[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
Nick:
> >But I don't understand your "lo'ie != lo'ei". Nobody has proposed
> >a "lo'ie", and the only person who thinks that "lo'e" or "loi'e"
> >= "lo'ei" is xorxes
>
> The latter was what I meant; but Jorge's acquiescence that his
> intensional is commensurate to the singularisation was critical in
> reaching the consensus
In his agreeing with it. He said he accepted the consensus without
trying to qualify it.
> I think what's emerged is (a) there is more than one way to
> singularise; (b) the lambda intensional generic is a singularisation;
> (c) it's not the only possible or useful singularisation; (d) it
> perhaps shouldn't be encompassed by lo'e as per "The Lion lives in
> Africa."
Yes. These have emerged.
> >The ability to use the singular generic for that is a byproduct
> >of its singularity. (Xorxes has been saying this all along, but it
> >has dawned on me only now. Yet more proof of the law of XIAR/JIAR.)
>
> We're having a lot of these "So that's why you said that!"
> revelations lately; I presume that's because we don't have the same
> backgrounds in logic. I mean, even with ten pages of formal semantics
> this morning, I can see why Jorge would conflate the Typical generic
> and the 'Any' Intensional generic; they are spoken of in the same
> breath in Formal semantics, and indeed are both expressed in the same
> way in English, as bare plurals. (The point of Carlson's paper is to
> argue that they are in fact the same thing --- but Carlson needed to
> account for English, so he would say that, right...)
Generics are expressed in English in a multitude of ways. 'Any'
readings are, I think, restricted to indefinites (_a(n)_ and
bare plurals (with coveral plural _a(n)_, IMO)).
Not to deny the relevance of differing backgrounds, tho.
> I dunno; I found myself yet again furious at Jorge, and I know that's
> counterproductive. So yes, Jorge can try expanding the meaning of a
> cmavo in direction X, because of perceived similarity Y; this may or
> may not mean the rest will follow
Yes, but if he thinks he's right, what else can he do? And if history
teaches us anything, it is that you'd be wise to bet that he is right.
> >Once there is only one broda, "I'm looking for a doctor" is like
> >saying "I'm looking for John" or "I'm looking for water", and the
> >opacity effect goes away
>
> .... until you start predicating extra stuff of the singularity,
> right? Because "any doctor" and "the average doctor" are certainly
> not the same thing
Agreed. I am now pretty much convinced that we need to distinguish
"Mr Doctor"/"the one doctor" from "the average doctor", and that
they should be expressed by different LE or LAhE.
> >> And that isn't what lo'e is defined as, dammit. lo'e is defined
> as 'typical'
> >> Even prototypical won't do, and And is being silly if he claims it
> >> does. Under no understanding of birds are penguins typical or even
> >> prototypical birds. Yet, if you're looking for a bird, and find a
> >> penguin, then you're satisfied
> >That's because the penguin IS Mr Bird! In Mr Bird's world, the penguin
> >in your garden and the sparrow in mine are the same individual
>
> And herein our misunderstanding. Because of course I didn't think of
> prototype in that way. And there is a long-running confusion as to
> what this Mr Bird is in Lojban; until recently, after all, we all
> assumed it was {loi cipni}. For all I know, it may in fact still be
> loi} and not {lo'e}
I know. At least we're coming out with a clearer understanding of
one another's positions, which goes quite a long way to working
out what viable consensual solutions might be.
> >You can see them as different individuals rather than being Mr Bird,
> >but then I can see Nick-in-2001 as a different individual from
> >Nick-in-2002, rather than both being Mr Nick
>
> Which is where Carlson's going. I don't think it's illegitimate for
> Lojban to have an ontological bias, like I said; it has a bias anyway
> by using Western logical machinery
I haven't got round to replying to that message -- BIIIG backlog.
But in brief, I don't accept that the logical machinery is biased
to the occident; I believe it is universal. I can live with biased
defaults -- e.g. bivalent logic -- so long as there is a marked
alternative (e.g. multivalent logic). And that seems to be Lojban's
general strategy. I don't like it if Lojban says "Look, we can see
the world in such and such a way, but we're not going to allow the
creation of a linguistic mechanism to express it, because most of
us don't see the world in that way".
> >You are construing "prototypical" to mean "a possibly imaginary
> >exemplar of all the properties category members typically have"
> >By "Prototype" I mean conceptualizing a category as an individual,
> >and that's the sense that {loi'e/lei'e} have. (I will cease to
> >ascribe it to {lo'e/le'e} because that former apparent-consensus
> >has lapsed into controversy.)
