  That Jill is asleep is not known by Jack.
> > The Lojban is true and the English false.
> >>
>I agree they are not equal, because the English us a translation of {lo
>du'u
>la djil sipna kei na se djuno la djak}, inversion and denial of {la djak
>djuno lo du'u la djil sipna}.

That's arguable, because negation in English does not
normally have such scope: "Some things are not known by Jack"
is not an inversion and denial of "Jack knows some things".
My suspicion is that "that Jill is asleep" is a singular
term in English, not an existentially quantified one.
>>

Well, if "that Jill is asleep" is a singular term in English (which seems plausible), then the scope of the negation is irrelevant, but the problem is about Lojban, where it is relevant. And the question then is which meaning is involved if "that" clauses are quantified.  The straightforward quantified case does not seem to me to be very informative, since the English rules for quantifiers and negations are rather different from the Lojban.  The crucial thing seems to me to be that, if Jack does not know that Jill is asleep, then he does not know any proposition that amounts to that claim, not just that he does not know some particular ones of that sort -- but does know others.  That, of course, relies on my understanding of what {du'u la djil sipna} means, and you may (and, apparently, do) disagree with that.  However, it is a predicate, so intended to apply to many things (even if it may not live up to intentions) and I see no reason why it does not do so here.  

<<
>there
>is no quantifier place, quantifiers are optional and so we can just leave
>them off.

Can we really? Can we use {lo broda} and NOT mean {su'o lo broda}?
>>
Yes, but it takes special circumstances, because the *convention* is to take {lo broda} as {su'o lo broda}.  The *grammar* does not require quantifiers (what it means to say that they are optional).  And, no, I cannot whip up a plausible case on the spur of the moment -- and, more importantly, in the limited context of an e-mail message.

<<
>So {lo'i nu broda} sometimes is the set of type {nu broda}s and
>in other contexts it is the set of token {nu broda}s?  Or is
>it always a set that includes both type and token {nu broda}s?
>(My position is that it is always a set of token {nu broda}s
>only, independently of context.)
> >>
>An interesting position, but how would you support it from the text?  Or in
>practice?

I'm not sure about supporting it from the text, there may be
support for any position in it, but I'm not an interpreter
of the true doctrine, I create my own. :)
>>
Not really a smiley matter, since it tends to confuse the issue of how to do it in Lojban rather than Xorban. Returning to that (the nominally specified topic here):

<<
As for practice, you know how I do it: I use {lo'e} for the types
and {lo/le} for the tokens. ({le'e} is also a type, obviously, but
I'm not sure yet how it would/could/should work in my scheme.)
>>
I am not sure that this is how I would describe what you do, but that is largely because I have yet to get a clear grip on (or a clear statement from you about) what your usage amounts to.  This descriptions seems different from others you have presented, but that may be just the differences in the predicates involved.  If I read what you say literally (which is probably grossly unfair -- but no more so than expecting someone to work out what the metaphor is wihtout some serious help) then  lo'e nu broda is a/the (?) type of which all members of lo'i nu broda are tokens and lo/le nu broda is/are particular token(s) of that or some type of lo'i nu broda.  This leaves the question of how to refer to the members of lo'i nu broda itself.  If, as it seems you have it, they are just the tokens themselves, then they are not nu broda, since that is abstract and the token events are concrete -- which goes against {lo} anyhow, if not {le} (which might, I suppose, be systematically used for specific concrete token events). 
In *your* scheme I suppose that {le'e} is to {lo'e} as {le} is to {lo} -- though that still leaves the impression that there is more than one type under the heading {nu broda}, so I probably have that wrong.

<<
>I think that lo'i nu broda is a set of abstracts always and that we can use
>reference to these abstracts to refer to occassions tht full under them.

But why not do the same for dogs and chocolates then? Why are
they priviledged (or underpriviledged rather) with respect to
events? Isn't one of the claims of Lojban that they are treated
the same?
>>
I am not sure I follow this at all.  Dogs and chocolates are different from events in that they are concrete and events are abbstract: or rather {gerku} and {cakla} doe not claim on their faces to be abstract (and so, by default, are assumed to be concrete, since good Gricean manners would require noting such a significant feature) while {nu broda}, etc. do have this on their faces.  I am unclear here what the privilege involved is.  The whole point of my argument is just that like things be treated alike: classes are classes of things satisfying the predicate defining the class, so members of classes of concreta are concrete, those of abstracta are abstract.
I suspect a deep (too deep for me to dig out so far) confusion here: maybe between concrete and abstract v. token and type, maybe between classes and other abstract entities, maybe something else entirely .  And I am not sure which of us is confused  -- or whether both of us in antagonistic ways. 

<<
>(how do you build types out of tokens unless
>you somehow have a criterion for their being "the same." i.e. have a type
>in
>utero at least.)

The type (the intension) is in the meaning of the word, of
course. The question is how you access it. Not with lo/le,
I say. lo/le give you the extension, what you quantify over,
the members of the class one by one.
>>

Or maybe between type and intension and meaning (sense)?  I don't get this at all.
Quantification does not enter ere at all in any interesting way.  Remember, in Lojban, all entities are on the same level in the grammar, so it is as easy to quantify over the members of a set of types or intensions (not necessarily the same thing, by the way) as of a set of dogs.  One hopes that the type of a thing is not in the meaning of the word for that thing, because that will lead to an infinite regress, circling forever around to get out the type, given only tokens.  Unless you wan to fall back on the bootstrapping empirical procedure of science and leave a forever incompletely specified type -- which makes talk of types pretty senseless against other locutions. --part1_182.e1fa1b9.2aaf7269_boundary--