From pycyn@aol.com Fri Jul 27 06:20:35 2001
Return-Path: <Pycyn@aol.com>
X-Sender: Pycyn@aol.com
X-Apparently-To: lojban@yahoogroups.com
Received: (EGP: mail-7_2_0); 27 Jul 2001 13:20:34 -0000
Received: (qmail 67616 invoked from network); 27 Jul 2001 13:20:28 -0000
Received: from unknown (10.1.10.26) by l8.egroups.com with QMQP; 27 Jul 2001 13:20:28 -0000
Received: from unknown (HELO imo-d02.mx.aol.com) (205.188.157.34) by mta1 with SMTP; 27 Jul 2001 13:20:27 -0000
Received: from Pycyn@aol.com by imo-d02.mx.aol.com (mail_out_v31.9.) id r.10.101fa3f2 (4554) for <lojban@yahoogroups.com>; Fri, 27 Jul 2001 09:20:26 -0400 (EDT)
Message-ID: <10.101fa3f2.2892c499@aol.com>
Date: Fri, 27 Jul 2001 09:20:25 EDT
Subject: Re: [lojban] Tidying notes on {goi}
To: lojban@yahoogroups.com
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="part1_10.101fa3f2.2892c499_boundary"
X-Mailer: AOL 6.0 for Windows US sub 10531
From: pycyn@aol.com

--part1_10.101fa3f2.2892c499_boundary
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: quoted-printable

In a message dated 7/26/2001 6:02:41 PM Central Daylight Time,=20
jjllambias@hotmail.com writes:


> >among those selected by the first quantifier,
>=20
> What? Quantifiers don't select anything. {su'o da poi prenu cu prami}
> is a statement about the set of all persons.
>=20

Well, yes and no -- and I suspect that it is this that is at the heart of a=
=20
part of this problem. On the "yes" side, your sentence strictly staes tha=
t=20
the set of persons intersects the set of lovers (every quantifier is a=20
second-order relation between classes). On the "no" side, the members of=20
that intersection can now function linguistically (even if not logically--=
=20
but I would say logically, too) as designated individuals -- the standard=20
English "a boy"~"the boy" alternation (logic "some x" "eta x" or "alpha x",=
=20
not "iota x").=20=20=20=20

<I don't think we can have one rule for {ro} and a different rule
for {su'o}, as it would cause all sorts of inconsistencies>

Same rule, different results. the class selected by {ro} is the whole clas=
s,=20
so you can use that class again to define the class on which the second=20
quantifier works. with other initial quantifiers, the class for the second=
=20
quantifier is already restricted.

<Consider this for example:

=A0 =A0 su'o da poi prenu ku'o naku zo'u da prami su'o da

which is logically equivalent to:

=A0 =A0 naku ro da poi prenu zo'u da prami su'o da>

Hmmm! Negation presents a problem here and I need to work out what happens=
.=20=20
I suspect that, as usual, prenex forms decide what the "previous quantifier=
"=20
actually is.

<This is also more or less what happens in natlangs in any case:
"No student took that class. They hate the teacher." "They"
obviously refers to all the students, not to "no student".
In Lojban that might go something like {no da poi tadni cu cilre
fo ko'a i ro da xebni le ctuca}.>

Now, this is an interesting case! We have, of course, several responses:=20
"Natural language is so illogical," "In the logically pure form this is jus=
t=20
a case of the sort we have already described" and probably others. In any=
=20
case, this need not count against the present rule, but might lead to=20
rethinking it.

<{su'o lo prenu} may refer to a different prenu every time it
is used. I don't understand how you could have a double binding
in this case.>

But {lo prenu} is not a variable (there is something odd about that sentenc=
e=20
in context).=20

<>I am unsure what that would mean for the {goi} case; probably gobbledygoo=
k
>unless la alphas was the same entity as la betas.

In {da goi la alfas} la alfas cannot have a previous referent.
If it does, then it is gobbledygook.>

Under which set of rules? Why can this not (under the present rules) not=20
just be the namely rider on {da}, "there is an x, namely Alpha?"

<>Yes {da'o} clears the xy assignment and the subsequent {da} is a new
>quantifier, not now restricted to xy.

That's what I thought. You will have to correct you demonstration
then, as you leave xy dangling unassigned in the middle of it:>

Ummm! I thought that was your example; it isn't mine (who else was in this=
=20
discussion?)

<What happens if The Book is in contradiction with Logic? Which one
wins?>

As Lojbab says, during the freeeze, the book does. But I am not yet=20
convinced that this is such a case. It does raise an issue, much discussed=
=20
in the 70's, about the difference between identifying and relational uses o=
f=20
quantifiers. That proved almost insoluble in formal logic, but can in fact=
=20
be solved easily in langauges meant for use. I am not sure that lojban has=
=20
done this very well and that may be the heart of issue here. Lojban does=20
certainly have a number of work-arounds that cover the problem, but does no=
t=20
face it square on.





