[analytic] Quine, Carnap, Ontology, and Modified Realism
Jan Dejnozka
Thank you very much for your thoughtful remarks.
"I subscribe neither to naive realism, nor to metaphysical realism.
Rather,
I hold to a position which has been called natural realism. It is in my
view just Quinean naturalism, but I am afraid it is not Quine as you have
interpreted him. I'll give you just two details which may suffice to
distinguish my position from what you understand as metaphysical
realism. First, natural realism is an empirical claim, not an a priori
one. It follows from physics, not metaphysics. Second, it does not
admit
of intensional objects, such as properties or relations."
You say you use "natural realism" to mean what Quine means by
"naturalism." I find it
confusing to proliferate terminology in this way, as it seems to blur the
distinction between his
naturalism and his robust realism. If I may quote from my "Quine: Whither
Empirical
Equivalence?,"
"Quine's thesis of naturalism is "the recognition that it is within
science itself, and not in some
prior philosophy, that reality is to be identified and described" (Quine
1981: 21). We "forgo the
transcendent or God's-eye view and only speak...from within our science"
(Quine 1994: 496). We
"recognize 'real' as itself a term within our scientific theory" (Quine
1994: 504)."
Naturalism is of ontological interest in that it concerns use of
the word "real." But it is not
a type of metaphysical realism. It is not a metaphysical thesis at all,
in the sense of asserting
anything about what there is, or what there may or may not be. It is an
epistemological and/or
linguistic thesis of how we can come to identify or describe relatity. It
says only that if we cannot
identify or describe something from within science, then we cannot
identify or describe it. (For
Quine, science is the totality of our theory, including logic and
mathematics.)
In contrast, Quine's robust realism is his express admission and
"unswerving" belief that
physical objects exist. This is obviously a type of realism. I take it to
be rather obviously a type
of metaphysical realism in conformance with Butchvarov's working
definition of "realism." To be a
metaphysical realism, a realism need not be specifically a substance
realism, any more than a
substance realism need be specifically materialist or physicalist. So I
need not even add that
Quine's physical objects are pretty close to being traditional substances
in many ways, or so I
argue in my book.
I see Quine's naturalism and his robust realism as not only
consistent, but as working
together in intimate cooperation in his philosophy. Namely, science in
his broad sense is
precisely the vehicle by which we can and do assert what we believe to to
be true about the world, and
refer to real things-- including real things that existed before
languages existed, for which the only
vehicle needed is a description to that effect (and I suspect that's what
Quine said; in your terms,
this would be a use as opposed to a mention of the description). I spell
this out and argue for this
intimate cooperation in my book, as quoted in at least one excerpt to
analytic (see "five excerpts
for the price of one"). This is as opposed to any suggestion that any
sort of radical metaphysical
relativity to scientific theory is implied by Quine's naturalism or is
thought by Quine to be
implied.
Therefore I'm perplexed by your statement that you are a "natural
realist" but not a
"metaphysical realist," since you appear to admit both what Quine would
call naturalism and
what he would call robust realism. and the latter is a type of
metaphysical realism. I hope the
confusion is only terminological, where metaphysical realism equates to
Moliere's "dormative property"
and your realism of physical objects equates to the capacity to put
people to sleep. But if you feel that
admitting physical objects is not a type of metaphysical realism, I think
the burden is on you to
explain why you would think that. Physical objects have been a standard,
even paradigmatic,
metaphysical category for many centuries.
I'm also perplexed by how you admit that Quine is a robust realist,
but appear to brush
that off as being somehow unimportant or irrelevant to his being a
metaphysical realist. Ignoring
his classes, that IS his metaphysical realism. What else could it be? Nor
is it a "moderate"
realism,
though it is a modified realism in my technical sense. What is moderate
about an unswerving
commitment to a world full of physical objects?
If you feel that naturalism is inconsistent with metaphysical
realism, and that you accept
the former but not the latter, I am also perplexed on how you could
accept "lumps before texts"
and be arguing that Bruce was wrong.
