From sentto-44114-15840-1032394468-lojban-in=lojban.org@returns.groups.yahoo.com Wed Sep 18 17:16:45 2002 Received: with ECARTIS (v1.0.0; list lojban-list); Wed, 18 Sep 2002 17:16:45 -0700 (PDT) Received: from n15.grp.scd.yahoo.com ([66.218.66.70]) by digitalkingdom.org with smtp (Exim 4.05) id 17rozr-0007lL-00 for lojban-in@lojban.org; Wed, 18 Sep 2002 17:16:43 -0700 X-eGroups-Return: sentto-44114-15840-1032394468-lojban-in=lojban.org@returns.groups.yahoo.com Received: from [66.218.66.95] by n15.grp.scd.yahoo.com with NNFMP; 19 Sep 2002 00:14:29 -0000 X-Sender: Pycyn@aol.com X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-8_1_1_3); 19 Sep 2002 00:14:28 -0000 Received: (qmail 64345 invoked from network); 19 Sep 2002 00:14:28 -0000 Received: from unknown (66.218.66.217) by m7.grp.scd.yahoo.com with QMQP; 19 Sep 2002 00:14:28 -0000 Received: from unknown (HELO imo-m08.mx.aol.com) (64.12.136.163) by mta2.grp.scd.yahoo.com with SMTP; 19 Sep 2002 00:14:28 -0000 Received: from Pycyn@aol.com by imo-m08.mx.aol.com (mail_out_v34.10.) id r.152.14477cfa (3980) for ; Wed, 18 Sep 2002 20:14:22 -0400 (EDT) Message-ID: <152.14477cfa.2aba70de@aol.com> To: lojban@yahoogroups.com X-Mailer: AOL 7.0 for Windows US sub 10509 From: pycyn@aol.com X-Yahoo-Profile: kaliputra MIME-Version: 1.0 Mailing-List: list lojban@yahoogroups.com; contact lojban-owner@yahoogroups.com Delivered-To: mailing list lojban@yahoogroups.com Precedence: bulk Date: Wed, 18 Sep 2002 20:14:22 EDT Subject: Re: [lojban] Re: I like chocolate and matters someone has related to it Content-Type: multipart/alternative; boundary="part1_152.14477cfa.2aba70de_boundary" X-archive-position: 1328 X-ecartis-version: Ecartis v1.0.0 Sender: lojban-list-bounce@lojban.org Errors-to: lojban-list-bounce@lojban.org X-original-sender: pycyn@aol.com Precedence: bulk Reply-to: lojban-list@lojban.org X-list: lojban-list --part1_152.14477cfa.2aba70de_boundary Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit In a message dated 9/18/2002 9:43:29 AM Central Daylight Time, arosta@uclan.ac.uk writes: << > where {tu'a da} is a cover for {tu'o du'u > #ce'u co'e da}, it's going to get the implication the other way as well. > > Is {tu'a da} a cover for {tu'o du'u da zo'u ce'u co'e da}? That is the > crux, and I think we all want the answer to be Yes. >> I don't see that it makes much difference, unless {co'e} contains something with other quantifiers than particular in it. If not, then (skipping the fringes) {co'e da} and {da zo'u co'e da} are equivalent. If {co'e} does have the buried quantfiers, then they are not, but {tu'a da} will be the unfronted form (I think: we can't lift the prenex alone since then we don have something of the form {co'e da}) Why is this important? << BTW, are you actually proposing locutions like {nelci tu'a lo cakla}, {nelci tu'o du'u ce'u co'e lo cakla}? To me, those don't mean the same thing as "I like chocolate". >> As the second, No -- did I mention such a form? It was probably meant to be {nu} rather than {du'u} (although the {du'u} form makes a sort of sense, too). But {nelci tu'a lo cakla} seems to me just right for "I like chocolate," in at least one reading of it, where the purposes for which I like it are hinted at but not given -- presumably eating for choclate, but maybe bathing or car waxing. jordan: << [ note to lionel: the default quantifier on da/de/di is su'o, which is where the ambiguity comes from: ] >> What ambiguity? << This doesn't support that a broda b != b se broda a in the general case. This merely shows that there is a different most-likely interpretation of the quantification of the da/de/di variables based on their order. Either of those two sentences *could* be interpreted as the other, but le gerku cu batci mi is precisely the same as mi se batci le gerku; both in possible meanings and in the most-likely interpretation. >> I'm as good a gricean as the next guy, but I don't see that there is any "most likely interpretation" here, this is a rock hard rule (and I'll put the {su'o}s in if it makes the case more clearly). The order of quantifier binding is the left to right order of the quantifier expressions (or their place, if they are implicit). The example using {le gerku cu batci mi} is not relevant, since it does not involve moving two different quantifiers ({mi} doesn't have a quantifier, so not a different one). << Furthermore, though the word order leads to different likely interpretation it doesn't change the possible meanings. ro da prami de Can mean "Everyone loves >=one other (the same) person" just as much as it can mean "Everyone loves someone (else)". >> In fact, it can't mean either of these, though it would be true it either of these was. But it would also be true in many cases where these are false: if each