From pycyn@aol.com Fri Jul 05 09:20:19 2002 Return-Path: X-Sender: Pycyn@aol.com X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-8_0_7_4); 5 Jul 2002 16:20:19 -0000 Received: (qmail 46530 invoked from network); 5 Jul 2002 16:20:19 -0000 Received: from unknown (66.218.66.216) by m8.grp.scd.yahoo.com with QMQP; 5 Jul 2002 16:20:19 -0000 Received: from unknown (HELO imo-d02.mx.aol.com) (205.188.157.34) by mta1.grp.scd.yahoo.com with SMTP; 5 Jul 2002 16:20:18 -0000 Received: from Pycyn@aol.com by imo-d02.mx.aol.com (mail_out_v32.21.) id r.125.13222192 (4230) for ; Fri, 5 Jul 2002 12:20:14 -0400 (EDT) Message-ID: <125.13222192.2a57213d@aol.com> Date: Fri, 5 Jul 2002 12:20:13 EDT Subject: Re: [lojban] pro-sumti question To: lojban@yahoogroups.com MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="part1_125.13222192.2a57213d_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 X-Yahoo-Message-Num: 14584 --part1_125.13222192.2a57213d_boundary Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Let me start by reminding everyone that I agree completely with xorxes on how Lojban *should* be on the issue of the implicit quantifiers on {lei}, but, as often noted, I am obligated to talk about Lojban as it *is*. One reason for doing so is the hope that I will come upon a case so bizarre that it will force the change that is needed. Meanwhile, I will continue to deal with xorxes (et al)'s remarks that are obviously correct but Lojbanically (ut nunc) wrong. In a message dated 7/4/2002 8:30:18 PM Central Daylight Time, jjllambias@hotmail.com writes: There is no reason why it can't be; but there is also no reason why it has to be, which is the point here. There is no reason why {pisu'o} can't be piro either, but it doesn't have to be. So it is probably safest to go with the minimum. <>That mass is tired if only the dog (or only >the cat) is, just as the mass chases the potman if only the dog does. That's your interpretation. The way I see it, the properties of tiredness and postman chasing do not add up that way. Just like the dog's weight is not the mass weight, the dog's tiredness is not that of the mass. So, how do they work? I see postman-chasing as directly analogous to piano-toting, which can be done by three guys together even if only two of them ever lay hands on the piano (or housepainting if you like that better). And I see being tired as being like being green, true of a mass if true of one element. (remember -- as I think you occasionally do not) that in Lojban, {lei} behaves in this respect exactly like {loi}. <>Masses >aren't as useless as sets, but they need to be treated carefully. It would be very hard to do away with masses, they are very useful. But the idea that the mass has all properties of the members (or that anything that applies to part of a mass applies automatically to the mass) is nonsense, and unfortunately very widespread in Lojban lore.> That masses have all the properties of any member may be (but I don't think really is) widespread in Lojban lore, but I certainly don't hold it (and have spoken against it within the last few days). It is, however, the default position when nothing else clearly has a role -- logical sum when neither numerical nor participatory applies. And it is hard to see how either numerical or participatory applies in the case of tiredness (a little clearer in the case of postman-chasing). Of course, there are almost certainly other kinds of exceptions and they should be counted in as soon as explained, but for now I don't see an explanation that works for tiredness aside from saying that it is sui generis -- which is not very convincing. I suspect this is your quite correct intuition about English and Spanish that you are trying to import into Lojban. <>Is {ko'a joi ko'e gunma ko'a ce ko'e ce ko'i} true or >not? We'd have to agree first on the place structure of {gunma} for me to answer something that is meaningful to you.> gunma = 1 is a mass/team/aggregate[I'm not at all sure about "aggregate"] of members of the set 2 (this derived partly from the gi'ste, partly from the definition of {lo/e/ai} ) True in general, but not the issue here, which is whether a mass of a subset is also a mass of the set as a whole. <>If true then your remark backs up my point about masses being only >partial. If false then, then {loi gerku cu gunma lo'i gerku} is also >false, >against a number of basic sematic principles. I don't see how you conclude any of this. Taking {gunma} as the relationship between a mass and a set with the same members, then {piro loi gerku cu gunma lo'i broda} should be true, and therefore {pisu'o loi gerku cu gunma lo'i broda} should be true as well. {piro loi broda cu brode} always entails {pisu'o loi broda cu brode}, as long as {piro loi broda} is not the empty mass!