I would have thought that you would be pleased to have naturalism
and metaphysical
realism working together in your philosophy, much as they do in Quine's.
"I'll give you just two details which may suffice to distinguish my
position from what you
understand as metaphysical realism. First, natural realism is an
empirical claim, not an a priori
one. It follows from physics, not metaphysics. Second, it does not
admit of intensional objects,
such as properties or relations."
The two details you mention don't distinguish your position from
metaphysical realism at
all. As to the first detail, there is no need to argue for metaphysical
realism on an a priori basis.
This is especially so for science-based versions of metaphysical realism.
That's the whole point
of arguing that science can tell us, at least in some ways, how the world
really is. Even that
argument itself would not be a priori for Quine, for whom nothing is a
priori, and for whom everything
rational is at bottom synthetic, a posteriori, and empirical. Why would
you think that
metaphysical realists are somehow required to argue for what they think
is real on a priori grounds? As I've
mentioned before, modified realism is not defined in terms of the
arguments you give, but in
terms of the position you hold. In order to be a modified realist, you
don't have to give any arguments
at all! The same goes for metaphysical realism in general. Even a
rationalist such as Descartes
requires contingent, a posteriori sense-perception if we are to have
evidence that physical things
exist. In fact, I can't think of anyone who ever admitted physical
objects who argued for their
existence in a purely a priori manner and without appealing to a
posteriori sense-experience.
(Even some denials that there are physical objects are based in part on
empirical observations.
Berkeley is full of empirical observations about how things look to
different observers. Other
denials of physical objects, such as Leibniz's infinite divisibility
regress, seem a priori.) It must
be remembered that rationalism versus empiricism is epistemological, not
metaphysical. The
distinction between a priori and a posteriori is epistemological. It's
about how we can know
things. That's enough reason right there to classify naturalism as
epistemological. In contrast,
metaphysical realism, including Quine's robust realism of phsyical
objects, is metaphysical. The
epistemological distinction does not define what the metaphysical
positions are, but only how
they might be known to be true or false.
As to the second detail, why would you think that metaphysical
realists are somehow
required to admit intensional entities?
"This follows directly from causality being a relation between
events."
I thought you rejected all relations construed as intensional! Is
causality not an essentially
intensional relation? (I assume you do not wish to assay it as mere
regularity, thus conflating it
with accidental generalizations.) What do you mean by "intensional" when
you reject intensional
entities? Causality is certainly intensional in Quine's sense. That is,
it's not truth-functional. Nor
do I see how you could "construe" it as truth-functional.
"To exist in space-time, is just to exist and also to have a
spatio-temporal
extension. Properties and relations, construed intensionally, do not
exist, and so do not have spatio-temporal extension. (Construed
extensionally, as sets, they do exist, but do so unremarkably.)"
Are asymmetric relations construed extensionally ordered sets? Do you
admit ordered sets? If so,
how do you assay the ordering relations-- as further ordered sets needing
further ordering
relations?
"Rocks can exist even when there are no conceptualizations of rocks."
This is pretty much what Butchvarov and I mean by metaphysical realism,
which is very close to
the ordinary use of "real."
"My claim is that there is no dependence in either direction. Rocks can
exist even when there are no conceptualizations of rocks. However, I
also
hold that the sentence `There are rocks' cannot be true unless there are
conceptualizations of rocks. The apparent contradiction is resolved by
paying careful attention to a certain use/mention distinction.
Let's consider the example I offered previously. We three are discussing
the counterfactual possibility that we might not have evolved and that
the
earth might have been barren with only rocks and other lifeless
objects. Would these rocks be ontologically dependent on
conceptualizations of rocks? No, they would not because had the world
been
lifeless, there would have been rocks and no minds, hence no
conceptualizers, hence no conceptulizations. Mindful of what might have
been, however, we did evolve. And not only did the human species evolve,
but the fact is that the three of us are here on this earth thinking
about
rocks. Perhaps we are thinking about rocks that might have been rather
than about rocks that are, but nonetheless we are thinking about rocks.
