[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [lojban] {zo'e} as close-scope existentially quantified plural variable
Martin Bays, On 05/11/2011 17:22:
* Saturday, 2011-11-05 at 12:06 +0000 - And Rosta<and.rosta@gmail.com>:
I think the essential difference between us is indeed the
semantic-metaphysics. On one view, the universe comes with
a ready-made set of individuals, to which predicates apply;
propositions make claims about those individuals. On the other view,
the universe is one blob that can be split into uncountably infinitely
many subtypes, defined by differentiation criteria.
Here's a solution (v) then: have a couple of cmavo that mark these two
views, the Ready-Made and the Blobular. I really think that would
work.
Obviously you're a Ready-Madeist, while me and xorxes are
Blobularists. Traditional logic (i.e. what John Clifford calls
Traditional Western Logic) and formal semantics is Ready-Madeist.
Cognitive and natural-language-inspired approaches to semantics are
Blobularist.
One of the main strengths of lojban, and a crucial difference between it
and natural languages, is the ability it gives us to precisely specify
the scope of quantifiers in a sentence. The rules aren't wholly
specified, but that's a temporary problem.
Setting aside the unfinishedness of the rules, quantifier scope is unambiguous in Lojban, regardless of whether the sentence is uttered in Ready-Made or Blobular. That is, the logical form is unambiguous. If we inhabit a blobular universe, then the applicability of the logical form to the universe is ambiguous, precisely because logical forms can be applied only post-differentiationally.
The question then is how to use these powerful mechanisms in actual
communication. Because of the quantifier-switching phenomenon we've been
discussing, these mechanisms are useful only if the listener understands
which levels the speaker means to refer to - where I define 'level' as
whatever it is that we go up one of when we get from an AE sentence to
a witness for the corresponding EA sentence.
The obvious way to solve this problem (and the one I had been assuming
until xorlo came along) is along the lines of your "Ready-Made view"
- certain predicates isolate certain levels. e.g. if lions cinfo then
lionkind (if that's at a level above) doesn't, and nor do lion-stages
(if a lion is at a level above its stages). This doesn't mean we have to
decide once and for all what constitutes a lion, as "Ready-Made" might
suggest, just that we have to specify cinfo well enough that there can
be no ambiguity between levels.
To reiterate the point: allowing cinfo to be ambiguous between levels
is, by the definition of 'level', effectively equivalent to allowing the
logical structure of sentences which involve quantifying over cinfo to
be ambiguous. Since ambiguity in logical structure is a no-no in lojban,
so should be such effective ambiguity, and hence so should be such
level-crossing ambiguity in the meaning of cinfo.
I am (still!) surprised that this could be controversial.
Ready-Made is not a *solution* to the communication problem, because it is inapplicable to a blobular world. How do blobularists communicate about a blobular world?
A Ready-Madeist may be aghast at the Blobularist universe, but the Ready-Madeist can do nothing about that. The Blobularist universe is a fact, and banning linguistic representations of it is hardly acceptable.
So where does this leave Blobularism? I fear it leaves it needing to
find a way to specify the levels its carving the Blob to. Sorry.
First of all, if you want to insist that Blobularism has to specify levels, then why not also insist that Ready-Made must too? Otherwise you'd be requiring that almost all sentences expressible in Blobularist Lojban would be unexpressible in Ready-Madeist Lojban.
Second, why must the levels be specifiable? The idea is nonsensical to Blobularism, because even the number of levels is uncountably infinite. Ready-Madeists horrified by that can take refuge in Ready-Madeist Lojban.
If Barbie-like Beret is a malkind, then (B) is derivable
from (A) only if it is also the case that all frenchmen wear the same
beret; if they all wear different berets, you can't derive (B).
Hmm? Doesn't (A) imply that all french people wear Barbie-Beret?
Only metatruly. Under Blobularity, you first have to apply
differentiation criteria to the universe before you can make claims
about it.
Yes. And again: the problem is that we need to be able to *communicate*
what differentiation criteria are being used (at least to an extent
which rules out cross-level ambiguity), because otherwise we have
effective ambiguity in logical form.
That effective ambiguity is an inescapable fact of the Blobularist universe. That sort of predifferentiational disambiguation is impossible.
So it seems to me that either (A) doesn't entail (B) malkindfully or
that xorxesianism is not malkindful.
I don't see what you've done here.
I hadn't realized you were talking about metatruth rather than truth.
Truth would be assessed relative to a post-differentiational universe.
Metatruth is assessed relative to the set of all possible
post-differentiational universes: claims X and Y are
metatruth-conditionally equivalent if there is a predifferentiational
Blobular universe such that there are differentiation criteria that
yield from it a postdifferentiational universe of which X is true and
there are differentiation criteria that yield from it
a postdifferentiational universe of which Y is true.
Yes; and the issue is that, informal conventions and contextual hints
aside, two sentences which are metatruth-conditionally equivalent
communicate the same information.
Two such sentences *encode* different information. Two such *utterances* may communicate the same information, but firstly utterances do have a pragmatic context -- that's how language works, and a context-free language would be impossibly impoverished -- and secondly it's not an issue, it's a settled fact.
Sure, we know what the difference between one lion and two lions is.
But there are these cases where you can't tell the difference. And
I think that these cases in which the speaker can't tell the
difference should be generalized into a case where for whatever reason
the speaker doesn't tell the difference.
But do we really need to create a new entity to do that? In examples
like the "lion(s) in your garden every day", we can just give a vague
count - {su'o cinfo}, in that case.
Yes, but it looks like one lion, not like a group of one or more
lions.
Then {pa ju'o ru'e cinfo}?
That doesn't sound like a very Baysian solution...
If you think there's only one lion but you're not sure, you should just
say so.
Okay, but in the case under discussion, you've got something that looks like one lion but might be several. The speaker is sure it looks like one lion and sure that it mightn't be one. All the available diagnostics point to it being one lion, but not enough diagnostics are available. If you were to draw it or describe it, it would be like drawing or describing one lion. Still, I suppose Ready-Madeism would have to just use {su'o cinfo}.
--and.
--And.
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.