[lojban] Re: na scope. Again.
Martin Bays
>
> --- Robin Lee Powell wrote:
>
> > Does {mi na nelci gi'e djica} mean:
>
> The three meanings can be unambiguously expressed thusly:
>
> (1) mi ge na nelci gi djica
> (2) mi ge na nelci gi na djica
> (3) mi na ge nelci gi djica
>
> The parser would suggest that {mi na nelci gi djica}
> corresponds to (1), but sometimes we don't pay any heed
> to what the parser says in these matters, especially
> when {na} is involved.
{mi naku nelci gi'e djica} would still be (3) though, right?
Martin
Robin Lee Powell
> Jorge Llamb?as wrote:
> >What it would change is simple negations like {mi e do na klama
> >le zarci}. Instead of meaning that either I don't go, or you
> >don't go, or both, it would mean that neither I nor you go.
>
> I would find that rather weird (lojbanically - it makes sense if
> you want to make Lojban closer to English), and also think it
> would defeat the point of using "na" rather than "na'e". If "na"
> doesn't mean "it is not the case that [brivla]", what does it mean
> that isn't covered by a different negative?
The goal would be to make "na" equivalent to a "na ku" in the same
place, because otherwise the scoping is really wierd and people
often get it wrong.
-Robin
--
http://www.digitalkingdom.org/~rlpowell/ *** http://www.lojban.org/
Reason #237 To Learn Lojban: "Homonyms: Their Grate!"
Proud Supporter of the Singularity Institute - http://singinst.org/
Jorge Llamb�as
--- Robin Lee Powell wrote:
> > > >
> > > > (1) mi ge na nelci gi djica
> > > > (2) mi ge na nelci gi na djica
> > > > (3) mi na ge nelci gi djica
>
> something?
(2) says I don't do either of them, (3) says I don't do both, i.e.
in (3) I may do one or none, but not both.
(2) is equivalent to:
(2') mi na ga nelci gi djica
and (3) is equivalent to:
(3') mi ga na nelci gi na djica
> > > {mi naku nelci gi'e djica} would still be (3) though, right?
> >
> > Right. And {mi nelci na gi'e djica} is unambiguously (1).
>
> And {mi nelci na gi'e nai djica} is unambiguously (2).
Yes.
mu'o mi'e xorxes
__________________________________
Do you Yahoo!?
Check out the new Yahoo! Front Page.
www.yahoo.com
robin
> --- Jorge Llamb�as wrote:
>
>>--- Robin Lee Powell wrote:
>>
>>
>>>
>>>>I am doing the section on NA, which should cover some of that. My
>>>>intention is to propose an interpretation where the scope of {na}
>>>>is restricted to what follows it.
>>>
>>>This would change the meaning of simple negations like "mi na klama
>>>le zarci", would it not? Good mabla luck.
>>
>>Not at all, how would that change? There are no quantifiers
>>or connectives to interact with {na} there.
>
>
> What it would change is simple negations like
> {mi e do na klama le zarci}. Instead of meaning that
> either I don't go, or you don't go, or both, it would
> mean that neither I nor you go.
>
I would find that rather weird (lojbanically - it makes sense if you
want to make Lojban closer to English), and also think it would defeat
the point of using "na" rather than "na'e". If "na" doesn't mean "it is
not the case that [brivla]", what does it mean that isn't covered by a
different negative?
--
"His youngest brother, Tendzin Choegyal, says one of the Dalai Lama's
greatest finds of recent years was super-glue -- second, in fact, only
to the more recent discovery of super-glue remover."
Robin Turner
IDMYO
Bilkent Universitesi
Ankara 06533
Turkey
Jorge Llamb�as
--- Robin Lee Powell wrote:
> On Wed, Nov 03, 2004 at 07:12:16AM -0800, Jorge Llamb?as wrote:
> > I am doing the section on NA, which should cover some of that. My
> > intention is to propose an interpretation where the scope of {na}
> > is restricted to what follows it.
>
> This would change the meaning of simple negations like "mi na klama
> le zarci", would it not? Good mabla luck.
Not at all, how would that change? There are no quantifiers
or connectives to interact with {na} there.
mu'o mi'e xorxes
Martin Bays
> --- Martin Bays wrote:
> > * Tuesday, 2004-11-02 at 13:08 -0800 - Jorge Llamb?as
> > <jjllamb...@yahoo.com.ar>:
> > > (1) mi ge na nelci gi djica
> > > (2) mi ge na nelci gi na djica
> > > (3) mi na ge nelci gi djica
> > >
> > {mi naku nelci gi'e djica} would still be (3) though, right?
>
> Right. And {mi nelci na gi'e djica} is unambiguously (1).
Yes, confusingly enough.
While we're on the subject... Is the BPFK or anyone else ba'o a ca a pu'o
working on the various problems with the interaction between negation,
unprenexed quantifiers and infix connectives, as raised e.g. by pycyn on the
wiki some months (years?) ago
(http://www.lojban.org/tiki/tiki-index.php?page=Logic+Language+Draft+3.1)?
Last I heard much of the semantics in even slightly complicated cases was
ill-defined. Is this still the case?