If
we had not evolved, and we three had not lived to think about rocks,
hence
provide the conceptualizations of rocks, then there would not have been
any
sentence `There are rocks' to be true. And yet, it would have been true
that there are rocks. Here is the use/mention distinction to which I
want
to draw attention. When I say that the sentence `There are rocks' would
not have been true, I am *mentioning* a sentence that would not have been
(i.e., would not have existed), whereas when I say that it would have
been
true that there are rocks, I am *using* a sentence that is (i.e., that
does
exist). When I use a sentence that does exist, it does not much matter
that I might not have lived to use it, because I *do* live and I *do* use
it."
I'm no longer confident that your use-mention distinction is helpful.
There are two distinctions
which you rely on, the use-mention and the actual-counterfactual. I think
it is simplistic to
associate mention with actual and use with counterfactual. The truth is
that only an actual
sentence
can be actually used OR mentioned. And a counterfactual sentence-- a
sentence which could
exist-
- could be used OR mentioned. The only distinction actually doing any
work is the actual-
counterfactual. You are merely saying that actual sentences involve
actual conceptualizations,
which I think neither Bruce nor I would deny.What Bruce is in effect
denying is that we can
make
sense of, and be warranted in asserting, the counterfactual. You and I
are very definitely siding
against Bruce because we both accept the counterfactual. So you are not
taking any sort of third
position. You are taking my position!
>Fifth, compounding the difficulty is whether Rodrigo's expression "the
>object conceptualized" means just the rock, or whether it means "the
rock
>conceptualized" as opposed to the rock itself.
"I don't go in for Kantian notions of things-in-themselves as opposed to
things-as-they-seem. There are just things. I used the qualifier
`conceptualized' rhetorically to emphasize that I was not neglecting the
fact that all objects are conceptualized. I consider it redundant."
I had nothing specifically Kantian in mind. Please feel free substitute
"the rock" for "the rock
itself." The "itself" was merely emphatic. My question remains.
'The difference between Carnap and Quine, as I see it, is that for Quine
all
questions, including those of ontology, are internal, Carnap's
sense. Every question is answered with respect to a background theory
(in
place of Carnap's "linguistic framework"). But this suggests greater
differences than are there. Carnap may have insisted that traditional
questions of ontology were metaphysical nonsense. But he would also
readily admit to sensical internal questions of the form, "Are there
X's?". He did not recognize these questions as ontological because he
considered them internal to a linguistic framework. In my view their
difference amounts to a verbal difference in their use of the word
"ontology". Carnap accepts the term as describing a metaphysical
inquiry,
external to science and evidence. Quine doesn't admit to there being
*anything* outside of science and evidence, so he co-opts the term to
just
mean inquiry into what there is, whatever the nature of the inquiry.
Both
Carnap and Quine agree that there is only one kind of inquiry into what
there is, and that is empirical inquiry in search of truths of the form
"(Ex)(Fx)". Carnap just doesn't call this "ontology" and Quine does.'
This is a very plausible and sympathetic reading. I think there is much
truth in it. Unfortunately,
to paraphrase the famous saying, there's nothing like a sordid quote to
slay a beautiful
interpretation. My interpretation is based on quotations from Carnap and
Quine in my earlier
message detailing seven differences between Carnap and Quine and does
little, if anything, but
work out the logical implications of the quotations.
The difference between Quine and Carnap is not merely a verbal
difference in their
respective uses of the word "ontology." You are right that Carnap allows
"(Ex)Fx" to be
construed as either an internal or as an external question. But that
makes no difference to
my point. As you say yourself, the mere ambiguity of the label
"ontological" shows nothing. So
I'll just rephrase my point. Instead of saying that ontological
statements are nonsense for Carnap,
I'll just say more correctly that they are nonsense if construed
externally, and are in general (if
not always) analytic if construed internally. Either way, they will
still be in general devoid of factual
content for Carnap. And they will still have factual content and to that
extent be about the world for Quine.