--part1_10.101fa3f2.2892c499_boundary
Content-Type: text/html; charset="ISO-8859-1"
Content-Transfer-Encoding: quoted-printable

<HTML><FONT FACE=3Darial,helvetica><BODY BGCOLOR=3D"#ffffff"><FONT SIZE=3D=
2>In a message dated 7/26/2001 6:02:41 PM Central Daylight Time,=20
<BR>jjllambias@hotmail.com writes:
<BR>
<BR>
<BR><BLOCKQUOTE TYPE=3DCITE style=3D"BORDER-LEFT: #0000ff 2px solid; MARGIN=
-LEFT: 5px; MARGIN-RIGHT: 0px; PADDING-LEFT: 5px">&gt;among those selected =
by the first quantifier,
<BR>
<BR>What? Quantifiers don't select anything. {su'o da poi prenu cu prami}
<BR>is a statement about the set of all persons.
<BR></BLOCKQUOTE>
<BR>
<BR>Well, yes and no -- and I suspect that it is this that is at the heart =
of a=20
<BR>part of this problem. &nbsp;On the "yes" side, &nbsp;your sentence stri=
ctly staes that=20
<BR>the set of persons intersects the set of lovers (every quantifier is a=
=20
<BR>second-order relation between classes). &nbsp;On the "no" side, the mem=
bers of=20
<BR>that intersection can now function linguistically (even if not logicall=
y--=20
<BR>but I would say logically, too) as designated individuals -- the standa=
rd=20
<BR>English "a boy"~"the boy" alternation (logic "some x" "eta x" or "alpha=
x",=20
<BR>not "iota x"). &nbsp;&nbsp;&nbsp;
<BR>
<BR>&lt;I don't think we can have one rule for {ro} and a different rule
<BR>for {su'o}, as it would cause all sorts of inconsistencies&gt;
<BR>
<BR>Same rule, different results. &nbsp;the class selected by {ro} is the w=
hole class,=20
<BR>so you can use that class again to define the class on which the second=
=20
<BR>quantifier works. &nbsp;with other initial quantifiers, the class for t=
he second=20
<BR>quantifier is already restricted.
<BR>
<BR>&lt;Consider this for example:
<BR>
<BR>=A0 =A0 su'o da poi prenu ku'o naku zo'u da prami su'o da
<BR>
<BR>which is logically equivalent to:
<BR>
<BR>=A0 =A0 naku ro da poi prenu zo'u da prami su'o da&gt;
<BR>
<BR>Hmmm! &nbsp;Negation presents a problem here and I need to work out wha=
t happens. &nbsp;
<BR>I suspect that, as usual, prenex forms decide what the "previous quanti=
fier"=20
<BR>actually is.
<BR>
<BR>&lt;This is also more or less what happens in natlangs in any case:
<BR>"No student took that class. They hate the teacher." "They"
<BR>obviously refers to all the students, not to "no student".
<BR>In Lojban that might go something like {no da poi tadni cu cilre
<BR>fo ko'a i ro da xebni le ctuca}.&gt;
<BR>
<BR>Now, this is an interesting case! &nbsp;We have, of course, several res=
ponses:=20
<BR>"Natural language is so illogical," "In the logically pure form this is=
just=20
<BR>a case of the sort we have already described" and probably others. &nbs=
p;In any=20
<BR>case, this need not count against the present rule, but might lead to=20
<BR>rethinking it.
<BR>
<BR>&lt;{su'o lo prenu} may refer to a different prenu every time it
<BR>is used. I don't understand how you could have a double binding
<BR>in this case.&gt;
<BR>
<BR>But {lo prenu} is not a variable (there is something odd about that sen=
tence=20
<BR>in context).=20
<BR>
<BR>&lt;&gt;I am unsure what that would mean for the {goi} case; probably g=
obbledygook
<BR>&gt;unless la alphas was the same entity as la betas.
<BR>
<BR>In {da goi la alfas} la alfas cannot have a previous referent.
<BR>If it does, then it is gobbledygook.&gt;
<BR>
<BR>Under which set of rules? &nbsp;Why can this not (under the present rul=
es) not=20
<BR>just be the namely rider on {da}, "there is an x, namely Alpha?"
<BR>
<BR>&lt;&gt;Yes {da'o} clears the xy assignment and the subsequent {da} is =
a new
<BR>&gt;quantifier, not now restricted to xy.
<BR>
<BR>That's what I thought. You will have to correct you demonstration
<BR>then, as you leave xy dangling unassigned in the middle of it:&gt;
<BR>
<BR>Ummm! &nbsp;I thought that was your example; it isn't mine (who else wa=
s in this=20
<BR>discussion?)
<BR>
<BR>&lt;What happens if The Book is in contradiction with Logic? Which one
<BR>wins?&gt;
<BR>
<BR>As Lojbab says, during the freeeze, the book does. &nbsp;But I am not y=
et=20
<BR>convinced that this is such a case. &nbsp;It does raise an issue, much =
discussed=20
<BR>in the 70's, about the difference between identifying and relational us=
es of=20
<BR>quantifiers. &nbsp;That proved almost insoluble in formal logic, but ca=
n in fact=20
<BR>be solved easily in langauges meant for use. &nbsp;I am not sure that l=
ojban has=20
<BR>done this very well and that may be the heart of issue here. &nbsp;Lojb=
an does=20
<BR>certainly have a number of work-arounds that cover the problem, but doe=
s not=20
<BR>face it square on.
<BR>
<BR>
<BR>
<BR></FONT></HTML>

--part1_10.101fa3f2.2892c499_boundary--