Message-Id: <199601101939.VAA27393@xiron.pc.helsinki.fi>
Date:         Wed, 10 Jan 1996 10:29:43 -0800
Reply-To: "John E. Clifford" <pcliffje@CRL.COM>
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: "John E. Clifford" <pcliffje@CRL.COM>
Subject:      opaque
To: Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 14160
Lines: 275

i,n:
Sure, you can define lexical items that way, but I reckon what
you end up with is no longer a predicate - it's some other
kind of operator,  one whose arguments cannot be fully
interpreted (as referring to anything), but must
be treated in some other fashion.  (The traditional solution
for this is treat arguments as text, but I can conceive that
there might be su'o intermediate solution.)
Lojban gismu are "sold" as predicate words, and I've assumed
that the same applies to other selbri in general.  I could
accept other kinds of "lexical item" in the language, but I
would much prefer them to be clearly marked as such.
pc:
They are still predicates, because Lojban (and languages
generally) are type-theoretically flat.  The critters that go into
those slots are either perfectly normal but evaluated oddly or odd
things evaluated perfectly normally -- for the sorts of things they
are, take your pick.  Generally,  I have gone along with normal
things evaluated oddly, since that seems easiest to work with
ontologically (I don't much like transworld objects in my object
language), but the other makes some other things easier to do, at
first glance anyhow.  Taking the odd things as texts -- unless you
mean something novelly subtle by that -- is an idea that tends (on
the half-dozen occasions it has been suggested in 2500 years) to
die in the worst ways in about a week of scrutiny; in what
tradition has it hung on?

i,n:
As for pictures, I think that what is depicted is some sort of
abstraction (in the most general sense) of/from the object in
question, which might be appropriately represented by a Lojban
abstraction (NU), but again probably not involving {lo broda}.
pc:
_lo broda_ has to get in somewhere if it is a picture of  a(some)
broda(s).

i,n:
Given that these concepts are not well understood, I suspect
that the best we can do at the moment is to represent them
by something like {tu'a ce'u broda}, assuming that
a) {ce'u} is Cowan's proposed lambda pseudo-quantifer
b) {tu'a} provides a sufficiently closed context to
bind the {ce'u}.
({lo ka ce'u da broda} is more-or-less just \x:broda(x),
from which I believe we could in principle derive something
suitable, but I don't think Lojban has any mechanisms to
process lambda abstractions, which are in any case fairly new.)
pc:
_ce'u broda_ is just _broda_ and so _tu'a ce'u broda_ is ill-formed,
_tu'a_ requiring a sumti. The lambda paper sometimes talks as
though \xFx were a sumti but, in fact, it is whatever F is, here a
predicate). _lo ka ce'u da broda_ is not \x: x broda, but
^\x:x broda, not the predicate "is a broda" but the property of
being a broda, brodatude.  Neither of these (_broda_, however
disguised, nor brodatude) seem to have any obvious help in this
area (each has a place in another, often related, problem), but
events (or whatever you want to call 'em) do seem to help a bit.

i,n:
In another post you say something about making it
difficult to say simple things.  I contend that one of the
things that Lojban in particular highlights is that things
that we express "simply" in natlangs frequently aren't at
all simple.
pc:
No argument.  Still, if they can be said easily in most natural
languages -- and are often needed -- I think they should be easy to
say in Lojban, though with suitable warnings attached somehow.

&:
If we translate "any old ... whatsoever" into pred logic we appear to
get one of two patterns: (I) universal quantification with scope over
an irrealis element; (II) existential quantification within a subordinate
proposition. These may amount to the same thing but I'm not logician
enough to work it out.
pc:
When, as is usual, the irrealis is the antecedent of a conditional
and the scope of the subordinate existential is that same
antecedent, then the two are equivalent in standard logic --
provided the bound variable does not occur in the consequent as
well.