Let's take a look, for example, at just one of the many consequences
I drew from that, namely
the consequence that Quine can allow genuine factual beliefs in
ontological statements. What
seems important to you is the admittedly very similar character of their
methodologies. Both say
they are open and liberal about allowing and exploring different theories
and frameworks. But that is where Carnap stops and Quine begins. Carnap
simply remains at that level, and so all his theories have a provisional,
merely methodological character. Quine is more complex. Precisely by
making all questions
internal and factual, he achieves tremendous stability at the
logico-mathematico-theoretical
phsyical internal core of factual theory. His core consists of statements
which are factual for him,
but which are not factual for Carnap, so Carnap cannot have a similar
factual core. For Carnap,
Quine's factual core is analytic if construed internally, and thus devoid
of factual content. For
Carnap, no private language argument can be given factual content if
construed internally, and he
apparently does not give one because it should lead to a factual
assertion that there is an external
world for us to communicate about independent of language, which invites
characterization of it
as external and nonsensical. But Quine's private language argument for
mind-independent
objective objects is and can be both internal and factual. Therefore, of
course, it is not immune
from revision and is not a priori on his own lights. But it is virtually
immune from revision and
for all practical purposes might as well be a priori. So he can achieve a
stable and unserving, not
to say overwhelmingly convincing, a posteriori and empirical commitment
to a robust reality of
physical objects. The provisional methodological character of his
framework becomes
comparatively a mere technicality, a fig leaf hiding the epistemological
feasibility of genuine
belief in a real world based on his factual private language argument. I
just don't see anything
like that epistemological feasibility in Carnap.
They are both pragmatists, and ironically Quine is more radically
extensive in his
pragmatism than Carnap, yet it is Quine who can and does genuinely
believe and affirm. This is
one of the consequences which I drew earlier of Quine's rejection of
Carnap's analytic-synthetic
distinction, now spelled out a little more fully in terms of the private
language argument. I could
have more directly and boringly appealed to Quine's naturalized
epistemology as factual.
Existential statements which for Carnap can only be either analytic
(when construed
internally) or literally nonsensical but covertly linguistic proposals
which are in any case neither
true nor false (when construed externally) become for Quine construable
only internally, but
always in uneliminable part factual and empirical questions-- and true or
false. This definitely
includes statements like "There are numbers," and arguably should include
"There are physical
objects," though "There exists an Eiffel Tower" would not be analytically
true for Carnap when
construed internally. So the word "ontological" is not merely used
differently by Carnap and
Quine. What's important is that the different use reflects a different
understanding of what
is and can and should be preserved of traditional ontology within
naturalism. For Carnap,
traditional ontology is literally nonsense and covertly linguistic
proposals, and either way it is
neither true nor false. So nothing can be preserved of its alleged
factual content. For Quine, on
his naturalism, even traditional ontology has an uneliminable factual and
empirical content, can
be intelligibly said to be true or false, and there can even be evidence
for or against its truth or
falsehood. This is paradoxical in light of Quine's verificationism, which
is itself ultimately in part
factual and empirical in character. But it is merely paradoxical insofar
as the factual plausibility
of verificationism outweighs the factual plausibility of the excesses of
traditional ontology. (This
may not be Quine's express argument, but it is structurally his
argument.) In contrast, Carnap's
verificationism is as a priori and devoid of factual content as the
traditional ontology it condemns
as nonsensical when construed externally.
"A question: Did Quine ever write about ontological hierarchies, with
some
things more real than others? It strikes me as unlike him. What is
there? Everything. He was very matter of fact about the simplicity of
being."
