From pycyn@aol.com Wed Apr 18 20:00:38 2001 Return-Path: X-Sender: Pycyn@aol.com X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-7_1_2); 19 Apr 2001 03:00:38 -0000 Received: (qmail 86405 invoked from network); 19 Apr 2001 03:00:37 -0000 Received: from unknown (10.1.10.142) by l10.egroups.com with QMQP; 19 Apr 2001 03:00:37 -0000 Received: from unknown (HELO imo-m02.mx.aol.com) (64.12.136.5) by mta3 with SMTP; 19 Apr 2001 03:00:37 -0000 Received: from Pycyn@aol.com by imo-m02.mx.aol.com (mail_out_v29.14.) id r.15.12ea6034 (3876) for ; Wed, 18 Apr 2001 23:00:20 -0400 (EDT) Message-ID: <15.12ea6034.280faec4@aol.com> Date: Wed, 18 Apr 2001 23:00:20 EDT Subject: Re: [lojban] RE:not only To: lojban@yahoogroups.com MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="part1_15.12ea6034.280faec4_boundary" Content-Disposition: Inline X-Mailer: AOL 6.0 for Windows US sub 10519 From: pycyn@aol.com X-Yahoo-Message-Num: 6679 --part1_15.12ea6034.280faec4_boundary Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable In a message dated 4/18/2001 8:05:17 PM Central Daylight Time,=20 jjllambias@hotmail.com writes: > la pycyn cusku di'e >=20 > > > pa da nenri le tanxe i le bolci ji'a cu go'i > > > "Exactly one thing is in the box, the ball too is in the box." > > > >{ji'a} is the wrong discursive to use at this > >point; {sa'e} would be better or some other "namely rider." The "in=20 > >addition" > >is the remark, not the thing mentioned in the remark: " the ball too" is > >sloppy translating for "moreover the ball." >=20 > The same thing applies to {po'o} then. It doesn't change truth values > if {ji'a} doesn't. >=20 I think it is clear that I don't understand what {po'o} means, other than=20 that it has nothing to do with English "only". "The ball is the only thing in the box but the screw is in there too" is=20 contradictory because the screw is not the ball and the first sentence says= =20 that everything in the box is identical with the ball. "There is only one= =20 thing in the box and moreover the screw is in there too" is odd but not=20 contradictory: it tells me that the thing in the box is the screw, so it ad= ds=20 information, but information that would normally be put with a namely rider= ,=20 like "that is, the screw is in there." It is contradictory if the second=20 part means that the screw is in there in addition to the exactly one thing= =20 that is in there, but the {ji'a} does not say that, since it is a discursiv= e:=20 you need a non-agentive addition of the screw to the other thing (te sumji?= ).=20 Merely flagging an additional *remark*, doesn't mark an additonal thing. <>"only my wife likes olives and even she can't stand them" That works, but it entails that someone can like something that they can't stand. On the other hand, if you say "only my wife likes olives and even she doesn't like them", there is a flat contradiction for me, and the phrase loses all its humour.> Actually, I think it is funnier, largely because it cuts across the persons= =20 *expectations*. There is a perfectly good implicature there, but not an=20 entailment, so the denial is a shock but not a contradiction. Probably not= =20 even a shock around me, since I use it a lot. Then {po'o} affects truth value and should not be a discursive -- and shoul= d=20 probably be replaced by the appropriate form noted earlier, under general=20 lojbanic principles. Note the difference even you make between {po'o} requiring that my wife lik= e=20 olives (as you say) and {ji'a} being inappropriate when there are no other= =20 olive likers. That would be the difference between entailment and=20 implicature in a nutshell -- if it held. But your {po'o}, if it were reall= y=20 related to "only," would not be merely inappropriate if there were another= =20 relevant olive liker, it would be false. The {ji'a} expression is only fal= se=20 if your wife doesn't like olives, but it may be starkly odd if no one has=20 mentioned liking olives in the conversation up too the utterance. That wasn't my aim. I only want to show that if they both function like=20 {ji'a} then {po'o} has nothing to do with ^only"^(concept, not English word= )=20 and that if {po'o} does function like ^only^ then it functions very=20 differently from {ji'a} (these amount to the same thing, of course). I thi= nk=20 the upshot is that {ji'a} is a perfectly good discursive and {po'o} is a=20 bloody mess. <>Since Lojban has consistently refused to give existential import to >universally quantified terms (and it has repeatedly over nearly 50 years), I am extremely glad to hear you say this. In our August '95 discussion you were holding the opposite view, that {ro} had existential import!> Gotta keep are arguments straight here (and I don't always manage). My not= es=20 show two things: that all versions of "All S is P" : ro S cu P, ro da poi S= =20 cu P, roda ganai da S gi da P, rolo S cu P, and probably some I've forgotte= n=20 do none of them entail the corresponding existential: su'o S cu P, dapoi S = cu=20 P, da ge da S gi da P, su'o lo S cu P and thus all are the universals are=20 true if there are no Ss (this is usually the crucial point). And I have=20 argued for a quarter century (even winning for one short spell) that at lea= st=20 one of these forms (I like rodapoi) should have existential import, to matc= h=20 up with the implicature, which is there for all the cases, though least for= =20 roda ganai.=20=20 The problem is that, in standard logic, the raw universal does have=20 existential import, even though the restricted one does not. So, ro da gan= ai=20 da S gi da P does imply=20 su'o da ganai da S gi da P, whatever that means and all of the other forms= =20 have, hidden away in them somewhere an equally murky form. And so, when on= e=20 uses a universal quantifier not in an A sentence, one does in fact commit t= o=20 the partiuclar form as well. In particular in the internal quantifiers in= =20 descriptions (the case in point again) roloro S commits one to there being= =20 S's, by the internal quantifier, not the outer one. What I meant when I sa= id=20 a while ago that we had the wrong quantifiers on these guys.=20 <><, but "only the cat likes that chair" does imply >that the cat likes that chair. It is not just a case of >"only Ss are Ps".> > >What is it a case of, then?=A0 Surely the fact that the subject is singula= r >does not alter the logic so completely -- especially if it can take the=20 >same >quantifier expression as the general case. It is not being singular that makes it different, "only my two cats like that chair" would work just the same. Unlike "all", "the" does have existential import. At least in Lojban this is very clear: {le broda} is {ro le su'o broda}.> So, we agree that we need existential import to make the inference that the= =20 subject of ^only^ actually has the property and we agree that the universal= =20 taken alone does not have that import. Hence it does not entail that some = S=20 is P. Why the argument then? Ah yes, you don't believe the original point= =20 that "only S is P" is "All P is S" and the fact that you have cats helps no= t=20 at all, since it is not proven that there are chair likers -- though it is= =20 implicated.=20=20 --part1_15.12ea6034.280faec4_boundary Content-Type: text/html; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable In a message dated 4/18/2001 8:05:17 PM Central Daylight Time,=20
jjllambias@hotmail.com writes:


