[lojban] Re: tu'o usage
And Rosta
jjlla...@hotmail.com writes:
[...]
#<<
#I said that changing inner {ro} to {me'iro} was nonsense, not
#that the passage of a negation boundary did not affect the inner
#quantifier. If the inner quantifier is {ro}, then nothing is changed,
#because {ro} as inner quantifier in fact adds nothing, neither
#claim nor presupposition: {lo'i broda} always has ro members
#by definition.
#>>
#Let's see, negation boundaries do affect inner quantifiers except in the case
#of the most common one. That does seem to violate the notion that they are
#affected -- a rule is a rule after all and the effects of negation boundaries
#on the universal quantifier is one of the best established of such rules.
So called "inner quantifiers" should be called "inner cardinality indicators"
-- just as PA does not always function as a quantifier (e.g. in {li pa}), so
in {lo PA broda} it functions as an indicator of cardinality, not as a
quantifier.
Negation boundaries affect all inner cardinality indicators, but since ro
does not ascribe any cardinality to the set, it is vacuously affected.
#As for {ro} adding nothing, it does at least exclude {no} (I know you disagree,
#but this is my turn) and, further, as the default, can be stuck in anywhere
#nothing is explicit (which is why I take it that nebgation does not affect
#it). What about {le broda}, where the default is {su'o} : does {naku le
#broda} go over to {ro le no broda naku}? If not, why not?
Not. Because *everything* within a le- phrase IS presupposed -- that is
the very nature of le-.
&:
#<<
#> Even
#> mathematicians and linguists pretty much get this right.
#
#The the confusion may be about what "this" is.
#>>
#That "all" has existential import. I guess I have to take back "linguists"
#-- but, gee, my people (Partee, Bill Bright, various Lakoffs) and McCawley
#had it right.
Does McCawley deal with it in _Everything linguists always wanted to
know about logic_?
I think that perhaps part of the issue concerns whether restricted
quantification exists in Lojban -- whether {da poi broda cu brode} means
something different from {da ge broda gi brode}. I suspect you
would say that the former but not the latter entails {da broda}.
If I'm right about this, at least I can understand where you're coming
from, and will be in a position to think properly about the issues.
[...]
#[Calling citation -- or the threat of such -- Argument from Authority is
#prejudicial, even when modified by "legitimately": loading.]
As I said, I think threatened citation and Arg from Auth is legitimate,
but I don't see much difference between them.
#<<
#My brand of English has "all" and "every" as nonimporting, and
#"each" as importing, but "each" quantifies over a definite class
#(i.e. it means "each of the"), so the importingness is probably
#an artefact of the definiteness.
#>>
#I'll take your word for it, even though I have found (as have more formal
#empiricial researchers on the issue) that people are not very clear about
#this and often display patterns incompatible with their conscious beliefs on
#the topic. In particular, though, people who allow both importing and
#non-importing meanings usually group "every" with "each" (as it is
#historically as well = "ever each"), so you constitute a group either new or
#too small to have been noted before. Your explanation for the position of
#"each" probably accounts for your case, which is basically a "no importing"
#one.
Everybody groups "every" and "each" together separate from "all", because
the former are distributive: "Every thing is", "Each (thing) is", "All (things) are".
If you can give me references on the importingness of "all" and "every" I
will go and look them up. I am skeptical about there being dialect differences,
but I shouldn't prejudge.
#<<
#If you have the logical formula:
#
# P and ASSERTED: Q
#
#how should that be expressed grammatically so that it comes out
#like
#
# Q PRESUPPOSED: and P
#>>
#I don't follow the formula, I think. Suppose that P presupposes Q. Then the
#whole situation is "P funny-and Q."
At a presyntactic/prelexical level I think it is "P and I-HEREBY-ASSERT Q"
#Negating this would be "not P funny-and Q,"
Polishly "funny-and Q not P", not "not funny-and Q P", I take it you mean.
#{na'i}ing it would be either "not(P and Q)" ("and" not at all funny) or (better)
#"not Q whether P" ("whether" = Lojban {u}).
The former would Griceanly imply the latter.
#lioNEL:
#<<
#Indeed, I take the opposing views. As xorxes pointed it out, the whole
#issue seems to decide wether the INNER part is claimed or presupposed.
#IMO it is naturally claimed (the ro case being special, see below):
#I would find it very strange, to say the least, to consider something
#explicitly stated as something presupposed.
#>>
#Me too. But INNER is not stated, merely displayed and, thus, open to a
#variety of interpretations, of which "presupposed" is one. "Asserted" is
#another, but I can't find any cases of it actually working that way anywhere
#and many cases of the presupposing version, even without {ro}.
Can you cite some of the many cases of the presupposing version without
ro?
--And.
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sell a Home for Top $
http://us.click.yahoo.com/RrPZMC/jTmEAA/MVfIAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Lionel Vidal
>>lionel:
> > But to be consistent, this should also be true in when INNER actually
set
> > the cardinality of the underlying subset of broda, as in{lo ci broda cu
> > brode},
> > which I would read as {ge lo'i broda cu ci mei gi lo broda cu brode},
> > and has such is indeed affected by negation boundaries. Or do you
consider
> > than this cardinality is never really asserted, but belongs to {na'i}
> > domain,
> > i.e. be the same kind of presupposed implications, despite being
explicitly
> > stated?
> >pc:
> > I would claim that it is true in the case of {lo ci broda} as well
> > and thus that the expansion you propose is not correct. That is,
> > {lo ci broda na brode} doesn't come out as {ro lo na'e ci broda naku
> > brode}. That is, yes, INNER is part of the {na'i} domain (I thought
> > I said that explicitly. Sigh!)
>>and:
> You had said that explicitly, but I think Lionel, like me, was taking
> the opposing view.
Indeed, I take the opposing views. As xorxes pointed it out, the whole
issue seems to decide wether the INNER part is claimed or presupposed.
IMO it is naturally claimed (the ro case being special, see below):
I would find it very strange, to say the least, to consider something
explicitly stated as something presupposed.
xorxes
>So if the inner quantifier is claimed, the manipulation rules are
>not at all simple,
That is what I was trying to show with my negation of
{lo ci broda cu brode}.
>except when the inner is non-importing ro,
>which makes no claim or presupposition. Yet another argument
>in favour of non-importing ro.
IMO, for me it is now the main argument in its favour, as it solves my
negation moving problems in a satifactory way.
mu'omi'e lioNEL
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Home Selling? Try Us!
http://us.click.yahoo.com/QrPZMC/iTmEAA/ySSFAA/GSaulB/TM
py...@aol.com
<<
I don't remember it being settled and decided (by whom?) the way
you want. For me {ro} is non-importing.
>>
Actually, on 15-03-02 you set forth (again) your system, acknowledging that it was aberrant, and claiming for it a simplicity that it turned out not to have when actually applied or worked out theoretically. That aside you acknowledged the correctness -- within Lojban of the importing system. Your {ro} is just {ro ni'u}, which is rarely useful and on those occasions is easily reached by falling back to standard Logic notation (your claim that ordinary {ro} can be reached in the same way from {ro ni'u} is true, but hardly an efficient suggestion. Of course, we still disagree about whether "every" -- you probably say "all" -- really has existential import.)
<<
So for you {ga broda ginai broda} can be false for selected broda?
For me it's a tautology.
>>
I'm not sure that I understand this, but I suppose you mean {lo brode ga broda ginai brode} can be false. Yes, it can, if there are no brode. But, note, {naku le brode ga broda ginai brode} is false as well, so tautological status is not affected -- the sentence is merely ill-formed at a low level.
| Yahoo! Groups Sponsor | |
|
|
 |
py...@aol.com
<<
After reading the nice page of xorxes on the Wiki exposing clearly
the controversal sentences, I am not sure any more, and you and
xorxes may be just plain right on the ground of the same practical
and ease of use I advocated. I have to think more about it
>>
I must read -- and respond to -- xorxes piece (the short stays aol is allowing me on-line cuts into my wiki reading). I think I know what it says, though, and while it is probably mostly correct, the conclusions would draw are not supposrted by the evidence adduced. Gieve Logic and common snese a turn before you make up your mind (not that either of thses have much of a record of success in the logical language).
<<
consider {OUTER lo INNER broda na brode}
Would you say that this is true when:
the brode relationship is false
or the cardinality of the underlying set of broda given by INNER is false
or the cardinality of the broda involved in the relationship given
by OUTER is false
(with of course inclusive or).