> Ignoratio elenchi -- and petitio principi. This is not the question at issue and it presupposes one answer in the discussion -- namely that the members of the set and of the mass are the same. Also, {piro loi broda} can't refer to an empty mass, since it always entails {pisu'o loi broda} -- but we do genuinely disagree about that (or, rather, you think it should be otherwise and I say it is this way -- and don't think it even should be otherwise). <>{gunma} means -- like most >predicates -- "is A mass" not "The complete mass" from some set. Assuming it does (even though that's not what the gi'uste says) {le gunma} is still used to refer to one (or more) of those masses from some set, not to parts of those masses from some set. Once you identify what {le gunma} is, it is that mass, not just any part of that mass.> Hey, I have my revisionist streaks, too -- especially in the case of clear inconsistencies, as here, where a submass would not be a mass at all. But the point again is whether this mass from less than all the members of a set is a mass from that set. If it is, {le remei} can be just the dog (or the mass that consists of just the dog, for which the inference from member to mass seems to go through unscathed). {remei} only identifies the size of the set, not of the mass. On the other hand, if it is not, then lei gerku na gunma le'i gerku, even when dogs are meant all around. <>{remei} >notice talks about the size of the set underlying, not about completeness >either. I'm not sure what you mean here. Hopefully {lo sovda 12mei} does not claim that the set of all eggs has only 12 members! {lo remei} is "a dozen", not some part of an underlying set of 12.> No one -- so far as I can tell -- thinks that {lo sovda paremei} means that there are only twelve eggs altogether in the world forever, etc. I am less sure that it means "a dozen eggs", since I don't generally take that as a mass (except in a recipe: "add a dozen eggs" and maybe a few other places). Generally, I think "a dozen eggs" is just {pare sovda}, since I intend then to be used one by one. Or maybe {le sovda se paremei}. I can't now work out what {le remei} actually means. How would it differ >from {lei remei}? le remei = each of the pairs lei remei = all the pairs together> I disagree with the latter, of course, and would say "some of the pairs together." Lojban again, not common sense. <>It is, however likely that {lu'o le gerku .e le mlatu cu tatpi} doesn't >mean >that all of them are tired. Otherwise an officer telling his superior "lei >nanmu cu tatpi" would not be telling the truth. For {lei nanmu cu tatpi} to be true, it is not necessary that {ro le nanmu cu tatpi} be true. However, {pa le nanmu cu tatpi} does not entail {lei nanmu cu tatpi}. If the officer tells his superior {lei nanmu cu tatpi} just because one of them is, he would not be telling the truth. The group as a whole should be tired.> I'm not sure {lu'o le gerku e le mlatu cu tatpi} gets what is wanted here: the scope of {lu'o} ought to run out at the end of {gerku}. Oddly, it does not, meaning that -- contrary to intuitions -- {le gerku e le mlatu} are no longer logically connected, because of the {lu'o}, but now amount to {joi} without the {lu'o}. OK, it is correct -- though {joi} is safer, I think. Admittedly, this way -- if we could keep the logic of {e} -- would make my case better, but be just too weird for words. You still us an explanation of "the group as a whole should be tired" that is different from both "one member of the group is tired" and "all the members of the group are tired." "Some plurality of the group is tired"? What? But notice that, absent that explanation, the default rule holds, and that is the same for {lei} as for {loi}. <>I think this is one place where the book is right in the wrong way. The >implicit quantifiers on lei don't work as they should. A mass, considered >as >a mass, is either tired or isn't (otherwise masses are as useful as sets). >We can't have a situation where {lei nanmu cu tatpi .ije naku lei nanmu cu >tatpi} is true. Right. At least not any more than {mi ge tatpi ginai tatpi}.> Well, the case to watch for is {lei nanmu cu tatpi ije lei nanmu cu naku tatpi}. The others work because they are logical contradictions, not because they say anything useful about masses (or anything else). Maybe, but they don't work very well at all. In *Lojban*. <>ciksi). It makes no sense to be able to say {loi cmacu cu crinu} because I >decided that dying my own hair green wasn't enough. But notice that {pisu'o} is the right quantifier for {loi}, just like {su'o} is for {lo}. {loi smacu cu crino} says that some mice are green, about the same as {lo smacu cu crino}, since there is not much difference here in being green together or individually.> Thanks for emphasizing this difference; it does get lost occasionally in these discussions. --part1_125.13222192.2a57213d_boundary Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: 7bit Let me start by reminding everyone that I agree completely with xorxes on how Lojban *should* be on the issue of the implicit quantifiers on {lei}, but, as often noted, I am obligated to talk about Lojban as it *is*.  One reason for doing so is the hope that I will come upon a case so bizarre that it will force the change that is needed.  Meanwhile, I will continue to deal with xorxes (et al)'s remarks that are obviously correct but Lojbanically (ut nunc) wrong.