la pycyn cusku di'e

> > pa da nenri le tanxe i le bolci ji'a cu go'i
> > "Exactly one thing is in the box, the ball too is in the box.= "
>
>{ji'a} is the wrong discursive to use at this
>point; {sa'e} would be better or some other "namely rider." The "in= =20
>addition"
>is the remark, not the thing mentioned in the remark: " the ball to= o" is
>sloppy translating for "moreover the ball."

The same thing applies to {po'o} then. It doesn't change truth values
if {ji'a} doesn't.

I think it is clear that I don't understand what {po'o} means, other th= an=20
that it has nothing to do with English "only".
"The ball is the only thing in the box but the screw is in there too" i= s=20
contradictory because the screw is not the ball and the first sentence = says=20
that everything in the box is identical with the ball.  "There is = only one=20
thing in the box and moreover the screw is in there too" is odd but not= =20
contradictory: it tells me that the thing in the box is the screw, so i= t adds=20
information, but information that would normally be put with a namely r= ider,=20
like "that is, the screw is in there."  It is contradictory if the= second=20
part means that the screw is in there in addition to the exactly one th= ing=20
that is in there, but the {ji'a} does not say that, since it is a discu= rsive:=20
you need a non-agentive addition of the screw to the other thing (te su= mji?).=20
 Merely flagging an additional *remark*, doesn't mark an additonal= thing.

<>"only my wife likes olives and even she can't stand them"

That works, but it entails that someone can like something
that they can't stand. On the other hand, if you say "only my wife
likes olives and even she doesn't like them", there is a flat
contradiction for me, and the phrase loses all its humour.>
Actually, I think it is funnier, largely because it cuts across the per= sons=20
*expectations*.  There is a perfectly good implicature there, but = not an=20
entailment, so the denial is a shock but not a contradiction.  Pro= bably not=20
even a shock around me, since I use it a lot.