>>
1) yes
2)yes -- but because {na'i} is true (i.e., {OUTER lo INNER broda cu brode} is also false.)
3) yes
<<
I am not sure of
what you mean with 'presupposition implicature'.
>>
I'm not sure I can give the official definition (it is a piece of jargon, so don't worry if you can't work it out from its parts). The idea, as it affects the present case, certain things must be true for a particular sentence to be asserted or, classically, a particular question asked. To be meaningfully asked "Have you stopped beating your wife" requires at least that you have a wife and that you have beaten her. Failing this the question is meaningless. To answer either "Yes" or "No" to the question is to admit to all the presuppositions -- and to add either that you have now stopped or that you are continuing the beating.
So the point here is that uttering a sentence with {lo INNER broda} in it -- even if INNER is implicit -- commits you to there being INNER broda. If there are not, then the whole is meaningless, {na'i}-false -- and so is its denial. Negations and negation boundaries do not affect this inner value. We do not say that the negation of {lo broda cu brode}, {lo brode na brode} is going to result in {ro lo me'iro brode naku brode} when we move the negation through, but just {ro lo broda naku brode} where {lo broda} is still implictly {lo ro broda} (I'm not even sure just what {me'iro} might mean as an INNER).
py...@aol.com
<<
la pycyn cusku di'e
>All of the supposed complications are exactly paralleled for your system,
Not really. In my system, ro = no naku = naku su'o naku = naku me'iro.
Some of those don't work with other systems. That's what makes them
complicated
>>
What won't work? And, by the way, which of the half dozen systems I have suggested and played with (including what I take is now yours) is being labelled "pc's system?"
<<>and
>more likely to need to be used there, since the non-importing {ro} is less
>common in actual usage than the importing.
How can you tell? In most usage we don't deal with empty sets,
so it makes no difference. A clearly non-importing case would
be saying something like "the only world where every politician
is honest is a world with no politicians" (we don't like
politicians much around here these days).
>>
Do you know a place where they do (I can't remember what hand-in-the-till or bumping-off-opponents ploy Argentina has been through most recently)?
I take the fact that we don't usually deal with empty sets as a reason to say that inporting {ro} is basic: it is the one we usually need.
I don't see what importing/non-importing has to do with the "clearly non-importing case:" "for all worlds, if there is a world where every politi --- Oh, never mind, just got it.
<<
>Also, since Lojban is following
>formal logic, it is more or less forced to the importing form that that
>logic
>uses (the apparent exception being an aberration that ran briefly form
>about
>1858 to 1958).
Are those the dates of some particular events?
>>
Boole's Laws of Thought to my first paper on the subject (class, not published). Boole gave a (not quite the first modern) expression to the non-importing reading of "All S is P" (but, of course, using the external importing "all" and something equivalent to conditionalization of the subject-in-the-predicate).
&:
<<
The only place where the importing/nonimportingness of ro makes
an obvious difference, as far as I can see, is as an inner quantifier,
and to my mind it is very useful that {lo'i broda}={lo'i ro broda} not exclude
{lo'i no broda}. If I want to exclude {lo'i no broda} I can say
{lo'i su'o broda}.
The rest of the discussion is too abstruse and too angelic-pinhead-
terpsichorean even for me! I humbly place my faith in xorxes to
show me the light & guide me on the True Path.
>>
I think there is a better solution to the problem of empty classes that allowing {ro} to include {no} (quite beside the obvious one to make {su'o} the implicit INNER everywhere).
The rest is pretty abstruse and largely irrelevant, since we end up about the same place regardless and xorxes' view is no wose -- and no better -- than any of the others.
xorxes:
<<
la pycyn cusku di'e
>So the point here is that uttering a sentence with {lo INNER broda} in it
>--
>even if INNER is implicit -- commits you to there being INNER broda.
case that there are ro broda with non-importing ro, and there is
therefore no commitment. (The outer {su'o} of course does require
there to be at least one.)
>>
This does sound like ypou agree with me, albeit by the back door. I'll take that, since the use is what matters for the moment, not the theory.
<<
>We do not say that the negation of {lo broda cu brode}, {lo brode na brode}
>is going to result in {ro lo me'iro brode naku brode} when we move the
>negation through,
claimed or presupposed.
>>
And this looks like unqualified -- even enthusiastic -- agreement!
<<
>but just {ro lo broda naku brode} where {lo broda} is still
>implictly {lo ro broda} (I'm not even sure just what {me'iro} might mean as
>an INNER).
and {me'iro} can't be {ro}.
>>
I take this to mean that whatever number we put in as INNER, it is, in fact {ro} (assuming we have it right) and so {me'iro} makes no sense. Check.
<<
"Inner quantifiers" are not quantifiers. They make a claim or
a presupposition about the _cardinality_ of the underlying set,
they do not quantify over it. (In the case of non-importing {ro}
no claim is made nor presupposed about the cardinality, so the
question does not even come up.)
>>
Well, I don't quite see how this use of PA is radically different from the use in OUTER, except about the identity of the set involved, but that doesn't matter in the present discussion, whose point was just that the passage of a negation boundary over a description did not change the inner quantifiers (or whatever) and so they have a different status from the outer one.
lioNEL:
<<
I agree, but I would have found more 'natural' for a logical language
to avoid these special cases by having no conv-implic and maybe
some explicit mechanism (special cmavos maybe) to allow it on demand.
Truth value affectations would have been much cleaner.
>>
It is not clear that we can get away from presuppositions altogether (actually, it is clear that we can't), but Lojban did work at minimizing them, Still some remain, although they could (and probably will) be eliminated in a general theory of the logicalizing of Lojban sentences -- one of And's projects, I think.
<<
But to be consistent, this should also be true in when INNER actually set
the cardinality of the underlying subset of broda, as in{lo ci broda cu
brode},
which I would read as {ge lo'i broda cu ci mei gi lo broda cu brode},
and has such is indeed affected by negation boundaries. Or do you consider
than this cardinality is never really asserted, but belongs to {na'i}
domain,
i.e. be the same kind of presupposed implications, despite being explicitly
stated?
>>
<<
> The rest of the discussion is too abstruse and too angelic-pinhead-
> terpsichorean even for me! I humbly place my faith in xorxes to
> show me the light & guide me on the True Path.
Hear, hear.
>>
Well, yes, but if you sign up for a language based on formal logic, you have to expect that a little formal logic turns up from time to time. This is a time. (It will go away fast into a three-way again.)
py...@aol.com
<<
where there is dispute about whether some
piece of meaning is within the scope of what is asserted or
outside it (i.e. presupposed/conventionally implicated), the
default/null hypothesis is that it is within. This is because
Lojban makes little if any use of presupposition/conventional
implicature (outside of UI, at least), does not discuss it in
Woldy, and has no established tradition of acknowledging its
existence in Lojban.
>>
I am hesitant to agree to such a sweeping principle, lest it be wielded without looking at the case at issue and hence stifle debate. However, I thnk that there are a variety of facts that suggest that internal quantification is presuppositional. Several have been mentioned already in this discussion, but the main one has not: the implicit {su'o} and {ro} and actually stated numbers as well, are never changed at the passage of a negation boundary: the implicit value with {le} is {su'o} throughout, and for {lo}, {ro}. These values can be inserted in any context without changing the utterance as a whole.
| Yahoo! Groups Sponsor | |
|
|
 |
And Rosta
> a.r...@lycos.co.uk writes:
> <<
> FWIW, my schooling is such that I automatically take ro broda and
> ro da poi broda to NOT entail da broda. So if for no other reason
> than sheer habit, I prefer nonimporting ro.
> >>
> Interesting. What were you schooled as and where?
As a linguist, at University College London. I was only ever taught
by linguists (counting formal semanticists as such), never by out and
out logicians.
> Even
> mathematicians and linguists pretty much get this right.
The the confusion may be about what "this" is.
> But, since
> what you say is sorta mixed categories, I suppose you might have
> gotten that all from someone confused by a semieducation in the area.
> I suppose you mean {ro broda cu brode} and {ro da poi broda cu
> brode} entail {da broda} (or you mean "implicate" rather than
> "entail").
I mean that {ro broda cu brode} and {ro da poi broda cu brode}
DON'T ENTAIL {da broda}. (Caps for emphasis, not shouting.)
That is, they are equivalent to {ro da ga na broda gi brode}.
In saying that, I'm just describing my habits of interpretation.