In a message dated 7/4/2002 8:30:18 PM Central Daylight Time, jjllambias@hotmail.com writes:

<No. There is no reason why {le remei} can't be {piro lei re danlu}.
There is no reason why the implicit {pisu'o} of {lei} has to be
transferred to {le remei}.>

There is no reason why it can't be; but there is also no reason why it has to be, which is the point here.  There is no reason why {pisu'o} can't be piro either, but it doesn't have to be.  So it is probably safest to go with the minimum.

<>That mass is tired if only the dog (or only
>the cat) is, just as the mass chases the potman if only the dog does.

That's your interpretation. The way I see it, the properties of
tiredness and postman chasing do not add up that way. Just like the
dog's weight is not the mass weight, the dog's tiredness is not
that of the mass.

So, how do they work?  I see postman-chasing as directly analogous to piano-toting, which can be done by three guys together even if only two of them ever lay hands on the piano (or housepainting if you like that better).  And I see being tired as being like being green, true of a mass if true of one element.  (remember -- as I think you occasionally do not) that in Lojban, {lei} behaves in this respect exactly like {loi}.

<>Masses
>aren't as useless as sets, but they need to be treated carefully.

It would be very hard to do away with masses, they are very
useful. But the idea that the mass has all properties of the
members (or that anything that applies to part of a mass applies
automatically to the mass) is nonsense, and unfortunately very
widespread in Lojban lore.>

That masses have all the properties of any member may be (but I don't think really is) widespread in Lojban lore, but I certainly don't hold it (and have spoken against it within the last few days).  It is, however, the default position when nothing else clearly has a role -- logical sum when neither numerical nor participatory applies.  And it is hard to see how either numerical or participatory applies in the case of tiredness (a little clearer in the case of postman-chasing).  Of course, there are almost certainly other kinds of exceptions and they should be counted in as soon as explained, but for now I don't see an explanation that works for tiredness aside from saying that it is sui generis -- which is not very convincing.  I suspect this is your quite correct intuition about English and Spanish that you are trying to import into Lojban.

<>Is {ko'a joi ko'e gunma ko'a ce ko'e ce ko'i} true or
>not?

We'd have to agree first on the place structure of {gunma} for
me to answer something that is meaningful to you.>

gunma = 1 is a mass/team/aggregate[I'm not at all sure about "aggregate"] of members of the set 2  (this derived partly from the gi'ste, partly from the definition of {lo/e/ai} )

<Assuming {ko'a} and {ko'e} refer to individuals, then
{ko'a joi ko'e} refers to a mass of two individuals, and
{ko'a broda} being true does not imply that {ko'a joi ko'e broda}.>

True in general, but not the issue here, which is whether a mass of a subset is also a mass of the set as a whole.

<>If true then your remark backs up my point about masses being only
>partial.  If false then, then {loi gerku cu gunma lo'i gerku} is also
>false,
>against a number of basic sematic principles.

I don't see how you conclude any of this. Taking {gunma} as the
relationship between a mass and a set with the same members, then
{piro loi gerku cu gunma lo'i broda} should be true, and therefore
{pisu'o loi gerku cu gunma lo'i broda} should be true as well.
{piro loi broda cu brode} always entails {pisu'o loi broda cu brode},
as long as {piro loi broda} is not the empty mass!>

Ignoratio elenchi -- and petitio principi.  This is not the question at issue and it presupposes one answer in the discussion -- namely that the members of the set and of the mass are the same.  Also, {piro loi broda} can't refer to an empty mass, since it always entails {pisu'o loi broda}  -- but we do genuinely disagree about that (or, rather, you think it should be otherwise and I say it is this way -- and don't think it even should be otherwise).

<>{gunma} means -- like most
>predicates -- "is A mass" not "The complete mass" from some set.