<It's extremely hard for me to believe that "only my wife likes
olives" does not entail "my wife likes olives", but it doesn't
really matter if this is how English works. In Lojban, if you
say {le mi speni ku po'o nelci lo'e rasygrute} then you are
commited to {le mi speni cu nelci lo'e rasygrute}, no matter
whether the English "equivalent" requires it or not. Also,
{po'o} is appropriate only if there is no other relevant olive
liker. Just like in {le mi speni ku ji'a nelci lo'e rasygrute},
{ji'a} would be appropriate only if there is some other relevant
olive liker.>
Then {po'o} affects truth value and should not be a discursive -- and s= hould=20
probably be replaced by the appropriate form noted earlier, under gener= al=20
lojbanic principles.
Note the difference even you make between {po'o} requiring that my wife= like=20
olives (as you say) and {ji'a} being inappropriate when there are no ot= her=20
olive likers.  That would be the difference between entailment and= =20
implicature in a nutshell -- if it held.  But your {po'o}, if it w= ere really=20
related to "only," would not be merely inappropriate if there were anot= her=20
relevant olive liker, it would be false.  The {ji'a} expression is= only false=20
if your wife doesn't like olives, but it may be starkly odd if no one h= as=20
mentioned liking olives in the conversation up too the utterance.

<I don't think you have shown that {po'o} is in any way less
logical than {ji'a}.>
That wasn't my aim.  I only want to show that if they both functio= n like=20
{ji'a} then {po'o} has nothing to do with ^only"^(concept, not English = word)=20
and that if {po'o} does function like ^only^ then it functions very=20
differently from {ji'a} (these amount to the same thing, of course). &n= bsp;I think=20
the upshot is that {ji'a} is a perfectly good discursive and {po'o} is = a=20
bloody mess.

<>Since Lojban has consistently refused to give existential impor= t to
>universally quantified terms (and it has repeatedly over nearly 50 = years),

I am extremely glad to hear you say this. In our August '95
discussion you were holding the opposite view, that {ro} had
existential import!>
Gotta keep are arguments straight here (and I don't always manage). &nb= sp;My notes=20
show two things: that all versions of "All S is P" : ro S cu P, ro da p= oi S=20
cu P, roda ganai da S gi da P, rolo S cu P, and probably some I've forg= otten=20
do none of them entail the corresponding existential: su'o S cu P, dapo= i S cu=20
P, da ge da S gi da P, su'o lo S cu P and thus all are the universals a= re=20
true if there are no Ss (this is usually the crucial point).  And = I have=20
argued for a quarter century (even winning for one short spell) that at= least=20
one of these forms (I like rodapoi) should have existential import, to = match=20
up with the implicature, which is there for all the cases, though least= for=20
roda ganai.  
The problem is that, in standard logic, the raw universal does have=20
existential import, even though the restricted one does not.  So, = ro da ganai=20
da S gi da P does imply=20
su'o da ganai da S gi da P, whatever that means and all of the other fo= rms=20
have, hidden away in them somewhere an equally murky form.  And so= , when one=20
uses a universal quantifier not in an A sentence, one does in fact comm= it to=20
the partiuclar form as well.  In particular in the internal quanti= fiers in=20
descriptions (the case in point again) roloro S commits one to there be= ing=20
S's, by the internal quantifier, not the outer one.  What I meant = when I said=20
a while ago that we had the wrong quantifiers on these guys.=20

<><, but "only the cat likes that chair" does imply
>that the cat likes that chair. It is not just a case of
>"only Ss are Ps".>
>
>What is it a case of, then?=A0 Surely the fact that the subject is = singular
>does not alter the logic so completely -- especially if it can take= the=20
>same
>quantifier expression as the general case.

It is not being singular that makes it different, "only my
two cats like that chair" would work just the same. Unlike
"all", "the" does have existential import. At least in Lojban
this is very clear: {le broda} is {ro le su'o broda}.>

So, we agree that we need existential import to make the inference that= the=20
subject of ^only^ actually has the property and we agree that the unive= rsal=20
taken alone does not have that import.  Hence it does not entail t= hat some S=20
is P.  Why the argument then? Ah yes, you don't believe the origin= al point=20
that "only S is P" is "All P is S" and the fact that you have cats help= s not=20
at all, since it is not proven that there are chair likers -- though it= is=20
implicated.  
























--part1_15.12ea6034.280faec4_boundary--