> It is quite true that for many people much of the time
> "All broda are brode" does not entail "There are broda," but by the
> same token, {ro broda cu brode} or {ro da poi broda cu brode} are not
> translations of that sentence (in that sense),
Right. As I understand it, this is your position, legitimately backed
up by an Argument from Authority, which I'm not confident I'm capable
of understanding, while Jorge takes the contrary view.
I am saying that I hope Jorge is right, so as to spare me having to
unlearn my habits. Of course, if I'm thereby committing some horrible
logical fallacy I would want to recant, but I don't (yet) see why
{ro broda cu brode} and {ro da poi broda cu brode} can't be strictly
equivalent to {ro da ga na broda gi brode}.
> rather {ro da zo'u
> ganai da broda gi da brode} is, just like we learned in Logic 01.
> {ro broda cu brode} etc. translate what is in my dialect "Every broda
> is a brode" or "Each broda is a brode." Some native speakers of
> English claim that their dialect does not make this distinction, but,
> curiously, they then divide into two groups over which of the two
> possibilities there uniform universal is -- with most going for the
> non-importing admittedly.
My brand of English has "all" and "every" as nonimporting, and
"each" as importing, but "each" quantifies over a definite class
(i.e. it means "each of the"), so the importingness is probably
an artefact of the definiteness.
> <<
> But I go along with the general desire to minimize presupposition
> (though Lionel's suggestion of an explicit marker of presupposition
> might be nice, though I'll leave it to someone else to propose it,
> since I'm weary of incurring the scorn of Jay and Jordan).
> >>
> The trick seems to be a metaconjuction that works at one level like
> an ordinary conjunction but at another level is not attached until
> all the other operations have been gone through (see some of the
> stuff about interdefining the various types of quantifiers earlier this year).
I take it that the 'operations' are 'gone through' from inside to
outside, i.e. mainly right to left in a Lojban-style syntax? That is,
if X has scope over Y, then Y is processed before X? In that case,
yes.
But it's in fact not easy to see how to turn it into a concrete
proposal. If you have the logical formula:
P and ASSERTED: Q
how should that be expressed grammatically so that it comes out
like
Q PRESUPPOSED: and P
?
--And.
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Plan to Sell a Home?
http://us.click.yahoo.com/J2SnNA/y.lEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
py...@aol.com
<<
The difference between 'stated' and 'displayed' in a sentence that the
origanotor means to be true seems fallacious to me.
But anyway, there is a example in CLL p.131, where Cowan says:
'Using exact numbers as inner quantifiers.....you are STATING that exactly
that many things exist'
and also
'lo ci gerku cu blabi'.... CLAIMS also that there are only three dogs
in the universe!
>>
The difference would appear when someone claimed it was false (English not being good at distinctions here) or when a negation was moved through. On the latter, I have yet to see a case of INNER change.
As for CLL, I have already noted that it seems in many ways oblivious to presuppositions. What I suppose is going on there is just that, if there are not exctly three dogs, we would not accept the claim even if all the dogs there were were white (I'm not, by the way, sure this is true -- many might say "Yeah, except there are actually four dogs" or some such thing).
Jorge Llambias
la pycyn cusku di'e
>So the point here is that uttering a sentence with {lo INNER broda} in it
>--
>even if INNER is implicit -- commits you to there being INNER broda.
But when INNER is {ro} (which is the default) it is always the
case that there are ro broda with non-importing ro, and there is
therefore no commitment. (The outer {su'o} of course does require
there to be at least one.)
>We do not say that the negation of {lo broda cu brode}, {lo brode na brode}
>is going to result in {ro lo me'iro brode naku brode} when we move the
>negation through,
Of course not! That's nonsense whether the inner quantifier is
claimed or presupposed.
>but just {ro lo broda naku brode} where {lo broda} is still
>implictly {lo ro broda} (I'm not even sure just what {me'iro} might mean as
>an INNER).
{me'iro} is nonsense as inner, because the inner is always {ro},
and {me'iro} can't be {ro}.
"Inner quantifiers" are not quantifiers. They make a claim or
a presupposition about the _cardinality_ of the underlying set,
they do not quantify over it. (In the case of non-importing {ro}
no claim is made nor presupposed about the cardinality, so the
question does not even come up.)
mu'o mi'e xorxes
_________________________________________________________________
MSN Photos is the easiest way to share and print your photos:
http://photos.msn.com/support/worldwide.aspx
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Home Selling? Try Us!
http://us.click.yahoo.com/QrPZMC/iTmEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
And Rosta
> &:
> <<
> The only place where the importing/nonimportingness of ro makes
> an obvious difference, as far as I can see, is as an inner quantifier,
> and to my mind it is very useful that {lo'i broda}={lo'i ro broda} not exclude
> {lo'i no broda}. If I want to exclude {lo'i no broda} I can say
> {lo'i su'o broda}.
>
> The rest of the discussion is too abstruse and too angelic-pinhead-
> terpsichorean even for me! I humbly place my faith in xorxes to
> show me the light & guide me on the True Path.
> >>
> I think there is a better solution to the problem of empty classes
> that allowing {ro} to include {no} (quite beside the obvious one to
> make {su'o} the implicit INNER everywhere).
I don't see any advantage to making su'o the default inner. My sense
is that in the one case where no might plausibly be the cardinality,
viz after lo'i, one generally wants to allow for the possibility of
no.
What other solutions are there?
> The rest is pretty abstruse and largely irrelevant, since we end up
> about the same place regardless and xorxes' view is no wose -- and no
> better -- than any of the others.
Yes. I was just excusing myself from participation.
FWIW, my schooling is such that I automatically take ro broda and
ro da poi broda to NOT entail da broda. So if for no other reason
than sheer habit, I prefer nonimporting ro.
> lioNEL:
>
> <<
> I agree, but I would have found more 'natural' for a logical language
> to avoid these special cases by having no conv-implic and maybe
> some explicit mechanism (special cmavos maybe) to allow it on demand.
> Truth value affectations would have been much cleaner.
> >>
> It is not clear that we can get away from presuppositions altogether
> (actually, it is clear that we can't), but Lojban did work at
> minimizing them, Still some remain, although they could (and
> probably will) be eliminated in a general theory of the logicalizing
> of Lojban sentences -- one of And's projects, I think.
That's rather complimentary! Excessively so. I thought we were all
(i.e. those who give a shit) engaged in the logicalizing of Lojban
sentences, as far as we can. But anyway, as I pointed out to Lionel,
I do hold that Lojban has conventional implicature/presupposition
-- I gave some UI and {le}-series gadri as examples.
But I go along with the general desire to minimize presupposition
(though Lionel's suggestion of an explicit marker of presupposition
might be nice, though I'll leave it to someone else to propose it,
since I'm weary of incurring the scorn of Jay and Jordan).
> <<
> But to be consistent, this should also be true in when INNER actually set
> the cardinality of the underlying subset of broda, as in{lo ci broda cu
> brode},
> which I would read as {ge lo'i broda cu ci mei gi lo broda cu brode},
> and has such is indeed affected by negation boundaries. Or do you consider
> than this cardinality is never really asserted, but belongs to {na'i}
> domain,
> i.e. be the same kind of presupposed implications, despite being explicitly
> stated?
> >>
> I would claim that it is true in the case of {lo ci broda} as well
> and thus that the expansion you propose is not correct. That is,
> {lo ci broda na brode} doesn't come out as {ro lo na'e ci broda naku
> brode}. That is, yes, INNER is part of the {na'i} domain (I thought
> I said that explicitly. Sigh!)
You had said that explicitly, but I think Lionel, like me, was taking
the opposing view.
> taral (quoting &)
> <<
> > The rest of the discussion is too abstruse and too angelic-pinhead-
> > terpsichorean even for me! I humbly place my faith in xorxes to
> > show me the light & guide me on the True Path.
>
> Hear, hear.
> >>
> Well, yes, but if you sign up for a language based on formal logic,
> you have to expect that a little formal logic turns up from time to
> time. This is a time. (It will go away fast into a three-way again.)
"Hear, hear" is not incompatible with "We're glad pc & Jorge thrash
these issues out satisfactorily without us all having to get
involved". That's what I meant, anyway.
--And.
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sell a Home for Top $
http://us.click.yahoo.com/RrPZMC/jTmEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
py...@aol.com
<<
My sense
is that in the one case where no might plausibly be the cardinality,
viz after lo'i, one generally wants to allow for the possibility of
no.
What other solutions are there?