>
> OK. In which case, honest question: is the penguin encompassed by the
> prototype of 'bird' in prototype semantics? My impression was no, and
> that's why the prototype bird excludes penguins. In your take it does
> --- vacuously so --- and your prototype is probably indeed
> indistinguishable from lx.bird(x). But I didn't think that's how
> 'prototype' was used in semantics
Here's my attempt at an honest answer. The essence of Prototype Semantics
is that a category concept is not a list of membership criteria
(= Lojban ka) but rather an individual. In your mind's eye, the Bird
Prototype looks like a bird, not a membership application form.
The actual properties the Bird Prototype has are, on the whole, those
that are typical in our experience (though there are other factors
besides mere typicality). So a penguin definitely does not resemble
the Bird Prototype. However -- and here is the key point --, if we
decline to discriminate among instances of Bird (i.e. if we decline
to see them as not the same individual) then the world contains
just one individual that manifests Bird. I tend not to see Bird
that way; instead I see different individuals that all qualify
as instances of Bird by virtue of sufficient resemblance to Bird.
But I do tend to see Nick that way -- as just one individual that
manifests the Nick Prototype. Just as I can (and do) see as Nick Nick
on a day when he is not typical Nick, so I can see as Bird a penguin.
Now, in Lojban if I want to Nick as manifest by different instances,
then I can use {me la(i) Nick}. But what if I want to see Bird
as just one individual? Then I use {loi'e cipni}.
That, then, is an attempt to explain {loi'e}. Whether that meaning
or the "the average" meaning gets ascribed to {lo'e} is more a
political issue than a semantic one.
> >> So it now looks to me that, if we resolved "I'm looking for a doctor"
> >> in Lojban as {mi sisku leka ce'u mikce}, then we should resolve this
> >> intensional article with {ka}, as something like {le jai ka ce'u
> >> mikce} (or {se ka}, or whatever.) And if this discussion were not so
> >> utterly free form, I'd be able to find where someone had an objection
> >> to that
> >Sisku x2 (ka) is an abomination. I prefer djica x2 (nu) . But as I say
> >above, singularization annuls opacity, so copes with nitcu x2 ('object')
>
> sisku x2 may or may not be an abomination, but it is a fact of
> Lojban. That said, the "I want to talk to a doctor/any doctor" is a
> far far better illustration of this intensional doctor, because we
> don't need to get sidetracked by {buska}. (Sorry, Jorge, but I do
> think that backfired.)
Let's forget that buska stuff.
> Walk me through the last bit. Singularisation annuls opacity. So, an
> individual unsingularised population allows an ambiguity between "I
> need a/any doctor" and "there's this particular doctor I need" ---
> because in the former you're allowing arbitrary choice in a
> population on the spot (opacity), and in the latter the choice has
> been made way beforehand by the speaker (transparency).
There's a three way ambiguity, which I can illustrate with djica:
A. (OPAQUE)
mi djica loi nu da poi cukta zo'u mi ponse da
= mi djica loi nu mi ponse lo cukta
"I want to own a book, any book"
B. (TRANSPARENT)
da poi cukta zo'u mi djica loi nu mi ponse da
"There's a book that I want to own"
C. (SPECIFIC)
le cukta ku goi ko'a zo'u mi djica loi nu mi ponse ko'a
= mi djica loi nu mi ponse le cukta
"A certain book is such that I want to own it"
= "I want to own a certain book"
Because nitcu doesn't take an abstraction x2, we can't get reading
A with it.
However, the difference between readings A and B evaporates
when reference to a unique is involved:
mi djica loi nu mi ponse lai praidnpredjydis
= lai praidnprejydis goi ko'a zo'u mi djica loi nu me ponse ko'a
"I want to own Pride and Prejudice.
Hence {mi nitcu loi'e cukta} treats Book as a unique, and so
annuls the contrast between readings A and B.
> When you have
> a population of one --- and singularity does that --- there's no
> difference. So in "I need Mr Doctor", there's no distinction between
> choosing any one and having already chosen one; there's only one
> there. Right?
Right. Or more exactly, there's no distinction between me needing that
there be a doctor and there being a doctor that I need there to be.
> OK, I see why you'd be saying that. But still, the minute you allow
> anything extra to be predicated of that Mr Doctor, it becomes useless
> as an intensional article. So either: lo'e mikce is lx.mikce(x) *and
> never* The Average/Median Dr, or lo'e mikce is The Average/Median Dr
> and never lx.mikce(x). Yes?
Yes.
> In which case, we're back to a cmavo split. And we're also back to
> deciding (a) which sense is more useful, so that it gets the shorter
> cmavo lo'e; or alternatively (b) which sense is closer to the CLL
> prescription, so that it keeps lo'e
Yes. CVVV has the same syllable count as lo'e, so I don't think (a)
is terribly important.