&:
I agree that {kae} can do the job of {cia}, so long as {kae} really
is defined just as "is true in some world (not necessarily this one)".
I see nothing else that can do the same job, except for really
long circumlocations like "da poi it-is-true-in-some-world fa
le duu kea broda" (instead of "lo kae/cia broda").
pc:
Whether we have stuck to the announced value of _ka'e_ I do not
know, but let's assume we have.  Even then, of course, your
reading for sumti will not help, since the referents of _lo ka'e
broda_ are the things *in this world* which may be brodas in this
or some other world.  But we will only be happy when we are
hunting them if we find them in a world where they are brodas
(which may not be this one), and there we would be just as happy
if we found things which are not in this world at all but are brodas
in that.  So your form doubly misses the mark.

&:
Is the "natural solution" an absence of rules, blind to the selbri,
for exporting to prenexes? I take it that instead you think that
such rules should be selbri-specific? I prefer the modern version
you decry.
pc:
Well, if we are going to claim to have all these words one of
whose main features is the opacity of a place and which are
unintelligible without that feature, we are stuck with something
like this.  I would be delighted if they were distinct so that a rule
could be written for them, since I dislike memorizing another list
(the nu-sumti list and the ka-sumti list are problems enough)
except that we all (should) know which the cases are already.
Perhaps we need a small list of these critters of various sorts and
make all the rest compounds with a distinctive ending (Sure, redo
morphology again! but that might be better than a slough of
useless gismu that do not mean what they seem to and cannot be
used to get to it -- or for anything else).

&:
Ultimately, Lojlan is speakable predicate logic, with various added
sugar.
pc:
Not recently (maybe not since 1980 or so) and even the sugar is
fermenting to vinegar pretty fast.

&:
I think we agree that semantically, all these things involve subordinate
bridi. The issue is just whether the syntax should mirror the semantics
(i.e. that the relevant predicates have sumti that are bridi), or whether
it should provide short cuts. It is unfair to criticize syntactic
forms without short cuts (i.e. no more or less complex than the semantics)
as "profoundly complex locutions".
pc:
Well, since you all seem to want to shortcut constantly -- omitting
the _ganai -gi_ with all your non-importing restricted _ro_s, for
example -- where the logic is against you and the natural language
is at best ambiguous, I see no reason for not insisting that we
shortcut when logic and natural language agree on the simpler
form.  Especially since the language already has forms which
appear to be the ones needed and have no other purpose.

&:
I feel similarly. I would prefer that the relevant gismu be redefined
so that, e.g. {djica} *must* have an x2 referring to a bridi, and, say,
{sisku} and {kalte} are redefined as the intended result of seeking
and hunting, so that one would then say {mi troci kuau/le duu/le nu
mi sisku lo cukta} "I seek a book".
pc:
What is the intended result of seeking and hunting? Finding and
TAKing, I guess.  But "find" is surely already a gismu, though I
can't find it, and TAKE is just a disjunction of a variety of
takings, the appropriate one of which would be substituted in each
case.  That is, these once useful old gismu would be superfluous
and replaced by complexes to the same effect as the gismu once
had.  So you have not yet found a new purpose for the displaced
forms.

&:
If you don't get your way, it's more likely due to people not understanding
you than to them not deferring to you. (But, undeferentially I opine
that I'm glad you didn't get your way on the matter you rue so.)
pc:
I have no evidence that anyone failed to understand my point
about the universal quantifier -- a matter that they could check in
a couple of minutes with a logic book -- yet in over two years you
are the first person to admit that I was right (if that is what you
did, rather than just stop arguing but going on as before).  The
present matter is somewhat different.  I cannot find when the
change was argued or, for that matter, when it was promulgated.  I
am less concerned about not getting my way than I am about not
having had a say at all in such a radical change.  I prefer open
discussion of all issues -- a part of the original Lojban plan, as I
understood it -- but, failing that, I want to be in the cabal (an
ancestor was, they say, the C) that does the deciding.

& (i,n inserted):
"hunt" means. So if {kalte} is a gismu, and it means "hunt", then it
means "x1 try for it to be the case that x1 'takes' x2".
> one whose arguments cannot be fully interpreted (as referring to
>  anything), but must be treated in some other fashion.
That doesn't follow. We seem to be agreed that x2 of kalte is quantified
in the same prenex as x1 ("Ex Ey, x try for it to be the case that x
'takes' y").
pc:
You can't have it both ways.  If x2 is subordinated to a "try" then
it need not be evaluated in the present universe.  If it is evaluated
only in the present universe than _kalte_ often does not mean
"hunt."  We can, of course, specify in various ways that we mean
for x2 to be evaluated in the present universe, but that takes an
extra effort (or, rather, does in natural langauges and ought to in
Lojban).