"That logico-mathematical objects are real only by analogy does ring a
bell. Maybe I do remember reading this. However, again I ask whether
this
reification by analogy results in a lesser status of being. I don't
think
so. Rather, he was just detailing the process by which, starting with
hard-line nominalism, one yet must accept some abstract entities."
Once again, there's nothing like a sordid quotation to slay a
beautiful interpretation.
I note in my book that he thinks hierarchically only for a 21 year
period, 1960-81, and that
he expressly abandons such thinking from 1981 on. It is just as much his
repudiation of his earlier
hierarchic thinking as his earlier thinking itself that provides textual
support. When you say it is
unlike him, it is indeed unlike post-81 Quine. I give five quotations of
what seem to be
arguments for hierarchy in my book, p. 267. I give three quotations and a
(fourth) cite of what seems to be
his repudiation of his previous hierarchism, pp. 267-68. The gist of the
repudiation is a
repudiation of the robustness of physical objects. That is, in 1987 he is
not leveling the field by
making classes or sets robust, but by anemicizing physical objects. I
think there were strong
earlier
tendencies in that direction, but they overwhelmed the conflicting
considerations favoring
hierarchy only sometime after 1981. If you look up my nine citations in
context, I think you'll
find
my interpretation very natural.
There are two things to warn against. First, it's a big mistake to
think that Quine does not
admit degrees of reality because he uses the same existential quantifier
for every object and
applies the doctrine that to be is to be the value of a variable to every
object equally. That is
totally irrelevant. I have already shown that Frege makes physical
objects more real than classes,
and that has nothing to do with the fact that he uses the same
existential quantifier for all objects
and applies his 'no entity without identity' doctrine of being to every
object equally. The very idea
is a categorial confusion. The concepts of existence and reality are not
even on the same logical
level. The former is a second-level concept predicated of first-level
concepts, and cannot be
significantly predicated of objects precisely because all objects exist.
The latter is a first-level
concept predicated of objects precisely because it does apply only to
certain paradigmatic objects.
Perhaps it is no coincidence that Quine studied and revered Frege so
highly, but if I am right,
exactly the same Fregean type-levels would apply to Quine's concepts of
existence and reality.
The second big mistake is to think that just because you can use
arbitrary proxy functions
to cook up indefinitely many empirically equivalent ontologies, reducing
physical objects to
numbers or vice versa, that Quine has no robust realism of physical
objects in which physical
objects "are the things, first and foremost."Quine himself throws a wet
blanket on the idea that
that is relevant at all in _Theory and Things_. You need to look at first
page 9, "It is a wrong
question...," then p. 21, "But I also expressed..." which comes AFTER all
the proxy function.
swticharounds of translation, and then above all at the immediately
following paragraph:
"The semantical considerations that seemed to undermine all this [robust
realism of physical
objects] were concerned not with reality but with analyzing methodology
and evidence. They
belong not to ontology but to the methodology of ontology, and thus to
epistemology. These
consdiderations showed that I could indeed turn my back on external
things and ride the proxy
functions to something strange and different without doing violence to
any evidence [and he had
just presented a whole series of those including rewriting physical
object talk as pure set talk]. But
all ascription of REALITY must come rather from within one's theory of
the world; it is
incoherent otherwise [precisely because of all these empirically
equivalent and mainly absurd
translation manuals]...."
Then at the bottom of the same page 21, he even repeats the point:
"We are free to switch, without doing violence to any evidence.... But it
is a confusion to suppose
that we can stand aloof and recognize all the alternative ontologies as
true in their several ways, all
the envisaged worlds as REAL. It is a confusion of TRUTH with evidential
support. Truth is
IMMANENT, and there is no higher...."
So, exactly as I said in my book, my home language is precisely my
vehicle of reference to the
real world for Quine, and so long as we do not rock the boat, we can and
do talk meaningfully of
what is real, and truth and reference can be ours. See my book, pp. 262,
263, 264.