>>
<<
FWIW, my schooling is such that I automatically take ro broda and
ro da poi broda to NOT entail da broda. So if for no other reason
than sheer habit, I prefer nonimporting ro.
>>
<<
But I go along with the general desire to minimize presupposition
(though Lionel's suggestion of an explicit marker of presupposition
might be nice, though I'll leave it to someone else to propose it,
since I'm weary of incurring the scorn of Jay and Jordan).
>>
| Yahoo! Groups Sponsor | |
|
|
 |
py...@aol.com
<<
I did not accept that it was correct for Lojban. Only that it would
not be logically inconsistent. But I consider it a bad choice,
because it is more complicated, and thus incorrect for Lojban in
that sense.
>>
All of the supposed complications are exactly paralleled for your system, and more likely to need to be used there, since the non-importing {ro} is less common in actual usage than the importing. Also, since Lojban is following formal logic, it is more or less forced to the importing form that that logic uses (the apparent exception being an aberration that ran briefly form about 1858 to 1958).
<<
>Do you really, by the way, want {ro da zo'u ganai da broda gi da brode} to
>be
>true even if there is nothing in the world at all?
Yes, vacuously true. I can't imagine a context where it
would come up, though.
>>
Oops! See how hard it is to even think of non-importing affirmative universals. I meant to say {ro da broda} but immediately fell into the formula needed in normal discourse to make "non-importing" claims.
And Rosta
> a.r...@lycos.co.uk writes:
> <<
>
> where there is dispute about whether some
> piece of meaning is within the scope of what is asserted or
> outside it (i.e. presupposed/conventionally implicated), the
> default/null hypothesis is that it is within. This is because
> Lojban makes little if any use of presupposition/conventional
> implicature (outside of UI, at least), does not discuss it in
> Woldy, and has no established tradition of acknowledging its
> existence in Lojban.
>
> >>
> I am hesitant to agree to such a sweeping principle, lest it be
> wielded without looking at the case at issue and hence stifle debate.
The advantage of having sweeping principles is precisely that
they give a default -- that ceteris paribus they tell us what X means,
when X is a case that has not properly been considered hitherto.
Generally the default principle is the one with the greatest generality.
That does not been that motivated exceptions aren't allowed. For example, the
rule that a sumti is quantified in the local bridi is overridden by tu'a-marked
sumti on sound utilitarian grounds.
> However, I thnk that there are a variety of facts that suggest that
> internal quantification is presuppositional. Several have been
> mentioned already in this discussion,
Can you remind me what they were? They must have passed me by without
my noticing.
> but the main one has not: the
> implicit {su'o} and {ro} and actually stated numbers as well, are
> never changed at the passage of a negation boundary: the implicit
> value with {le} is {su'o} throughout, and for {lo}, {ro}. These
> values can be inserted in any context without changing the utterance
> as a whole.
Can you give examples? What you say doesn't seem true to me, but
I think I may be missing your intended point.
The point needs to be stated about {lo}, not {le}, because the internals
of {le} sumti scope outside everything. E.g. {le broda na brode} is
not true if the referent is brode but not broda.
--And.
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sell a Home for Top $
http://us.click.yahoo.com/RrPZMC/jTmEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
And Rosta
> aro...@uclan.ac.uk writes:
> <<
>
> So called "inner quantifiers" should be called "inner cardinality indicators"
> -- just as PA does not always function as a quantifier (e.g. in {li pa}), so
> in {lo PA broda} it functions as an indicator of cardinality, not as a
> quantifier.
>
> Negation boundaries affect all inner cardinality indicators, but since ro
> does not ascribe any cardinality to the set, it is vacuously affected.
>
> >>
> from quantifier DeMorgan, but what exactly. Where is a case of it
> (other than ro = ro) applied?
{lo ro broda cu brode} means something like "some broda is brode
and the cardinality of the set of broda is the cardinality of the
set of broda" -- I can't think of a better way of putting it,
unfortunately. The Lojban is not tautologous like my English version,
but it is as vacuously uninformative. Maybe "some broda is brode
and lo'i broda has a cardinality" might be better.
So in that case, {lo ro broda na brode} means "it is not the case
that both some broda is brode and the cardinality of the set of
broda is the cardinality of the set of broda (or, alternatively,
lo'i broda has cardinality". Since it obviously is the case that the cardinality
of lo'i broda is the cardinality of lo'i broda -- or that
lo'i broda has a cardinality -- the inevitable inference is that
it is not the case that some broda is brode.
In summary, then, the 'claim' made by {ro} in {lo ro broda} is
uninformative and true by definition. When it is negated it is
analogous to:
p and x=x
negated yields:
not (p and x=x)
from which the only valid conclusion is "not p".
> <<
> Not. Because *everything* within a le- phrase IS presupposed -- that is
> the very nature of le-.
> >>
> Huh? I guess we mean totally different things by "presupposed". What
> is referred to may not be literally as it is described, but that is
> not presupposition, only the vagaries of human humor (when it occurs
> -- it is not very common so far in Lojban or any other language I can
> think of). Even if thiis is presupposition, I haven't anywhere seen
> it suggested before that (always remembering that with me memory is
> not a pramana) the INNER was also subjectively defined.
Firstly, but less importantly, 'definite descriptions' -- closely
comparable to Lojban le -- are fairly standard exx of presupposition,
I think.
Second, the essence of le is specificity, with nonveridicality something
of a by-product. My personal view is that logically specificity
involves existential quantification outside the scope of the operator
that carries illocutionary force (e.g. assertive force). If the
sumti tail is held to also be outside the scope of the illocutionary
operator, then nonveridicality is an automatic consequence. It is
also my personal view that the essence of presupposition/conventional
implicature that it is outside the scope of the illocutionary
operator. Therefore I see specificity as "presuppositional existential
quantification".
Finally, I too don't recall it ever having been established that
the inner cardinality indicator is nonveridical (=subjectively
defined), but I think that is the more consistent position to take.
>
> <<
> #That "all" has existential import. I guess I have to take back "linguists"
> #-- but, gee, my people (Partee, Bill Bright, various Lakoffs) and McCawley
> #had it right.
>
> Does McCawley deal with it in _Everything linguists always wanted to
> know about logic_?
> >>
> Yes (I don't have my copy handy, so I can't give a citation). Right
> after he deals with the logicians' view (which he gets almost right
> -- he just doesn't note that the explicit quantifier in the
> quantified conditional form is importing), he remarks that the actual
> language situation is pretty clearly importing (though I don't think
> he uses that exppression) and gives some examples.
Thanks. I'll read this up.
> <<
> I think that perhaps part of the issue concerns whether restricted
> quantification exists in Lojban -- whether {da poi broda cu brode} means
> something different from {da ge broda gi brode}. I suspect you
> would say that the former but not the latter entails {da broda}.
> If I'm right about this, at least I can understand where you're coming
> from, and will be in a position to think properly about the issues.
> >>
> On the contrary, I would have to insist that these two are
> equivalent. At most I have a problem with {ro broda cu brode} or {ro
> da poi broda cu brode} and {ro da zo'u ganai d a broda gi da brode}.
> (As xorxes points out, it is really only A --- and maybe occasionally
> O -- and maybe even more occasionally E -- that is a problem).
So I'm back to square one then, understandingwise. Never mind, I'll
see if McCawley enlightens me.
> <<
> Everybody groups "every" and "each" together separate from "all", because
> the former are distributive: "Every thing is", "Each (thing) is",
> "All (things) are".
> >>
> But you just didn't, although the question was not about
> distribution.
What I meant was that for most people the salient criterion for
grouping is not importingness, and that each+every vs all is a
salient grouping.
> I suppose that one source of the typical (but hardly
> universal, obviously) tendency to take "each" and "every" as
> importing is the difficulty of imagining distribution in a null set.
That makes sense, yet it's easy to come up with examples of
nonimporting "every": "everyone who answers all questions successfully
will pass the course" -- this does not claim that some has answered
or will answer all questions successfully.
> <<
> If you can give me references on the importingness of "all" and "every" I
> will go and look them up. I am skeptical about there being dialect
> differences,
> but I shouldn't prejudge.
> >>
> The loc class is Zeno Vendler, "Each and Every, Any and All" Mind,
> v.71 (April 1962), pp 145 - 160. This is reprinted in a collection
> Vendler's papers, information on which I can't find (I'll sure be
> glad when I have used my "organized" library enought that I can
> start finding thing in it again). The results are summarized in a
> section "Any and All" in the Encyclopedia of Philosophy v1 pp 131-3
> in the first edition. Vendler there cites a number of other sources,
> including somewhat linguistic ones like Klima "Negation in English"
> in Fodor and Katz.