> My opinion is still that (b) wins out,
I agree.
> that the Median is closer to the CLL's lion,
Either that or they're equally close -- I could argue it either way.
But I couldn't argue that Unique is closer.
BUT, if there is option (c):
(c) which sense is more useful, so that it (and only it) gets an
official cmavo
Then I am less happy to say "Median wins, by virtue of satisfying
(b)". This is because Median satisfies (b) only marginally more
than Unique, and because I think Unique satisfies (c) better, and
because I think that when it's a close call, it is more important
to have official cmavo with the 'best' meanings than with meanings
that are very very marginally more consistent with CLL.
As you will have realized, the outcome I am agitating for is
official gadri or LAhE for both Median and Unique.
> and that something *like* {se ka} or {jai ka} would
> be good for the intensional lo'e, since it would directly and
> mnemonically tie in with the {sisku x2} (a fact on the Lojban ground.)
No, {se ka} or {jai ka} or {poi'i} are of no help at all. It requires
a gadri or a LAhE. (Of course, a lujvo could express it -- pavycmima
or suchlike -- but that is beside the point, because lujvo could
replace other gadri too; they are needed not because they have
no paraphrase but because they are short.)
> I may yet yield and accept lo'e is not {le fadni be lo'i broda}, but
> this intensional. I haven't yet. And I persist on fundamentalist
> grounds 9and by the by, because I'm still unclear on how le'e fits in
> to this.) Now, that said, yes, I may well be overinterpreting or
> misconstruing CLL, and as you say, John hinted something in this
> direction. So did Bob, by saying "typical? sure, prototypical"
It comes down to whether lo'e actually encodes typicality, or whether
it is merely a generic, typicality being an epiphenomenon of genericity.
I think the choice is pretty arbitrary, if the only point at issue
is which meaning is attributed to lo'e, and not which meaning gets
an official gadri.
As for le'e, it's easy. For Median, it means "The average member of
le'i" (i.e. what you get from averaging all the members of le'i).
For Unique, it means "The one and only member of le'i". Peasy.
> But I don't think we understand the same thing by 'prototypical'.
> You're talking Mr Bird, which includes penguins. I'm talking cultural
> default 'bird', which excludes them. And of course, this is why I
> said {lo'e cinfo na xabju le gugdrxirana}, whereas Jorge would admit
> {lo'e cinfo cu xabju le sralo}
Right.
> That said, Jorge had {lo'e xunre labno} living in the wilds of the
> South, and {loi xunre labno} living in captivity. Which to me sound
> like my version of prototypical: "proper red wolves" live in the
> wilds of the South, "the mass of red wolves as we veridically know
> them" live in captivity. I think
Well I have Jorge living in Argentina, even though he lived in
Australia for four years. And I would say that "And is a to'e
angry person" is true, and "And is an angry person" is false,
even though I very occasionally do get angry.
IOW just because an instance of Broda is brode, one does not
necessarily find it appropriate to see Broda as brode.
> I don't know where this is going. "any doctor" and "the typical
> doctor" are clearly separate things. Logicians may well have found
> ways to conflate them; but I think we should keep them separated in
> Lojban anyway, to avoid confusion. So if we accept that, we have to
> pick whether fundamentalism or utility determines which keeps lo'e;
> which one the baseline hints at; and whether some sort of {jai ka}
> trick will work after all
Okay, as per my above comments.
> >Hang on, though. Until today, only Jorge said "lo'ei" is equivalent
> >to "lo'e/loi'e", *without disputing the meaning of "lo'e/loi'e"*
> >As of today, I finally see what he meant
>
> Please spell it out, just so I'm sure I get it.
See above.
> As long as there's
> any singularisation, you'll get *a* Mr Doctor, and if you want such
> an abstraction as Mr Doctor, you're clearly going to settle for any
> avatar of Mr Doctor, otherwise you'd have said something else. Is
> that it?
>
> Perhaps. Like I say, though, I don't know that Mr Doctor is lo'e
> mikce. In fact, I don't know yet that s/he isn't {loi mikce} after
> all. Massification suppresses individuation, so the choosing of
> opacity vs. transparency is sidestepped there too. Isn't it?
Yes, it is.
mi djica loi nu mi ponse re cukta
when applied to nitcu could be
mi nitcu loi'e cukta re mei
or
mi nitcu loi cukta re mei
but the latter is also consistent with a mental scenario in which
you need many different pair of books, while the loi'e version
isn't (because there is only one pair of books).
--And.