This is not to mention that the very term "robust realism" is
repeatedly applied to physical
objects only. Classes or sets are never called robustly real. What more
evidence could you
possibly want? Why, the fellow keeps calling physical objects his robust
realism! He always calls
physical objects robustly real, and he never calls anything else robustly
real! He's handing his
two-tiered reality view to you on a silver platter! The silver platter is
filled with physical objects
and has a large placard taped to it announcing THIS IS MY ROBUST
REALISM.! He's shouting
it from the rooftops! How could anyone possibly miss it? What's more, the
fellow is flatly telling
you that the indefinitely many systematic inversions of ontology through
translation manuals or
proxy functions, winding up even in replacing physical objects with a
pure set ontology, has
absolutely nothing to do with it!
Nor does he merely assert a two-tiered realism. He supports it with
reasons. My citations
indicate his reasons.
And yet many people have missed it. This seems to me to be another
widespread case of selective reading of the great analysts. To my mind,
reading Quine's giving of primacy of reality to physical objects out of
Quine's ontology as if it were never there is on a par with reading
Wittgenstein's many realist aspects out of _Investigations_ as if they
were
never there.
Quine does, I believe, eventually repudiate his two-tiered ontology
in favor of
a single tier, and I report the four structural reasons he had for doing
so on pages 287-88 of my
book. So I have considerable sympathy with those who missed the earlier
two-tiers view. They may be reading
Quine selectively, but they are certainly reading Quine. And I cannot say
the secondary literature has
emphasized Quine's degrees of reality. Again, this is on a par with those
who miss the realist aspects of _Investigations_-- it's understandable.
"I don't doubt that he expressed unswerving commitment. It did not
swerve,
after all. I just meant that his commitment was not a priori."
I never said or thought it was anything else! Nor does the
epistemological basis of his
commitment matter to the fact of his robust metaphysical realism one way
or the other.
> Also note that in assuming abstract objects over and above the
> physical objects, Quine goes on to make it clear that abstract objects
> less paradigmatically real than physical objects. _Theories and
> Things_, also in my book. ADD CITE. This is an ontological statement
> about what is in the world that would be unintelligible and
> nonsensical to Carnap.
"I am not convinced that it would be nonsense. As for physical objects
being paradigmatically real, yes, they set the standard by example, I
would
say. But is this talk of ontological foundations, or just guides by
which
to judge our theories?"
It would be nonsense as an external question. I think you are right that
it would not be nonsense
as an internal question. I am not clear on your distinction between
ontological foundations and
guides by which to judge our theories.
"To clarify, I meant not that his commitments should be doubted. Carnap
never called himself a phenomenalist, only a methodological
phenomenalist. Never a solipsist, only a methodological solipsist. It
is
critical that we understand what this qualifier was meant to express. My
interpretation is that, a priori, he acted epistemologically as a
phenomenalist or solipsist would, but that a posteriori, his commitments
would change. Implicitly, instead of a phenomenalism, he would conclude
with physicalism, and instead of solipsism, he would believe in other
minds. The reason he didn't outright adopt these positions
metaphysically
was basically his opposition to metaphysics. He thought it was
meaningless
to adopt any position a priori. He only adopted, provisionally,
phenomenalism and solipsism as epistemological starting points if you
will,
because they seemed ideally free of metaphysical baggage and because he
had
to start somewhere."
Your observation of a disconnect between Carnap's methodological
commitments and his actual
beliefs seems not only to support my interpretation, but to suggest that
our interpretations are not
so far apart as you might think.
"Quine's innovation was to start in the middle, thereby obviating the
necessity for "methodological" commitments. Quine's "method", if you
will,
was just as physicalist and realist as his conclusion, but not because he
had the a priori metaphysical commitments that Carnap lacked, rather it
was
because he didn't start with the a priori, preferring to deny its
availability altogether."
What Quine means by starting in the middle is starting by positing
medium-sized dry goods, _Word and Object_, p. 4. This is precisely the
initial (and revisable) commitment his methodology requires. I agree with
the second sentence, except that I would say instead that Carnap's a
priori commitment was to an anti-metaphysical verificationism intimately
related to an a priori analytic-synthetic distinction.