Martin
Jorge Llamb�as
--- Martin Bays wrote:
> While we're on the subject... Is the BPFK or anyone else ba'o a ca a pu'o
> working on the various problems with the interaction between negation,
> unprenexed quantifiers and infix connectives, as raised e.g. by pycyn on the
> wiki some months (years?) ago
> (http://www.lojban.org/tiki/tiki-index.php?page=Logic+Language+Draft+3.1)?
I am doing the section on NA, which should cover some of that. My
intention is to propose an interpretation where the scope of {na}
is restricted to what follows it.
> Last I heard much of the semantics in even slightly complicated cases was
> ill-defined. Is this still the case?
Yes. We always have the option of using the well defined
structure: i.e. explicit prenex quantification, negation
only in the prenex and forethought connectives. Hopefully we
will come up with consistent rules for how all other
structures can be expressed in terms of the well defined one.
Jorge Llamb�as
--- robin.tr wrote:
> Sorry, I meant "bridi". But it seems that if {mi e do na klama le zarci}
> means that neither I nor you go to the store, then what we are doing is
> modifying the brivla, i.e. we're saying "I and you do something other
> than go to the store" which is pretty much the same as saying {mi e do
> na'e klama le zarci}, isn't it?
{e}, like {na}, is a bridi operator.
We can apply first {e} and then {na}, or first {na} and then {e}.
Option 1:
mi e do na klama le zarci
= naku mi e do na klama le zarci
= naku ge mi klama le zarci gi do klama le zarci
= ga mi na klama le zarci gi do na klama le zarci
Option 2:
mi e do na klama le zarci
= ge mi na klama le zarci gi do na klama le zarci
In 1, {na} has scope over {e}. In 2, {e} has scope over {na}.
Option 2 does not say that you and I do something else to
the store. It just says that I don't go there and you don't
go there.
> Forgive me if I'm being dense here; I've always found negation hard to
> get my head round.
I think the problem is not so much negation but realizing that
connectives and quantifiers are bridi operators as much as
negation is, and the order in which they are applied with respect
to negation matters. This is separate from the issue of {na'e}
changing a brivla into a different brivla, which I don't dispute
at all.
robin
> --- robin wrote:
>
>>Jorge Llamb�as wrote:
>>
>>>What it would change is simple negations like
>>>{mi e do na klama le zarci}. Instead of meaning that
>>>either I don't go, or you don't go, or both, it would
>>>mean that neither I nor you go.
>>
>>I would find that rather weird (lojbanically - it makes sense if you
>>want to make Lojban closer to English),
>
>
> English whenever it makes sense, but the global scope of {na} makes
> no sense even from a strictly lojbanic point of view. It just
> doesn't fit well with everything else with scope, and it creates
> some problems that need more ad-hoc rules to solve them.
>
>>and also think it would defeat
>>the point of using "na" rather than "na'e".
>
>
> negates a bridi. I am in no way proposing to conflate them.
>
>
>>If "na" doesn't mean "it is
>>not the case that [brivla]", what does it mean that isn't covered by a
>>different negative?
>
>
>
Sorry, I meant "bridi". But it seems that if {mi e do na klama le zarci}
means that neither I nor you go to the store, then what we are doing is
modifying the brivla, i.e. we're saying "I and you do something other
than go to the store" which is pretty much the same as saying {mi e do
na'e klama le zarci}, isn't it?
Forgive me if I'm being dense here; I've always found negation hard to
get my head round.
Jorge Llamb�as
Right. And {mi nelci na gi'e djica} is unambiguously (1).
mu'o mi'e xorxes
Jorge Llamb�as
--- Jorge Llamb�as wrote:
>
> > On Wed, Nov 03, 2004 at 07:12:16AM -0800, Jorge Llamb?as wrote:
> > > I am doing the section on NA, which should cover some of that. My
> > > intention is to propose an interpretation where the scope of {na}
> > > is restricted to what follows it.
> >
> > This would change the meaning of simple negations like "mi na klama
> > le zarci", would it not? Good mabla luck.
>
> Not at all, how would that change? There are no quantifiers
> or connectives to interact with {na} there.
What it would change is simple negations like
{mi e do na klama le zarci}. Instead of meaning that
either I don't go, or you don't go, or both, it would
mean that neither I nor you go.
mu'o mi'e xorxes
Robin Lee Powell
> intention is to propose an interpretation where the scope of {na}
> is restricted to what follows it.
This would change the meaning of simple negations like "mi na klama
le zarci", would it not? Good mabla luck.
-Robin
Jorge Llamb�as
--- Robin Lee Powell wrote:
> Does {mi na nelci gi'e djica} mean:
>
> I do not enjoy, and I desire
>
> or
>
> I do not enjoy, and I do not desire
>
> ?
