From pycyn@aol.com Fri Apr 20 10:22:41 2001 Return-Path: X-Sender: Pycyn@aol.com X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-7_1_2); 20 Apr 2001 17:22:41 -0000 Received: (qmail 24033 invoked from network); 20 Apr 2001 17:22:40 -0000 Received: from unknown (10.1.10.26) by l8.egroups.com with QMQP; 20 Apr 2001 17:22:40 -0000 Received: from unknown (HELO imo-r16.mx.aol.com) (152.163.225.70) by mta1 with SMTP; 20 Apr 2001 17:22:40 -0000 Received: from Pycyn@aol.com by imo-r16.mx.aol.com (mail_out_v29.14.) id r.d0.149fb21b (9725) for ; Fri, 20 Apr 2001 13:21:59 -0400 (EDT) Message-ID: Date: Fri, 20 Apr 2001 13:21:58 EDT Subject: Re: [lojban] RE: "not only" To: lojban@yahoogroups.com MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="part1_d0.149fb21b.2811ca36_boundary" Content-Disposition: Inline X-Mailer: AOL 6.0 for Windows US sub 10519 From: pycyn@aol.com X-Yahoo-Message-Num: 6727 --part1_d0.149fb21b.2811ca36_boundary Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable In a message dated 4/19/2001 6:33:14 PM Central Daylight Time,=20 jjllambias@hotmail.com writes: > >I take it that everyone agrees that for the general case "Only S is P"=20 > >means > >"All P is S" and thus does not entail "Some S is P" or even "Something i= s=20 > >P". >=20 > Now I'm not sure I agree with that. In any case, the entailment, > if there is one, would never be "Some S is P" but "All S is P". >=20 Oh, Lord, why me in retirement?! It would NEVER be "All S are P" for then,= =20 from the true "Only women are pregnant" we could infer the (hopefully) fals= e=20 "All women are pregnant." "Some" is all you have been claiming in the gener= al=20 case up to now and is the most you have even an implicature for. This is not the general case -- you separated it off yourself. I happen to= =20 think, on the usual logical grounds, that it works just like the general ca= se=20 and that what you seem to regard as a logical horror is merely a Gricean=20 dirty trick. Actually, a Gricean dirty trick is more likely to be called=20 nonsense than is a flat out contradiction, which would be properly (Gricean= =20 sense) called a contradiction. I suspect that you are coming around to mi\= y=20 point of view, kicking and screaming the whole way. <>=A0 If there is some doubt about this, consider the following.=A0 For hum= ans it=20 >is >universally true that only females are pregnant. "It is universally true that only females can be pregnant" would be the normal way to say it.> But that says something very different, doesn't it? I said there is nothin= g=20 that *is* both preganant and non-female. You said there is nothing that=20 *could be* pregnant and is not female. Even assuming {ka'e} here for "coul= d=20 be," this is a much more risky proposition. and even if not risky, it is=20 clearly different. I picked the example with some care -- to change it and= =20 argue against your version is strawman, a fallacy enough to cancel any valu= e=20 to the rest of your argument.=20=20 This pattern is only superficially different, it has the same deep structur= e=20 on an competent grammar. So the answer to your question is "No, but they=20 implicate that," just as before. < We could always make a lujvo in Lojban with the following place structure: =A0=A0 x1 is/are the only one(s) with property x2 among x3 where (1) x1 is a member (or subgroup) of x3, (2) x1 (or every member of x1) must have property x2, and (3) no member of x3 which is x1 (or a member of x1) can have property x3.> Well, of course, in clause 3 you mean members of x3 NOT in x1 It does, with the proviso that x1 is a set, not your favorite type of thing= .=20=20 We could probably loosen it to a mass or even just to a list of members.. = So=20 now you have a way to say what you mean by "only" It will occasionally be= =20 false when ordinary "only" is true and it may occasionally leave you with=20 rather surprising empty sets, but it will work for what you want. --part1_d0.149fb21b.2811ca36_boundary Content-Type: text/html; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable In a message dated 4/19/2001 6:33:14 PM Central Daylight Time,=20
jjllambias@hotmail.com writes:



>I take it that everyo= ne agrees that for the general case "Only S is P"=20
>means
>"All P is S" and thus does not entail "Some S is P" or even "Someth= ing is=20
>P".

Now I'm not sure I agree with that. In any case, the entailment,
if there is one, would never be "Some S is P" but "All S is P".




Oh, Lord, why me in retirement?!  It would NEVER be "All S are P" = for then,=20
from the true "Only women are pregnant" we could infer the (hopefully) = false=20
"All women are pregnant." "Some" is all you have been claiming in the g= eneral=20
case up to now and is the most you have even an implicature for.

<For example: "Only the cat and the dog like that chair" would
entail that both the cat and the dog like the chair, not that
at least one of them likes the chair.

In your view, the following makes sense: "Only the cat and the dog
like that chair, but only the cat likes it, and the cat doesn't
like it."
I don't believe that is merely a violation of implicature,
to me it is strictly nonsense.>
This is not the general case -- you separated it off yourself.  I = happen to=20
think, on the usual logical grounds, that it works just like the genera= l case=20
and that what you seem to regard as a logical horror is merely a Gricea= n=20
dirty trick.  Actually, a Gricean dirty trick is more likely to be= called=20
nonsense than is a flat out contradiction, which would be properly (Gri= cean=20
sense) called a contradiction.  I suspect that you are coming arou= nd to mi\y=20
point of view, kicking and screaming the whole way.

<>=A0 If there is some doubt about this, consider the following.= =A0 For humans it=20
>is
>universally true that only females are pregnant.

"It is universally true that only females can be pregnant" would
be the normal way to say it.>

But that says something very different, doesn't it?  I said there = is nothing=20
that *is* both preganant and non-female.  You said there is nothin= g that=20
*could be* pregnant and is not female.  Even assuming {ka'e} here = for "could=20
be," this is a much more risky proposition.  and even if not risky= , it is=20
clearly different. I picked the example with some care -- to change it = and=20
argue against your version is strawman, a fallacy enough to cancel any = value=20
to the rest of your argument.  

<Let's consider a slightly different use of "only":

My wife is the only one who likes olives.
The cat is the only one that likes that chair.
Females are the only ones that can be pregnant.

Do these entail that my wife is one who likes olives, that the
cat is one that likes that chair and that females are ones that
can be pregnant?>

This pattern is only superficially different, it has the same deep stru= cture=20
on an competent grammar.  So the answer to your question is "No, b= ut they=20
implicate that," just as before.

<
We could always make a lujvo in Lojban with the following
place structure:

=A0=A0 x1 is/are the only one(s) with property x2 among x3

where
(1) x1 is a member (or subgroup) of x3,
(2) x1 (or every member of x1) must have property x2, and
(3) no member of x3 which is x1 (or a member of x1) can have
property x3.>

Well, of course,  in clause 3 you mean members of x3 NOT in x1

<I wonder whether {selte'i} already means that...>
It does, with the proviso that x1 is a set, not your favorite type of t= hing.  
We could probably loosen it to a mass or even just to a list of members= ..  So=20
now you have a way to say what you mean by "only"  It will occasio= nally be=20
false when ordinary "only" is true and it may occasionally leave you wi= th=20
rather surprising empty sets, but it will work for what you want.
--part1_d0.149fb21b.2811ca36_boundary--