"Physical objects are vital to Quine's philosophical views and they are
not similarly
vital to Carnap's philosophical views."
My point exactly.
"Quine's commitment to physical objects is *identical* to Carnap's. It
is
> > motivated and defended by the empirical science of physics, no more
> > and no less. Quine differs only in not allowing for a philosophy
> > prior to empirically motivated ontological commitments. For Quine,
> > this is not a recognition that an ontological commitment to
substance
> > is prior to anything else, and in particular prior to basic physics
> > and psychology. Rather, it is a denial that there is any prior
first
> > philosophy at all."
>
> ##### This is a misunderstanding. Carnap does not regard "his
> philosophical views as separable from and prior to his scientific
> views." Rather, Carnap repudiates ontological views, as asserted by
> ontological statements purporting to be about the world, as literally
> nonsensical.
"You're right, but I was referring to philosophical views other than
ontology. I was referring to his views on language, meaning, evidence,
truth, etc., even his views of being, but only in the abstract, since he
didn't make any a priori (read: metaphysical) ontogical commitments."
My point about ontology remains. I also disagree on the "differs only,"
in that I don't think you
are seeing all the systematic consequences of the difference that I do.
> First, Quine expressly affirms that he is a robust realist, and
> goes into abundant detail on the point. You won't find any such
> affirmations linguistically permissible to Carnap, I think for obvious
> reasons.
"Perhaps not so obvious. For Carnap, it goes without saying that we are
all
realists when it comes finally to scientific theory. Since he is at such
pains to show that metaphysical claims are meaningless, he would
understandably hesitate to claim robust realism, lest he be mistaken for
a
metaphysical realist."
But then there is the disconnect between methodology and actual (or
feasible) belief which you
mention. More precisely, please see my discussion of Carnap's distinction
between empirical
realism (roughly equating to ordinary belief) and metaphysical realism
(roughly equating to
"methodology") below.
"Quine, on the other hand, handily disavows traditional metaphysics as
easily as he holds to "robust realism". Only, it must be noticed that
for
Quine, unlike for Carnap's targetted metaphysicians, this realism is
supported by physics."
I never disagreed. I say all four great analysts disavow traditional
metaphysics on page 1 of my
book.
> Sixth, Carnap makes "There are physical objects," which is on its
> face a synthetic a posteriori statement, into an analytic a priori
> truth-- and true depending only on our choice of "linguistic
> framework"! I see that as preposterous. Quine, in contrast, correctly
> makes "There are physical objects" an ontological statement to which
> empirical evidence is relevant. What I see as preposterous in Quine is
> his making even "There are numbers" an ontological statement to whose
> truth-value empirical evidence is relevant.
I consider it no accident that you find preposterousness in both Carnap
and
Quine along such similar lines. I, by contrast, find nothing
disagreeable
or preposterous about either claim. I only distance myself from Carnap's
due to my not subscribing to the a/s distinction which is what
underwrites
the whole internal/external distinction to begin with. To the degree,
however, that it makes sense to speak of truths, analytic in degree, I
also
agree with Carnap.
I'm confused. Do you mean that insofar as you accept truths as analytic
to a degree, you accept
"There are physical objects" as analytic to a degree? What does it mean
for a statement to be
analytic to a degree? Carnaps rejects that there can be degrees of
analyticity. It's a big part of his
argument in "Empiricism, Semantics, and Ontology" that analytic internal
questions are either
true or false, and can admit of no degree, while the same questions
construed externally are neither
true nor false, but are only pragmatically useful to some degree. As I
see it, Quine rejects all talk
of degrees of analyticity along with yes-or-no analyticity. He admits
instead degrees of linguistic
content and converse degrees of empirical content.