person loved him/her self, for example. << Your mearly cheating with su'o to try to claim the grammar doesn't fully explain this. >> Me cheat? Surely you jest. The grammar certainly does not give the above explanations, or -- if it does -- is more thoroughly verruckt then I remember. << The non-ambiguous ways to make the two claims are: ro da poi prenu cu prami lo drata be vo'a Everyone loves someone other than themselves. (in practice the be vo'a would likely be elided and inferred through a zo'e). >> This does work for the "Everybody loves someone else" reading -- which is not "Everybody loves somebody" in any sense and certainly not {ro da prami de}. << ro da poi prenu cu prami le su'o prenu Everyone loves the one-or-more persons. >> This is "Everybody loves the people I have in mind." A new line altogether, but still one that entails {ro da prami de} (assuming restriction to prenu). Different from it though. << These two claims *are* the exact same if you flip the terms. (Except the former requires changing the vo'a to a vo'e). >> Nope. The second is, because the quantifiers are the same (the implicit quantifier on {le} is {ro} (the {su'o} is a cardinal for the whole set, not a quantifier). Shifting the first one to {lo drata be vo'e cu se prami ro da poi prenu} means something like (it is hard to get the cross references right in English) "Someone different from all people is loved by every person" A dog, maybe, or God or Gaia. In any case, it is very different from the unswitched version. << [ what's a rorne? ] >> A horn, referred to by some conflation of the last three or four vocabulary shuffles in LoCCan. I meant {jirna}, but foolishly forgot to look it up. << I was discussing this point with some people on IRC a while back, and bunk I say! bunk! Of course unicorns exist: they're concepts. If I say {mi djica lenu lo pavyseljirna cu klama ti} there's nothing wrong with the bridi, as I really do desire that su'o lo ro pavyseljirna come (even if ro = 0; the su'o is just the number I'm wanting). >> Unicorns don't exist (on Earth anyhow). If they did, they would be animals etc. etc. not concepts nor pictures nor stories nor.... the fact that we have pictures and stories and concepts of unicorns is a large part of why we need intensional contexts: so we can go from saying we saw a picture of a unicorn to saying that there is something whose pictue we saw (This is official doctrine; I've been toying with some variants in which there are unicorns. But they still don't exist.) {ro} can't equal 0, since {ro broda cu brode} implies {su'o broda cu brode} -- and in this context, {ro} must be at least as large as {su'o}, which is greater than 0. Your desire is nicely in an opaque context, so that nothing talked about there needs to exist in the real world. << zo'o mi nelci le su'o su'o pavyseljirna cu zasti .i zo'o lo no pavyseljirna cu zasti >> 'Tain't funny (haha -- but maybe peculiar). I assume the first {su'o} is {su'u}. You can like the idea (more or less) that unicorns exist, without unicorns existing. Indeed, you can like the event of unicorns existing, without unicorns existing. All abstracts exist, even if what is in them doesn't. The second sentence is grammatical but semantically ill-formed (i.e., nonsense). You can't get one or more things out of an empty set, but {[su'o] lo no pavyseljirna} claims to refer to just those things -- one or more of them. << Additionally, certainly you can dream a unicorn klama do, as unicorns *do* exist in dreams. With: da poi pavyseljirna zo'u mi senva ledu'u da klama mi says "there is a unicorn such that I dreamt it came to me". Which (assuming the speaker isn't lying) is perfectly fine. That pavyseljirna exists as whatever it is that dreams/concepts are from a biological standpoint, etc. >> Which is why the content of dreams is carefully isolated behind some abstactor or other (not always clear what the best one is). And one of the three things you cannot do with absrtract clauses is bind terms in them with quantifiers outside them (the others are moving existing quantifiers from inside to outside and replacing one term by another on the basis of an identity external to the abstraction). [Again, there is a prefectly good way of talking where in we could do the binding bit, but that would only show that there are unicorns, not that they exist. And this is not the approach Lojban has taken.] << It should be noted also, that if I had actually had a dream, since I have the unicorn in mind already, the better sentence would be mi senva ledu'u le pavyseljirna cu klama mi >> Yeah, it is hard to know what kind of abstraction a dream is -- and, indeed, people's dreams seem to differ (I used to dream in text or voiceover), so this may be right sometimes. I favor {li'i}, but that may