Thanks.
>
> <<
> #I don't follow the formula, I think. Suppose that P presupposes Q.
> Then the
> #whole situation is "P funny-and Q."
>
> At a presyntactic/prelexical level I think it is "P and I-HEREBY-ASSERT Q"
> >>
> I would have put it pretty much tother way round, since I do not
> assert Q in this situation, only P and Q is there to allow me to do
> that assertion (or denial for that matter).
I understand where you're coming from, treating presupposition as
kind of analogous to parenthesis. But from what I say above, you
can see that I see presupposition as basically a matter of scope
relative to the illocutionary operator.
> <<
> #Negating this would be "not P funny-and Q,"
>
> Polishly "funny-and Q not P", not "not funny-and Q P", I take it you mean.
> >>
> Yes; I am careful about my binders when not writing Polish.
>
> <<
> #{na'i}ing it would be either "not(P and Q)" ("and" not at all funny)
> or (better)
> #"not Q whether P" ("whether" = Lojban {u}).
>
> The former would Griceanly imply the latter.
> >>
> It does so in any case, since "not Q" entails "Either not P or not
> Q", i.e. "not (P and Q). The second is better precisely because it
> contains more information.
>
> <<
> Can you cite some of the many cases of the presupposing version without
> ro?
> >>
> Not in any normal sense (there is no standard location to cite from),
> but any case with a specific INNER that is not changed by negation
> passage will do.
Of course, but you seemed to imply that there were many such cases.
I suspect that there aren't, and that if you did find some, the
authors might feel that their usage was an inadvertent mistake.
--And.
------------------------ Yahoo! Groups Sponsor ---------------------~-->
4 DVDs Free +s&p Join Now
http://us.click.yahoo.com/pt6YBB/NXiEAA/MVfIAA/GSaulB/TM
And Rosta
#a.r...@lycos.co.uk writes:
#<<
#> {lo ro broda cu brode} means something like "some broda is brode
#> and the cardinality of the set of broda is the cardinality of the
#> set of broda" -- I can't think of a better way of putting it,
#> unfortunately. The Lojban is not tautologous like my English version,
#> but it is as vacuously uninformative. Maybe "some broda is brode
#> and lo'i broda has a cardinality" might be better.
#>
#> So in that case, {lo ro broda na brode} means "it is not the case
#> that both some broda is brode and the cardinality of the set of
#> broda is the cardinality of the set of broda (or, alternatively,
#> lo'i broda has cardinality". Since it obviously is the case that the
#> cardinality
#> of lo'i broda is the cardinality of lo'i broda -- or that
#> lo'i broda has a cardinality -- the inevitable inference is that
#> it is not the case that some broda is brode.
#>>
#That is, of course, interpretation. What is said is that the cardinality is
#ro, just another number.
But ro functions here as a cardinal number and means "the number equal
to the cardinality of lo'i broda". As a cardinality indicator, it's hard to see
ro as anything other than a dummy filler; it is utterly uninformative.
#So, if negation is going to affect that, it changes it to {me'iro} or some
#other version of {na'e ro}.
But the claim made by the cardinality indicator -- lo'i broda cu PA mei --
is always linked by logical AND to some other proposition, and its
the conjunction that gets negated. So {lo ro broda na brode} means
{ga lo'i broda na ro mei gi su'o broda na brode}. Now if {lo'i broda
na ro mei}, then indeed ro would change to some version of na'e
ro. But that would be so nonsensical that the only plausible
interpretation of {ga lo'i broda na ro mei gi su'o broda na brode}
is as equivalent to {su'o broda na brode}.
#Now, that may in the end
#give an equally contradictory component of some compound, if that is how it
#develops, and tha component may drop out to give just the basic negative,
#without any component about cardinality. But what then justifies adding back
i#n a cardinality component, namely, the one jut dropped out. I suppose that
#it come back in because it is a tautology (why the negation dropped out).
#Well, at least this coheres so far, but I await a case where {lo PA broda}
#negated turns up as {lo na'e PA broda}. (As you point out later, this may be
#a long wait because no one ever uses {lo PA broda}.)
The three Graces, the seven seas, the 51 states of the USA -- those are
equivalent to (ro) lo PA broda.
#<<
#Firstly, but less importantly, 'definite descriptions' -- closely
#comparable to Lojban le -- are fairly standard exx of presupposition,
#I think.
#>>
#You mean as a way to dodge the Russell cases.
Yes. Bald French kings etc.
#Probably, although I
#personally go with the explicit formats, the eight and ninety ways remaining
#to do descriptions -- all of which are also right.
#
#<<
#Second, the essence of le is specificity, with nonveridicality something
#of a by-product. My personal view is that logically specificity
#involves existential quantification outside the scope of the operator
#that carries illocutionary force (e.g. assertive force). If the
#sumti tail is held to also be outside the scope of the illocutionary
#operator, then nonveridicality is an automatic consequence. It is
#also my personal view that the essence of presupposition/conventional
#implicature that it is outside the scope of the illocutionary
#operator. Therefore I see specificity as "presuppositional existential
#quantification".
#>>
#Both of these theories of yours are interesting and need some mulling (my
#first instinct is to like it a lot). But I don't see that the fact that you
#have these theories (even if they turn out to be correct -- i.e., pick of the
l#itter) requires us to use them to explain the present case, which isn't
#directly about that. But, of course, if specificity is a pre-illocution
#quantifier, then INNER, which is that quantifier, is presuppositional.
Hold on: my theory/claims are:
* Specificity is a pre-illocution quantifier. {le} = "pre-illocution-{lo}".
* Everything following {le} is preillocutionary/presuppositional (i.e. the
INNER and the rest of the sumti tail).
* Everything following {lo} is not preillocutionary/presuppositional (i.e.
neither the INNER nor the rest of the sumti tail).
#<<
#Finally, I too don't recall it ever having been established that
#the inner cardinality indicator is nonveridical (=subjectively
#defined), but I think that is the more consistent position to take.
#>>
#So, when I say {le ci mlatu} meaning those four dogs -- or even those four
#cats -- I did not misspeak myself, since I know what I mean -- but can't
#count -- and you know what I mean, even if you can count? Given the history
#of {le}, that seems plausible -- and thoroughly disastrous.
That's right, except it's not disastrous -- it's desirable. I want to provide
info within the le sumti to help you identify the referent set, but so long
as it aids with identification I don't want to *claim* that info is true.
E.g. I want to say that those people -- those in the group that looks,
from where we're standing, like a threesome -- over there are happy.
So I say {le ci prenu cu gleki}. If you identify which people I'm
talking about, agree that they're happy, but go and count them and
find that they are a foursome, I want you to answer {ja'a go'i}, not
{na go'i}, though you are very welcome to also answer {na'i go'i}
too.
(This is a point I picked up from McCawley, btw.)
#<<
#So I'm back to square one then, understandingwise. Never mind, I'll
#see if McCawley enlightens me.
#>>
#Sorry if I didn't fit your hypothesis. I have trouble imagining what you
#could have been thinking of. The importing forms (yes, restricted
#quantification) always entail the corresponding unrestricted ones. If some
#broda is brode then there are broda and something is brode -- the same thing,
#in fact.
I had trouble imaging what I could have been thinking of, too. I know
what I was trying to grope towards, but ended up talking nonsense.
#<<
#yet it's easy to come up with examples of
#nonimporting "every": "everyone who answers all questions successfully
#will pass the course" -- this does not claim that some has answered
#or will answer all questions successfully.
#>>
#But, I would never say that but rather "Any student ,,,"
OK. Maybe there really are interlectal differences, then.
#<<
#I understand where you're coming from, treating presupposition as
#kind of analogous to parenthesis. But from what I say above, you
#can see that I see presupposition as basically a matter of scope
#relative to the illocutionary operator.
#>>
#Neat. And I would add that negation then comes after the presuppositional
#part, either as part of or in the scope of the illocutionary operator.
I'm pleased you like the idea. Yes, certainly negation is within the scope of
the illocutionary operator, except of course for {na'i} which I (and you) would
take to be a negator with scope over everything else (including stuff outside
the scope of the illoc-op.
#<<
#Of course, but you seemed to imply that there were many such cases.