Assuming it does (even though that's not what the gi'uste says)
{le gunma} is still used to refer to one (or more) of those masses
from some set, not to parts of those masses from some set. Once
you identify what {le gunma} is, it is that mass, not just any part
of that mass.>

Hey, I have my revisionist streaks, too -- especially in the case of clear inconsistencies, as here, where a submass would not be a mass at all.  But the point again is whether this mass from less than all the members of a set is a mass from that set.  If it is, {le remei} can be just the dog (or the mass that consists of just the dog, for which the inference from member to mass seems to go through unscathed).  {remei} only identifies the size of the set, not of the mass.  On the other hand, if it is not, then lei gerku na gunma le'i gerku, even when dogs are meant all around.

<>{remei}
>notice talks about the size of the set underlying, not about completeness
>either.

I'm not sure what you mean here. Hopefully {lo sovda 12mei} does
not claim that the set of all eggs has only 12 members! {lo remei}
is "a dozen", not some part of an underlying set of 12.>

No one -- so far as I can tell -- thinks that {lo sovda paremei} means that there are only twelve eggs altogether in the world forever, etc.  I am less sure that it means "a dozen eggs", since I don't generally take that as a mass (except in a recipe: "add a dozen eggs" and maybe a few other places).  Generally, I think "a dozen eggs" is just {pare sovda}, since I intend then to be used one by one.  Or maybe {le sovda se paremei}. 

<la greg cusku di'e

>I can't now work out what {le remei} actually means. How would it differ
>from {lei remei}?

le remei = each of the pairs
lei remei = all the pairs together>

I disagree with the latter, of course, and would say "some of the pairs together."  Lojban again, not common sense.

<>It is, however likely that {lu'o le gerku .e le mlatu cu tatpi} doesn't
>mean
>that all of them are tired. Otherwise an officer telling his superior "lei
>nanmu cu tatpi" would not be telling the truth.

For {lei nanmu cu tatpi} to be true, it is not necessary that
{ro le nanmu cu tatpi} be true. However, {pa le nanmu cu tatpi}
does not entail {lei nanmu cu tatpi}. If the officer tells
his superior {lei nanmu cu tatpi} just because one of them
is, he would not be telling the truth. The group as a whole
should be tired.>

I'm not sure {lu'o le gerku e le mlatu cu tatpi} gets what is wanted here: the scope of {lu'o} ought to run out at the end of {gerku}.  Oddly, it does not, meaning that -- contrary to intuitions -- {le gerku e le mlatu} are no longer logically connected, because of the {lu'o}, but now amount to {joi} without the {lu'o}.  OK, it is correct -- though {joi} is safer, I think.  Admittedly, this way -- if we could keep the logic of {e} -- would make my case better, but be just too weird for words.
You still us an explanation of  "the group as a whole should be tired" that is different from both "one member of the group is tired" and "all the members of the group are tired." "Some plurality of the group is tired"?  What?  But notice that, absent that explanation, the default rule holds, and that is the same for {lei} as for {loi}.

<>I think this is one place where the book is right in the wrong way. The
>implicit quantifiers on lei don't work as they should. A mass, considered
>as
>a mass, is either tired or isn't (otherwise masses are as useful as sets).
>We can't have a situation where {lei nanmu cu tatpi .ije naku lei nanmu cu
>tatpi} is true.

Right. At least not any more than {mi ge tatpi ginai tatpi}.>

Well, the case to watch for is {lei nanmu cu tatpi ije lei nanmu cu naku tatpi}.  The others work because they are logical contradictions, not because they say anything useful about masses (or anything else). 

<The prosumti we want here is necessarily
vague, and {le remei} (or {le romoi}, or whatever number works best
in context) I think is the best we have. I used this method a few
times in the Alice translation.>

Maybe, but they don't work very well at all. In *Lojban*.

<>ciksi). It makes no sense to be able to say {loi cmacu cu crinu} because I
>decided that dying my own hair green wasn't enough.

But notice that {pisu'o} is the right quantifier for {loi},
just like {su'o} is for {lo}. {loi smacu cu crino}
says that some mice are green, about the same as {lo smacu cu crino},
since there is not much difference here in being green together
or individually.>

Thanks for emphasizing this difference; it does get lost occasionally in these discussions.


























--part1_125.13222192.2a57213d_boundary--