just be because I don't understand it. None of this helps unicorns to exist, though. << Can tu'a/jai be used to raise sumti out of relative clauses? It seems to me that we should be able to. For example: {tu'a le prenu} could raise from {le gerku poi pu batci le prenu} I'm unable to find anything in the book specifically prohibitive of this sort of thing; but naturally nothing suggesting it is ok. What do people think? >> I suppose it could, but I don't see the point from this example anyhow. Can you come up with one where I can imagine someone wanting to do it? {tu'a} certainly and, by implication (and logic) {jai}, mark their connected sumti as being in an intensional context and therefore not open to certain normal manipulations. Sumti in relative clauses are not under these restrictions, so why would we want to restrict them? &: << I still haven't had time to digest those ideas, but in the meantime I have remembered an old argument in favour of {lo'e} or {tu'o} in these exx. It seems to me that what is essentially going on in these exx -- and also generally with generic reference -- is that a category is being conceptualized as a single individual ("myopic singularization"). E.g. it is quite easy to think of Chocolate as a single individual, and "I like chocolate" means the same as "I like Chocolate". >> Yes, that notion was around often for {loi}, as one of several ways to deal with masses. I'm not sure it helps here. but that is mainly because ever attempt to cash that idea in for some definite already intelligible notions and operations, never got much beyond the stage of xorxes {lo'e}. We could try tet again, I suppose, but I don't really think it will come to much this time either. << So on this basis I understand your use of {lo'e} and agree with it. The question that remains in my mind is whether there is a difference between {lo'e broda} and {tu'o broda}. >> That suggests that you understand myopic singularization, which, since it was, I think, one of your early contributions, is to be expected. But I don't recall anyone else ever getting the knack of it, and certainly most now on this list don't know what you are talking about. I suggest a thorough exposition, if you think the notion will really help here (not that I am inclined to think that xorxes' idea is one we need to add to Lojban). << BTW, this automatically gives us a useful meaning for {le'e} -- it would mean {(ro) le pa}. >> I don't see how this is automatic nor even desirable or useful. And it certainly doesn't have any much connection with the cases at hand. (I infer from the o/e alternation and the standard meanings, that {le'e} is the subjective or selective version of {lo'e}. Does this mean that {lo'e} means {lo pa}?) xorxes: << >Is {tu'a da} a cover for {tu'o du'u da zo'u ce'u co'e da}? That is the >crux, and I think we all want the answer to be Yes. We all want that as far as where the quantifier goes, I'm sure about that. >> I don't think I do, as noted above. << But in general, {tu'a da} is a cover for any of: {tu'o du'u da zo'u ce'u co'e da} {tu'o du'u da zo'u co'e da} {tu'o? nu da zo'u co'e da} and possibly others. Which one it is depends on where it is used. It's a du'u in a place that accepts {du'u}s, a nu in a place that accepts {nu}s. >> And up in the air for those that accept more than one absraction. So, it is not as clear as the full form -- but a lot shorter, which may be enough if the situation is otherwise clear (usually even a good guess at {co'e}). << >BTW, are you actually proposing locutions like {nelci tu'a lo cakla}, >{nelci tu'o du'u ce'u co'e lo cakla}? To me, those don't mean the >same thing as "I like chocolate". I don't think he was proposing that. In this case, x2 of nelci accepts {nu}s, not {du'u}s, so it means {nelci ?? nu co'e lo cakla}. (I think ?? = {tu'o} or equivalently {lo'e}.) >> Damn, now we have {lo'e} being equivalent to (the so far totally unexplained ) {tu'o} -- and used with abstractors diectly to boot: the generic event of doing something with chocolate. Do note that these do not fit the examples so far (and I mean over the entire month+ of this discussion). cowan: on jorda n << > I say {mi djica lenu lo pavyseljirna cu klama ti} there's nothing wrong > with the bridi, as I really do desire that su'o lo ro pavyseljirna > come (even if ro = 0; the su'o is just the number I'm wanting). There *is* nothing wrong, because nu-events exist even if the things inside don't. But lo pavyseljirna cu blabi, "some unicorn is white", that's rubbish. >> No, just false, because there aren't any unicorns. --part1_152.14477cfa.2aba70de_boundary Content-Type: text/html; charset=US-ASCII Content-Transfer-Encoding: 7bit In a message dated 9/18/2002 9:43:29 AM Central Daylight Time, arosta@uclan.ac.uk writes:

<<
where {tu'a da}  is a cover for {tu'o du'u
#ce'u co'e da}, it's going to get the implication the other way as well.

Is {tu'a da} a cover for {tu'o du'u da zo'u ce'u co'e da}? That is the
crux, and I think we all want the answer to be Yes.

>>
I don't see that it makes much difference, unless {co'e} contains something with other quantifiers than particular in it.  If not, then (skipping the fringes) {co'e da} and {da zo'u co'e da} are equivalent.  If {co'e} does have the buried quantfiers, then  they are not, but {tu'a da} will be the unfronted form (I think: we can't lift the prenex alone since then we don have something of the form {co'e da}) Why is this important?

<<
BTW, are you actually proposing locutions like {nelci tu'a lo cakla},
{nelci tu'o du'u ce'u co'e lo cakla}? To me, those don't mean the
same thing as "I like chocolate".
>>
As the second, No -- did I mention such a form?  It was probably meant to be {nu} rather than {du'u} (although the {du'u} form makes a sort of sense, too). But {nelci tu'a lo cakla} seems to me just right for "I like chocolate," in at least one reading of it, where the purposes for which I like it are hinted at but not given -- presumably eating for choclate, but maybe bathing or car waxing.

jordan:
<<
[ note to lionel: the default quantifier on da/de/di is su'o, which is
where the ambiguity comes from: ]
>>
What ambiguity?

<<
This doesn't support that a broda b != b se broda a in the general
case.  This merely shows that there is a different most-likely
interpretation of the quantification of the da/de/di variables based
on their order.  Either of those two sentences *could* be interpreted
as the other, but le gerku cu batci mi is precisely the same as mi
se batci le gerku;  both in possible meanings and in the most-likely
interpretation.
>>
I'm as good a gricean as the next guy, but I don't see that there is any "most likely interpretation" here, this is a rock hard rule (and I'll put the {su'o}s in if it makes the case more clearly). The order of quantifier binding is the left to right order of the quantifier expressions (or their place, if they are implicit).  The example using {le gerku cu batci mi} is not relevant, since it does not involve moving two different quantifiers ({mi} doesn't have a quantifier, so not a different one). 

<<
Furthermore, though the word order leads to different likely interpretation
it doesn't change the possible meanings.
    ro da prami de
Can mean "Everyone loves >=one other (the same) person" just as much as it
can mean "Everyone loves someone (else)".
>>
In fact, it can't mean either of these, though it would be true it either of these was. But it would also be true in many cases where these are false: if each person loved him/her self, for example.

<<
Your mearly cheating with su'o
to try to claim the grammar doesn't fully explain this.
>>
Me cheat? Surely you jest.  The grammar certainly does not give the above explanations, or -- if it does -- is more thoroughly verruckt  then I remember.

<<
  The non-ambiguous
ways to make the two claims are:

    ro da poi prenu cu prami lo drata be vo'a
    Everyone loves someone other than themselves.
  (in practice the be vo'a would likely be elided and inferred
    through a zo'e).
>>
This does work for the "Everybody loves someone else" reading -- which is not "Everybody loves somebody" in any sense and certainly not {ro da prami de}.