#I suspect that there aren't, and that if you did find some, the
#authors might feel that their usage was an inadvertent mistake.
#>>
#Sorry about the implication (worse, I think I asserted it, without checking
#what data I had). There don't seem to be any cases at all one way or the
#other. So, no conclusion can be drawn from usage, althoug the absence of
#usage might suggest that people just don't quite know what to do with it.
#Or, more likely, that no occasion has arisen for both mentioning the size of
#the set (and precious few for that alone) and negating.
Right. The infrequency of {lo PA broda} is comparable to the infrequency
of {lo broda noi brode ku} -- they both provide information that does not
restrict the referent set.
--And.
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sell a Home for Top $
http://us.click.yahoo.com/RrPZMC/jTmEAA/MVfIAA/GSaulB/TM
Jorge Llambias
la pycyn cusku di'e
><<
> > There never was a difference between {pa lo su'o} and {pa lo ro},
> > so "any more" does not apply.
> >>
>True, only one thought to hold between {lo ro} and {lo su'o}.
Sorry, I don't understand that sentence.
>It hasn't come up because, as you well know after hammering away at it,
>there
>is now no non-importing {ro}
I don't remember it being settled and decided (by whom?) the way
you want. For me {ro} is non-importing.
>(though I cannot remember what the corresponding
>nonimporting expression is, it never being one I need). So {ro lo ro
>pavyseljirna cu blabi} is just false in this world (and its denial
>probably
>is too).
So for you {ga broda ginai broda} can be false for selected broda?
For me it's a tautology.
mu'o mi'e xorxes
_________________________________________________________________
MSN Photos is the easiest way to share and print your photos:
http://photos.msn.com/support/worldwide.aspx
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Home Selling? Try Us!
http://us.click.yahoo.com/QrPZMC/iTmEAA/ySSFAA/GSaulB/TM
py...@aol.com
<<
>What won't work?
Some of the negation relationships. Unless the simple forms are
assigned to (A-E-I+O+) (or (A+E+I-O-) but this one would be silly)
then some of those relationships don't work between the simple forms.
>>
Frinstance? And remember that the other assumption of all of this is that things talked about exist, that the Lojban axiom {ro da zasti} uses importing {ro} and works at the restricted level as well {ro broda cu zasti} (I know that you diagree with this -- at least that it is importing {ro} -- but non-importing {ro} makes precious little sense in the general case.)
<<
I didn't label any system as yours. I understand you argued at some
point for (A+E+I+O+) and at other times for (A+E-I+O-) for the simple
forms. A+/A- is the one we always disagree about, since I want
{ro broda cu brode} to be A- and you want it to be A+.
>>
And that it doesn't matter which, since we are dealing with non-empty sets. On the rare occasion when we want to deal with empty sets, we have to flag it anyhow, so what the basic system is is relatively insignificant.
<<
<<
> >(the apparent exception being an aberration that ran briefly form
> >about
> >1858 to 1958).
>
> >>
>Boole's Laws of Thought to my first paper on the subject (class, not
>published).
paper available online?
>>
I didn't claim that my paper caused the end of an era, only that it marks for me the point when I knew the era was over. If I got away with it in a class paper at UCLA, the logic hub of the universe, then it the old doctrine must have been dead indeed. That paper is not only not available on-line (thank God) but is long ago (well, not all that long ago, since only my last wife makes me clean out my files) destroyed.
<<
I'm glad Boole is on my side then
>>
Boole was, paradoxically, a lousy logician and was working almost entirely outside the tradition (or, rather, with a degenerate humanistic tradition that ran parallel to and against the logician tradition since the Renaissance). I have to admit sadly that Dodgson got sucked into the same tradition, with only minor rebellions (he used O- for a while).
<<
I said that changing inner {ro} to {me'iro} was nonsense, not
>>
Let's see, negation boundaries do affect inner quantifiers except in the case of the most common one. That does seem to violate the notion that they are affected -- a rule is a rule after all and the effects of negation boundaries on the universal quantifier is one of the best established of such rules. As for {ro} adding nothing, it does at least exclude {no} (I know you disagree, but this is my turn) and, further, as the default, can be stuck in anywhere nothing is explicit (which is why I take it that nebgation does not affect it). What about {le broda}, where the default is {su'o} : does {naku le broda} go over to {ro le no broda naku}? If not, why not?
<<
When the inner quantifier is something other than {ro}, then
there is an additional claim or presuposition that {lo'i broda}
has Q members. If it is a claim, then passing through {na}
will affect that claim, but not by changing the inner quantifier
into another inner quantifier. For example (asuming for the
moment that the inner is claimed rather than presupposed):
naku lo pa broda cu brode
= naku ge lo broda cu brode gi pa da broda
= ganai lo broda cu brode ginai pa da broda
= ga ro lo broda naku brode ginai pa da broda
And this cannot be written as {ro lo Q broda naku brode}.
So if the inner quantifier is claimed, the manipulation rules are
which makes no claim or presupposition. Yet another argument
in favour of non-importing ro.
>>
naku lo pa broda cu brode = ro lo pa broda naku brode
= ge ro lo broda naku gi pa da broda
I don't, btw, think doing this is very enlightening, since it leaves out some interesting information.
As for a non-importing {ro} in that position, the rule should still apply, making non-sense of the whole. But using importing {ro} also makes nonsense of the whole (well, a tautology -- or nearly so), so the result is that taking INNER as assertive is an error. On the other hand, taking INNER as presuppositional makes the same sense regardless of the nature of {ro}.
&:
<<
> Even
> mathematicians and linguists pretty much get this right.
>>
<<
> I suppose you mean {ro broda cu brode} and {ro da poi broda cu
> brode} entail {da broda} (or you mean "implicate" rather than
> "entail").
DON'T ENTAIL {da broda}. (Caps for emphasis, not shouting.)
That is, they are equivalent to {ro da ga na broda gi brode}.
>>
<<
In saying that, I'm just describing my habits of interpretation.
> "All broda are brode" does not entail "There are broda," but by the
> same token, {ro broda cu brode} or {ro da poi broda cu brode} are not
> translations of that sentence (in that sense),
up by an Argument from Authority, which I'm not confident I'm capable
of understanding, while Jorge takes the contrary view.
I am saying that I hope Jorge is right, so as to spare me having to
unlearn my habits. Of course, if I'm thereby committing some horrible
logical fallacy I would want to recant, but I don't (yet) see why
{ro broda cu brode} and {ro da poi broda cu brode} can't be strictly
equivalent to {ro da ga na broda gi brode}.
>>
[Calling citation -- or the threat of such -- Argument from Authority is prejudicial, even when modified by "legitimately": loading.]
<<
My brand of English has "all" and "every" as nonimporting, and
"each" as importing, but "each" quantifies over a definite class
(i.e. it means "each of the"), so the importingness is probably
an artefact of the definiteness.
>>
<<
> The trick seems to be a metaconjuction that works at one level like
> an ordinary conjunction but at another level is not attached until
> all the other operations have been gone through (see some of the
> stuff about interdefining the various types of quantifiers earlier this year).
outside, i.e. mainly right to left in a Lojban-style syntax? That is,
if X has scope over Y, then Y is processed before X? In that case,
yes.
>>
<<
If you have the logical formula:
P and ASSERTED: Q
how should that be expressed grammatically so that it comes out
like
Q PRESUPPOSED: and P
>>
would be either "not(P and Q)" ("and" not at all funny) or (better) "not Q whether P" ("whether" = Lojban {u}).
lioNEL:
<<
Indeed, I take the opposing views. As xorxes pointed it out, the whole
issue seems to decide wether the INNER part is claimed or presupposed.
IMO it is naturally claimed (the ro case being special, see below):
I would find it very strange, to say the least, to consider something
explicitly stated as something presupposed.
>>
Jorge Llambias
la pycyn cusku di'e
> > In my system, ro = no naku = naku su'o naku = naku me'iro.
> > Some of those don't work with other systems. That's what makes them
> > complicated
> >>
>What won't work?
Some of the negation relationships. Unless the simple forms are
assigned to (A-E-I+O+) (or (A+E+I-O-) but this one would be silly)
then some of those relationships don't work between the simple forms.
>And, by the way, which of the half dozen systems I have
>suggested and played with (including what I take is now yours) is being
>labelled "pc's system?"
I didn't label any system as yours. I understand you argued at some
point for (A+E+I+O+) and at other times for (A+E-I+O-) for the simple
forms. A+/A- is the one we always disagree about, since I want
{ro broda cu brode} to be A- and you want it to be A+.