It is definitely not the second one, but it might
be:
I do not both enjoy and desire
The three meanings can be unambiguously expressed thusly:
(1) mi ge na nelci gi djica
(2) mi ge na nelci gi na djica
(3) mi na ge nelci gi djica
The parser would suggest that {mi na nelci gi djica}
corresponds to (1), but sometimes we don't pay any heed
to what the parser says in these matters, especially
when {na} is involved.
mu'o mi'e xorxes
Adam D. Lopresto
>>>> The three meanings can be unambiguously expressed thusly:
>>>>
>>>> (1) mi ge na nelci gi djica
>>>> (2) mi ge na nelci gi na djica
>>>> (3) mi na ge nelci gi djica
<snip>
> I don't see a semantic difference between (2) and (3); am I missing
> something?
{mi na ge nelci gi djica} is "It is not that case that I both like
and want," so (by DeMorgan's Law) {mi ga na nelci gi na djica}. mu'a, if {mi
ja'a nelci .ije mi na djica} then 2 is false, but 3 is true.
--
Adam Lopresto
http://cec.wustl.edu/~adam/
"Bother," said Pooh as he bounced off the Starfury.
Robin Lee Powell
I do not enjoy, and I desire
or
I do not enjoy, and I do not desire
?
-Robin
Jorge Llamb�as
--- robin wrote:
> Jorge Llamb�as wrote:
> > {mi e do na klama le zarci}. Instead of meaning that
> > either I don't go, or you don't go, or both, it would
> > mean that neither I nor you go.
>
> I would find that rather weird (lojbanically - it makes sense if you
> want to make Lojban closer to English),
That's certainly not the goal. I don't mind Lojban differing from
English whenever it makes sense, but the global scope of {na} makes
no sense even from a strictly lojbanic point of view. It just
doesn't fit well with everything else with scope, and it creates
some problems that need more ad-hoc rules to solve them.
> and also think it would defeat
> the point of using "na" rather than "na'e".
{na'e} modifies a brivla, it is quite different from {na}, which
negates a bridi. I am in no way proposing to conflate them.
> If "na" doesn't mean "it is
> not the case that [brivla]", what does it mean that isn't covered by a
> different negative?
{na} means "it is not the case that [bridi]". That doesn't change.
Quantifiers are also bridi modifiers. For example, {roda zo'u [bridi]}
means:
For every x, it is the case that [bridi].
What happens when you have both {roda} and {na} present? They can
modify a bridi in two different orders:
It is not the case that: For every x it is the case that: [bridi]
or:
For every x it is the case that: It is not the case that: [bridi]
Normally, when we have two operators that act on a bridi, the
first one takes the second one under its scope. This happens
for example with {roda} and {naku} or {roda} and {su'ode}
or {roda} and {ga... gi...}, etc. For some reason, {na} was
excepted from this rule, so that it supposedly always trumps
any other bridi operator when it appears right in front of
the selbri. In simple cases, this is awkward, but it can be
managed. In some more complex cases, for example with {na}
inside a logically connected bridi, the whole things just
breaks down.
In addition, the rule for other bridi operators that can also
appear in front of the selbri, such as {roroi}, was never very
clear. Do they follow the {na} rule, or the usual one of order
of appearance?
robin
> --- robin.tr wrote:
>
>>Sorry, I meant "bridi". But it seems that if {mi e do na klama le zarci}
>>means that neither I nor you go to the store, then what we are doing is
>>modifying the brivla, i.e. we're saying "I and you do something other
>>than go to the store" which is pretty much the same as saying {mi e do
>>na'e klama le zarci}, isn't it?
>
>
> {e}, like {na}, is a bridi operator.
> We can apply first {e} and then {na}, or first {na} and then {e}.
>
> Option 1:
>
> mi e do na klama le zarci
> = naku mi e do na klama le zarci
> = naku ge mi klama le zarci gi do klama le zarci
> = ga mi na klama le zarci gi do na klama le zarci
>
> Option 2:
>
> mi e do na klama le zarci
> = ge mi na klama le zarci gi do na klama le zarci
>
> In 1, {na} has scope over {e}. In 2, {e} has scope over {na}.
> Option 2 does not say that you and I do something else to
> the store. It just says that I don't go there and you don't
> go there.
>
OK, I see what you mean now.
Robin Lee Powell
>
> --- Martin Bays wrote:
> > * Tuesday, 2004-11-02 at 13:08 -0800 - Jorge Llamb?as
> > <jjllamb...@yahoo.com.ar>:
> > > --- Robin Lee Powell wrote:
> > >
> > > > Does {mi na nelci gi'e djica} mean:
> > >
> > > The three meanings can be unambiguously expressed thusly:
> > >
> > > (1) mi ge na nelci gi djica
> > > (2) mi ge na nelci gi na djica
> > > (3) mi na ge nelci gi djica
> > >
> > > The parser would suggest that {mi na nelci gi djica}
> > > corresponds to (1),
That is also my belief and, in fact, I think it *must* mean (1),
because if it means either (2) or (3), there is no way to say (1) in
afterthought!
I don't see a semantic difference between (2) and (3); am I missing
something?
> > > but sometimes we don't pay any heed to what the parser says in
> > > these matters, especially when {na} is involved.
> >
> > {mi naku nelci gi'e djica} would still be (3) though, right?
>
> Right. And {mi nelci na gi'e djica} is unambiguously (1).
And {mi nelci na gi'e nai djica} is unambiguously (2).
-Robin