<<
   ro da poi prenu cu prami le su'o prenu
    Everyone loves the one-or-more persons.
>>
This is "Everybody loves the people I have in mind." A new line altogether, but still one that entails {ro da prami de} (assuming restriction to prenu). Different from it though.

<<
These two claims *are* the exact same if you flip the terms.  (Except
the former requires changing the vo'a to a vo'e).
>>
Nope.  The second is, because the quantifiers are the same (the implicit quantifier on {le} is {ro} (the {su'o} is a cardinal for the whole set, not a quantifier).
Shifting the first one to {lo drata be vo'e cu se prami ro da poi prenu} means something like (it is hard to get the cross references right in English) "Someone different from all people is loved by every person"  A dog, maybe, or God or Gaia.  In any case, it is very different from the unswitched version.

<<
[ what's a rorne? ]
>>
A horn, referred to by some conflation of the last three or four vocabulary shuffles in LoCCan.  I meant {jirna}, but foolishly forgot to look it up.

<<
I was discussing this point with some people on IRC a while back, and
bunk I say!  bunk!  Of course unicorns exist:  they're concepts.  If
I say {mi djica lenu lo pavyseljirna cu klama ti} there's nothing wrong
with the bridi, as I really do desire that su'o lo ro pavyseljirna
come (even if ro = 0; the su'o is just the number I'm wanting).
>>
Unicorns don't exist (on Earth anyhow).  If they did, they would be animals etc. etc. not concepts nor pictures nor stories nor....  the fact that we have pictures and stories and concepts of unicorns is a large part of why we need intensional contexts: so we can go from saying we saw a picture of a unicorn to saying that there is something whose pictue we saw (This is official doctrine; I've been toying with some variants in which there are unicorns.  But they still don't exist.)
{ro} can't equal 0, since {ro broda cu brode} implies {su'o broda cu brode}  -- and in this context, {ro} must be at least as large as {su'o}, which is greater than 0.  Your desire is nicely in an opaque context, so that nothing talked about there needs to exist in the real world.

<<
  zo'o mi nelci le su'o su'o pavyseljirna cu zasti
  .i zo'o lo no pavyseljirna cu zasti
>>
'Tain't funny (haha -- but maybe peculiar).  I assume the first {su'o} is {su'u}.  You can like the idea (more or less) that unicorns exist, without unicorns existing.  Indeed, you can like the event of unicorns existing, without unicorns existing.  All abstracts exist, even if what is in them doesn't.
The second sentence is grammatical but semantically ill-formed (i.e., nonsense).  You can't get one or more things out of an empty set, but {[su'o] lo no pavyseljirna} claims to refer to just those things -- one or more of them.

<<
Additionally, certainly you can dream a unicorn klama do, as unicorns
*do* exist in dreams.  With:
    da poi pavyseljirna zo'u mi senva ledu'u da klama mi
says "there is a unicorn such that I dreamt it came to me".  Which
(assuming the speaker isn't lying) is perfectly fine.  That
pavyseljirna exists as whatever it is that dreams/concepts are from
a biological standpoint, etc.
>>
Which is why the content of dreams is carefully isolated behind some abstactor or other (not always clear what the best one is).  And one of the three things you cannot do with absrtract clauses  is bind terms in them with quantifiers outside them (the others are moving existing quantifiers from inside to outside and replacing one term by another on the basis of an identity external to the abstraction). [Again, there is a prefectly good way of talking where in we could do the binding bit, but that would only show that there are unicorns, not that they exist. And this is not the approach Lojban has taken.]

<<
It should be noted also, that if I had actually had a dream, since I
have the unicorn in mind already, the better sentence would be
    mi senva ledu'u le pavyseljirna cu klama mi
>>
Yeah, it is hard to know what kind of abstraction a dream is -- and, indeed, people's dreams seem to differ (I used to dream in text or voiceover), so this may be right sometimes.  I favor {li'i}, but that may just be because I don't understand it.  None of this helps unicorns to exist, though.