>I take the fact that we don't usually deal with empty sets as a reason to
>say
>that inporting {ro} is basic: it is the one we usually need.
That doesn't make sense. When we don't deal with empty sets
the question of import does not even arise. Either importing or
non-importing work just as well. In those cases we don't need
to choose one over the other.
<<
> >(the apparent exception being an aberration that ran briefly form
> >about
> >1858 to 1958).
>
>Are those the dates of some particular events?
> >>
>Boole's Laws of Thought to my first paper on the subject (class, not
>published).
Nobody can accuse you of being too modest! :) Is your epoch making
paper available online?
>Boole gave a (not quite the first modern) expression to the
>non-importing reading of "All S is P" (but, of course, using the external
>importing "all" and something equivalent to conditionalization of the
>subject-in-the-predicate).
I'm glad Boole is on my side then.
><<
>"Inner quantifiers" are not quantifiers. They make a claim or
>a presupposition about the _cardinality_ of the underlying set,
>they do not quantify over it. (In the case of non-importing {ro}
>no claim is made nor presupposed about the cardinality, so the
>question does not even come up.)
> >>
>Well, I don't quite see how this use of PA is radically different from the
>use in OUTER, except about the identity of the set involved, but that
>doesn't
>matter in the present discussion, whose point was just that the passage of
>a
>negation boundary over a description did not change the inner quantifiers
>(or
>whatever) and so they have a different status from the outer one.
I said that changing inner {ro} to {me'iro} was nonsense, not
that the passage of a negation boundary did not affect the inner
quantifier. If the inner quantifier is {ro}, then nothing is changed,
because {ro} as inner quantifier in fact adds nothing, neither
claim nor presupposition: {lo'i broda} always has ro members
by definition.
When the inner quantifier is something other than {ro}, then
there is an additional claim or presuposition that {lo'i broda}
has Q members. If it is a claim, then passing through {na}
will affect that claim, but not by changing the inner quantifier
into another inner quantifier. For example (asuming for the
moment that the inner is claimed rather than presupposed):
naku lo pa broda cu brode
= naku ge lo broda cu brode gi pa da broda
= ganai lo broda cu brode ginai pa da broda
= ga ro lo broda naku brode ginai pa da broda
And this cannot be written as {ro lo Q broda naku brode}.
So if the inner quantifier is claimed, the manipulation rules are
not at all simple, except when the inner is non-importing ro,
which makes no claim or presupposition. Yet another argument
in favour of non-importing ro.
mu'o mi'e xorxes
_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sell a Home with Ease!
http://us.click.yahoo.com/SrPZMC/kTmEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Lionel Vidal
> Me too. But INNER is not stated, merely displayed and, thus, open to a
> variety of interpretations, of which "presupposed" is one. "Asserted" is
> another, but I can't find any cases of it actually working that way
anywhere
> and many cases of the presupposing version, even without {ro}.
The difference between 'stated' and 'displayed' in a sentence that the
origanotor means to be true seems fallacious to me.
But anyway, there is a example in CLL p.131, where Cowan says:
'Using exact numbers as inner quantifiers.....you are STATING that exactly
that many things exist'
and also
'lo ci gerku cu blabi'.... CLAIMS also that there are only three dogs
in the universe!'
Interpreting these CLL sayings as the presupposed version would be
far too streched for my liking. But maybe John could say what he
really meant here.
mu'omi'e lioNEL
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sell a Home for Top $
http://us.click.yahoo.com/RrPZMC/jTmEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Jorge Llambias
la pycyn cusku di'e
><<
> > other choice of import assignment. The way you present it makes
> > it look as if I had acknowledged it being better, something I do
> > not now and did not at that time consider to be true.
> >>
>Not better, just correct for Lojban (rarely the same, in your view).
I did not accept that it was correct for Lojban. Only that it would
not be logically inconsistent. But I consider it a bad choice,
because it is more complicated, and thus incorrect for Lojban in
that sense.
>Do you really, by the way, want {ro da zo'u ganai da broda gi da brode} to
>be
>true even if there is nothing in the world at all?
Yes, vacuously true. I can't imagine a context where it
would come up, though.
mu'o mi'e xorxes
_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com
------------------------ Yahoo! Groups Sponsor ---------------------~-->
4 DVDs Free +s&p Join Now
http://us.click.yahoo.com/pt6YBB/NXiEAA/ySSFAA/GSaulB/TM
py...@aol.com
<<
There never was a difference between {pa lo su'o} and {pa lo ro},
so "any more" does not apply.
>>
True, only one thought to hold between {lo ro} and {lo su'o}.
<<
nonimporting {ro}, but this has not come up in the present
discussion.
{ro lo ro pavyseljirna cu blabi} is true (in worlds with no unicorns)
with nonimporting {ro}, but false with importing {ro}.
{ro lo su'o pavyseljirna cu blabi} is false (in worlds with no
unicorns) both for importing and nonimporting {ro}.
>>
It hasn't come up because, as you well know after hammering away at it, there is now no non-importing {ro} (though I cannot remember what the corresponding nonimporting expression is, it never being one I need). So {ro lo ro pavyseljirna cu blabi} is just false in this world (and its denial probably is too). And the same goes for {ro lo su'o pavyseljirna cu blabi}.
| Yahoo! Groups Sponsor | |
|
|
 |
Lionel Vidal
> > I did not know that the case was settled. In any cases, the book is not
> > at all explicit about this and I think I remember a recent mail from
> > xorxes where he says he does include 0.
> Well, yes; I too think it includes 0.
> lionel:
> This being said, I agree that {ro} should not include the 0 case from
> > a logical and practical point of view.
After reading the nice page of xorxes on the Wiki exposing clearly
the controversal sentences, I am not sure any more, and you and
xorxes may be just plain right on the ground of the same practical
and ease of use I advocated. I have to think more about it
> > {lo pa broda naku brode} = {su 'o lo pa broda naku brode}
> > = {naku zu'o ro lo pa broda cu brode} = {ro lo pa broda na brode}
> I don't agree that the last 2 are equivalent to the first 2, since
> the first 2 mean:
> ge su'o broda na ku brode gi lo'i broda cu pa mei
> and the second two mean:
> na ku ge ro broda cu brode gi lo'i broda cu pa mei
I agree and so now we reach the problem that bothered me for a while:
consider {OUTER lo INNER broda na brode}
Would you say that this is true when:
the brode relationship is false
or the cardinality of the underlying set of broda given by INNER is false
or the cardinality of the broda involved in the relationship given
by OUTER is false
(with of course inclusive or).
I would say yes and this invalidates my previous claims on the implication
of the broda referent existence when using {na}.
And so {tu'o}, because of its lesser sensitivity to the problems
negations involve, seems indeed useful to me now: thank you
for your patient explanations.
and:
>This is because
>Lojban makes little if any use of presupposition/conventional
>implicature (outside of UI, at least),
pc:
>I thnk that there are a variety of facts that suggest that internal
>quantification is presuppositional
Sorry, I may have a problem with my english there: I am not sure of
what you mean with 'presupposition implicature'.
mu'omi'e lioNEL
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Home Selling? Try Us!
http://us.click.yahoo.com/QrPZMC/iTmEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
py...@aol.com
<<
I may have acknowledged that your position is as consistent as any
other choice of import assignment. The way you present it makes
it look as if I had acknowledged it being better, something I do
not now and did not at that time consider to be true.
>>
Not better, just correct for Lojban (rarely the same, in your view).
| Yahoo! Groups Sponsor | |
|
|
 |
Jorge Llambias
la pycyn cusku di'e
>All of the supposed complications are exactly paralleled for your system,
Not really. In my system, ro = no naku = naku su'o naku = naku me'iro.
Some of those don't work with other systems. That's what makes them
complicated.
>and
>more likely to need to be used there, since the non-importing {ro} is less
>common in actual usage than the importing.
How can you tell? In most usage we don't deal with empty sets,
so it makes no difference. A clearly non-importing case would
be saying something like "the only world where every politician
is honest is a world with no politicians" (we don't like
politicians much around here these days).
>Also, since Lojban is following
>formal logic, it is more or less forced to the importing form that that
>logic
>uses (the apparent exception being an aberration that ran briefly form
>about
>1858 to 1958).
Are those the dates of some particular events?