&:
Why do you reject a _duu/kuau_ clause? Those are, I understand, things
that are or aren't the case; that are or aren't true.
pc:
I don't reject them; I just don't understand what you mean by them
(as I said).  I am not sure what a predication is meant to be (it is
maybe time for another jargon dictionary to get issued; I lose
track of all the words based on _bridi_ for example, and this case
of what I suppose is a translation of one is not at all clearer).
_du'u_ is listed in NU but presented as a relation between a
predication and a sentence, which does not seem to be NU syntax.
As for kuau (illegal; kua'u or ku'a'u?), it is clearly experimental
and the only explanation I can find for it is a bare sketch (totally
unclear) in which it is contrasted with a misinterpretation (I think)
of another vaguely presented experimental proposal.  So, as I say,
I don't understand what you are suggesting.

&:
In my neck of the woods, situations comprise states, events, processes,
etc. They occupy time, and sometimes space. Storms and explosions are
situations. They contrast with states of affairs which are "the way the
world is (viewed panchronically)" - roughly speaking, I think a state
of affairs is an extensional object and a proposition a corresponding
intensional one. There is no temporal component to them.
I've hitherto thought that nu-thingies are situations, and duu-thingies
are propositions/not-necessarily-real-states-of-affairs. And everyone
has seemed to me to misuse/overuse nu, and underuse duu. Now, though,
it occurs to me that maybe nu-thingies are states of affairs - "way(s) the
panchronically viewed world is", and duu-thingies are their intensional
propositional counterparts.
I'm not trying to sell you my terminology. I'm trying to establish
whether we see the same distinctions.
pc:
Well, my terminology is in constant flux as I move from one
author to another, for none of the relevant fields have
standardized this part of terminology.  This seems to be due in
large measure to a lack of a standard semantic theory in any of
them and the disagreement among the dozen or so competing
theories about even what objects they need to countenance, let
alone what to call them.  When on my own, I tend to use "world"
for the largest possibility, a set of objects, assigned various
names,  and the extensions of all the predicates (roughly an
interpretation in the once-standard metalogical sense).  When
tense is not the issue, I tend to allow that worlds can have changes
in them over time, but, when doing tense logic, I call each
synchronic interpretation a world.  Situations are, for me, like
worlds except that they need not be complete (not every name and
predicate dealt with) and they need not (when it is relevant) be
consistent either.  States-of-affairs tend to be consistent situations,
and events small-scale ones (a couple of predicates and a few
objects, usually). All these are extensional (and abstract).  The
intensional side is even less clear.  Propositions are the intensions
of sentences and so what identifies the same sentence in different
worlds, but whether they are what is true or false (as opposed to,
say, holding or not) is less clear.  And the intensions of other
objects is even less clear; I suspect I use words lke "event" and
"situation" ambigously.  It would be nice to have a single theory
and then talk of these in terms of that, with, say,  lambdas and
cups and caps, but I don't see it happening soon.
        For now, what we need to do, before it starts to matter too
much, is agree on what constructions we have to deal with and
what their various peculiar characteristics are and then assign
some conventional name to them that we agree on.  What we
seem to be mainly talking about now is an event in my sense,
things having properties (or, usually, relations) where some of the
things need not be drawn from the present world and others
typically must be: the hunter yes, the hunted not.  Your problem
with these, I gather, is a difficulty in believing they exist when
they do not occur and that would make it nice if these were
intensional, since those are more believable when unfulfilled.
But there seem to be some problems with that in the literature,
though I cannot honestly say I understand them all (the fact that
one view thinks they're bugs and another that they're features may
be part of the problem).  To be continued, I'm sure.

&:
I beg your pardon. [No COI to say that with.]
pc:
COI suggests a conventional conversational move, on a par with the "Watch
out! I'm coming through!" and the "I am butting in here" versions of
"Pardon/excuse me."  My "Sure!" suggests that might be appropriate, though
a UI (u'u?) with a sarcasm marker (did that ever get in?) might be even
better.