<<
Can tu'a/jai be used to raise sumti out of relative clauses?  It seems
to me that we should be able to.  For example:

  {tu'a le prenu} could raise from {le gerku poi pu batci le prenu}

I'm unable to find anything in the book specifically prohibitive of
this sort of thing; but naturally nothing suggesting it is ok.  What
do people think?
>>
I suppose it could, but I don't see the point from this example anyhow.  Can you come up with one where I can imagine someone wanting to do it? 
{tu'a} certainly and, by implication (and logic) {jai}, mark their connected sumti as being in an intensional context and therefore not open to certain normal manipulations.  Sumti in relative clauses are not under these restrictions, so why would we want to restrict them?

&:
<<
I still haven't had time to digest those ideas, but in the meantime
I have remembered an old argument in favour of {lo'e} or
{tu'o} in these exx. It seems to me that what is essentially
going on in these exx -- and also generally with generic
reference -- is that a category is being conceptualized as
a single individual ("myopic singularization"). E.g. it is
quite easy to think of Chocolate as a single individual,
and "I like chocolate" means the same as "I like Chocolate".
>>
Yes, that notion was around often for {loi}, as one of several ways to deal with masses.  I'm not sure it helps here.  but that is mainly because ever attempt to cash that idea in for some definite already intelligible notions and operations, never got much beyond the stage of xorxes {lo'e}.  We could try tet again, I suppose, but I don't really think it will come to much this time either.

<<
So on this basis I understand your use of {lo'e} and agree
with it. The question that remains in my mind is whether
there is a difference between {lo'e broda} and {tu'o broda}.
>>
That suggests that you understand myopic singularization, which, since it was, I think, one of your early contributions, is to be expected.  But I don't recall anyone else ever getting the knack of it, and certainly most now on this list don't know what you are talking about.  I suggest a thorough exposition, if you think the notion will really help here (not that I am inclined to think that xorxes' idea is one we need to add to Lojban).

<<
BTW, this automatically gives us a useful meaning for
{le'e} -- it would mean {(ro) le pa}.
>>
I don't see how this is automatic nor even desirable or useful.  And it certainly doesn't have any much connection with the cases at hand.  (I infer from the o/e alternation and the standard meanings, that {le'e} is the subjective or selective version of {lo'e}.  Does this mean that {lo'e} means {lo pa}?)

xorxes:
<<
>Is {tu'a da} a cover for {tu'o du'u da zo'u ce'u co'e da}? That is the
>crux, and I think we all want the answer to be Yes.

We all want that as far as where the quantifier goes, I'm sure
about that.
>>
I don't think I do, as noted above.

<<
But in general, {tu'a da} is a cover for any of:
{tu'o du'u da zo'u ce'u co'e da}
{tu'o du'u da zo'u co'e da}
{tu'o? nu da zo'u co'e da}

and possibly others. Which one it is depends on where it is used.
It's a du'u in a place that accepts {du'u}s, a nu in a place that
accepts {nu}s.
>>
And up in the air for those that accept more than one absraction.  So, it is not as clear as the full form -- but a lot shorter, which may be enough if the situation is otherwise clear (usually even a good guess at {co'e}).

<<
>BTW, are you actually proposing locutions like {nelci tu'a lo cakla},
>{nelci tu'o du'u ce'u co'e lo cakla}? To me, those don't mean the
>same thing as "I like chocolate".

I don't think he was proposing that. In this case, x2 of nelci
accepts {nu}s, not {du'u}s, so it means {nelci ?? nu co'e lo cakla}.
(I think ?? = {tu'o} or equivalently {lo'e}.)
>>
Damn, now we have {lo'e} being equivalent to (the so far totally unexplained ) {tu'o} -- and used with abstractors diectly to boot: the generic event of doing something with chocolate.  Do note that these do not fit the examples so far (and I mean over the entire month+ of this discussion).


cowan:
on jorda n
<<
> I say {mi djica lenu lo pavyseljirna cu klama ti} there's nothing wrong
> with the bridi, as I really do desire that su'o lo ro pavyseljirna
> come (even if ro = 0; the su'o is just the number I'm wanting).

There *is* nothing wrong, because nu-events exist even if the things inside
don't.  But lo pavyseljirna cu blabi, "some unicorn is white", that's
rubbish.
>>
No, just false, because there aren't any unicorns.
Yahoo! Groups Sponsor
ADVERTISEMENT

To unsubscribe, send mail to lojban-unsubscribe@onelist.com

Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
--part1_152.14477cfa.2aba70de_boundary--