>Oops! See how hard it is to even think of non-importing affirmative
>universals. I meant to say {ro da broda} but immediately fell into the
>formula needed in normal discourse to make "non-importing" claims.
{ro da broda} would be true in an empty universe, yes. Is that
problematic?
mu'o mi'e xorxes
_________________________________________________________________
Join the world�s largest e-mail service with MSN Hotmail.
http://www.hotmail.com
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sell a Home with Ease!
http://us.click.yahoo.com/SrPZMC/kTmEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
py...@aol.com
{lo ro broda cu brode} means something like "some broda is brode
and the cardinality of the set of broda is the cardinality of the
set of broda" -- I can't think of a better way of putting it,
unfortunately. The Lojban is not tautologous like my English version,
but it is as vacuously uninformative. Maybe "some broda is brode
and lo'i broda has a cardinality" might be better.
So in that case, {lo ro broda na brode} means "it is not the case
that both some broda is brode and the cardinality of the set of
broda is the cardinality of the set of broda (or, alternatively,
lo'i broda has cardinality". Since it obviously is the case that the cardinality
of lo'i broda is the cardinality of lo'i broda -- or that
lo'i broda has a cardinality -- the inevitable inference is that
it is not the case that some broda is brode.
>>
That is, of course, interpretation. What is said is that the cardinality is ro, just another number. So, if negation is going to affect that, it changes it to {me'iro} or some other version of {na'e ro}. Now, that may in the end give an equally contradictory component of some compound, if that is how it develops, and tha component may drop out to give just the basic negative, without any component about cardinality. But what then justifies adding back in a cardinality component, namely, the one jut dropped out. I suppose that it come back in because it is a tautology (why the negation dropped out). Well, at least this coheres so far, but I await a case where {lo PA broda} negated turns up as {lo na'e PA broda}. (As you point out later, this may be a long wait because no one ever uses {lo PA broda}.)
<<
Firstly, but less importantly, 'definite descriptions' -- closely
>>
You mean as a way to dodge the Russell cases. Probably, although I personally go with the explicit formats, the eight and ninety ways remaining to do descriptions -- all of which are also right.
<<
Second, the essence of le is specificity, with nonveridicality something
>>
Both of these theories of yours are interesting and need some mulling (my first instinct is to like it a lot). But I don't see that the fact that you have these theories (even if they turn out to be correct -- i.e., pick of the litter) requires us to use them to explain the present case, which isn't directly about that. But, of course, if specificity is a pre-illocution quantifier, then INNER, which is that quantifier, is presuppositional.
<<
Finally, I too don't recall it ever having been established that
>>
<<
So I'm back to square one then, understandingwise. Never mind, I'll
>>
<<
yet it's easy to come up with examples of
or will answer all questions successfully.
>>
<<
>>
<<
Of course, but you seemed to imply that there were many such cases.
I suspect that there aren't, and that if you did find some, the
>>
| Yahoo! Groups Sponsor | |
|
|
 |
py...@aol.com
<<
So called "inner quantifiers" should be called "inner cardinality indicators"
-- just as PA does not always function as a quantifier (e.g. in {li pa}), so
in {lo PA broda} it functions as an indicator of cardinality, not as a
quantifier.
Negation boundaries affect all inner cardinality indicators, but since ro
does not ascribe any cardinality to the set, it is vacuously affected.
>>
Just what does "affected" mean here? Obviously something different from quantifier DeMorgan, but what exactly. Where is a case of it (other than ro = ro) applied?
<<
the very nature of le-.
>>
Huh? I guess we mean totally different things by "presupposed". What is referred to may not be literally as it is described, but that is not presupposition, only the vagaries of human humor (when it occurs -- it is not very common so far in Lojban or any other language I can think of). Even if thiis is presupposition, I haven't anywhere seen it suggested before that (always remembering that with me memory is not a pramana) the INNER was also subjectively defined.
<<
#-- but, gee, my people (Partee, Bill Bright, various Lakoffs) and McCawley
Does McCawley deal with it in _Everything linguists always wanted to
know about logic_?
>>
Yes (I don't have my copy handy, so I can't give a citation). Right after he deals with the logicians' view (which he gets almost right -- he just doesn't note that the explicit quantifier in the quantified conditional form is importing), he remarks that the actual language situation is pretty clearly importing (though I don't think he uses that exppression) and gives some examples.
<<
quantification exists in Lojban -- whether {da poi broda cu brode} means
something different from {da ge broda gi brode}. I suspect you
would say that the former but not the latter entails {da broda}.
If I'm right about this, at least I can understand where you're coming
from, and will be in a position to think properly about the issues.
>>
<<
Everybody groups "every" and "each" together separate from "all", because
the former are distributive: "Every thing is", "Each (thing) is", "All (things) are".
>>
<<
If you can give me references on the importingness of "all" and "every" I
will go and look them up. I am skeptical about there being dialect differences,
but I shouldn't prejudge.
>>
The loc class is Zeno Vendler, "Each and Every, Any and All" Mind, v.71 (April 1962), pp 145 - 160. This is reprinted in a collection Vendler's papers, information on which I can't find (I'll sure be glad when I have used my "organized" library enought that I can start finding thing in it again). The results are summarized in a section "Any and All" in the Encyclopedia of Philosophy v1 pp 131-3 in the first edition. Vendler there cites a number of other sources, including somewhat linguistic ones like Klima "Negation in English" in Fodor and Katz.
<<
#whole situation is "P funny-and Q."
At a presyntactic/prelexical level I think it is "P and I-HEREBY-ASSERT Q"
>>
I would have put it pretty much tother way round, since I do not assert Q in this situation, only P and Q is there to allow me to do that assertion (or denial for that matter).
<<
Polishly "funny-and Q not P", not "not funny-and Q P", I take it you mean.
>>
Yes; I am careful about my binders when not writing Polish.
<<
#"not Q whether P" ("whether" = Lojban {u}).
The former would Griceanly imply the latter.
>>
It does so in any case, since "not Q" entails "Either not P or not Q", i.e. "not (P and Q). The second is better precisely because it contains more information.
<<
Can you cite some of the many cases of the presupposing version without
ro?
>>
Not in any normal sense (there is no standard location to cite from), but any case with a specific INNER that is not changed by negation passage will do.
Jorge Llambias
la pycyn cusku di'e
><<
> > I don't remember it being settled and decided (by whom?) the way
> > you want. For me {ro} is non-importing.
> >>
>Actually, on 15-03-02 you set forth (again) your system, acknowledging that
>it was aberrant,
The only post I can find from me on the subject that day is:
http://groups.yahoo.com/group/lojban/message/13795
which disagrees with your position rather than acknowkedging
mine as aberrant.
>and claiming for it a simplicity that it turned out not to
>have when actually applied or worked out theoretically.
Everyone can judge that for themselves. I have presented my
reasons for preferring non-importing {ro} on the wiki.
>That aside you
>acknowledged the correctness -- within Lojban of the importing system.
I may have acknowledged that your position is as consistent as any
other choice of import assignment. The way you present it makes
it look as if I had acknowledged it being better, something I do
not now and did not at that time consider to be true.
>Your
>{ro} is just {ro ni'u}, which is rarely useful and on those occasions is
>easily reached by falling back to standard Logic notation (your claim that
>ordinary {ro} can be reached in the same way from {ro ni'u} is true, but
>hardly an efficient suggestion.
In your system maybe. In mine {ro} is plain {ro} and yours
was {ro ma'u}, though the ma'u/ni'u idea never took flight.
><<
>So for you {ga broda ginai broda} can be false for selected broda?
>For me it's a tautology.
> >>
>I'm not sure that I understand this, but I suppose you mean {lo brode ga
>broda ginai brode} can be false.
You said that some claims and their negations could be false at
the same time, so for example, {ga ro pavyseljirna cu blabi ginai
ro pavyseljirna cu blabi} for me is a tautology, but for you it
is not. In other words, for me {ga <bridi> ginai <same bridi>} is
always a tautology. For you, for some <bridi>, it is not.
>Yes, it can, if there are no brode. But,
>note, {naku le brode ga broda ginai brode} is false as well, so
>tautological
>status is not affected -- the sentence is merely ill-formed at a low level.
For me, negation of a falsehood gives a truth.
mu'o mi'e xorxes
_________________________________________________________________
MSN Photos is the easiest way to share and print your photos:
http://photos.msn.com/support/worldwide.aspx
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sell a Home with Ease!
http://us.click.yahoo.com/SrPZMC/kTmEAA/ySSFAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/