As to the preposterousness claim, let me embroider on that. If
"There are physical
objects" is analytic a priori, then we should be able to give a formal a
priori proof of the
existence of physical objects. It would be an ontological argument not
for God, but for physical
objects. It should be able to rely only on logical truths and putting
synonyms for synonyms, or
for Carnap, meaning postulates describing linguistic conventions of the
framework. To me, it
sound like pulling an ontological rabbit out of a logico-linguistic hat.
But of course, it is only an
internal rabbit, not an external rabbit. In both traditional and
pre-philosophical terms, it's a sham,
and that suggests that Carnap has strayed far from common sense in a way
that traditional
ontology has not.
"I agree, citing the very same essay, that the a/s distinction is
underwriting their differences. I only disagree that Carnap is
anti-realist."
As long as there is a disconnect between what you call his realism and
his philosophical views,
he is philosophically a radical relativist. And again, I don't think
you're seeing all the consequences
of Carnap's acceptance and Quine's rejection of the analytic-synthetic
distinction. To make it as
visible as I can, for Quine "God exists," "Numbers exist," "Classes
exist," and "Relations exist"
are all internal only, all have some factual, empirical content, and all
are a posteriori. They all
have a minimal empirical relationship to the world, and are all
assertions about the world to some
minimal degree. Furthermore, for Quine these statements I just made about
"God exists" and
"Numbers exist" are themselves a posteriori and empirical, and have some
minimal factual
content and are about the world. For Carnap, if the existential
statements are construed externally, they
are cognitive nonsense if taken literally, and they are covertly
linguistic proposals-- all of this being
an a priori and analytically true philosophical position. If they are
construed internally, they are
analytically true and empty of all factual and empirical content (I
assume that we are concerned
only with the ontological argument for God), and this is itself an a
priori and analytically true
philosophical position which is devoid of factual content and not about
the world. I find this
ample reason for finding Quine a metaphysical realist and Carnap not.
And let's look at what you call Carnap's realism. I assume you're
referring to his
distinction between empirical realism and metaphysical realism in the
_Aufbau_. This is
basically just the same as the distinction between existential statements
construed internally and
construed externally. Empirical realism is what the pre-philosophical
scientist would deem real
as opposed to an illusion, hallucination, dream, mirage, decoy, or fake.
It seems related to
Austin's theory of the ordinary use of "real" as well as at least
somewhat to Kant's conception of
empirical reality. Metaphysical realism is as opposed to other
metaphysical positions such as
idealism or solipsism. Thus calling Carnap an empirical realist is
certainly correct, but adds
nothing to the discussion. We've already discussed the distinction using
other terms. To say that
Quine collapses the internal-external distinction so as to make all
existential statements internal
yet also genuinely factual and about the world, is just to say that Quine
collapses the empirical
realism-metaphysical realism distinction so as to make all existential
statements empirically
realist, yet also genuinely factual and about the world. So basically the
_Aufbau_ is at one with
"Empiricism, Semantics, and Ontology." There is basically one position
with two terminologies,
and Quine has a fundamental disagreement with that position. I'm
referring to section 170ff. of
the _Aufbau_. If Carnap's approach to metaphysical realism in the
_Aufbau_ is somewhat muted
and more cautious than his forthright, definitive theory in "Empiricism,
Semantics, and Ontology,"
it makes no significant difference to the present discussion.
Carnap says in _Aufbau_, sect. 170:
'The only concept of reality which occurs in the empirical sciences we
shall call *the concept of
empirical reality*. It is this concept which distinguishes a
geographically determined mountain
from a legendary or a dreamed mountain, and an experienced emotion from a
simulated one. The
question as to what is real, when it is formulated with the aid of
constructable concepts, can only
be concerned with this empirical reality; it alone can be posed and
treated within the constructional
system; hence, we speak here of the 'constructional' or 'empirical'
problem of reality, in contrast
to the 'metaphysical' problem of reality which will be discussed in the
sequal (sect. 175ff.), when
we shall be concerned with a different, a 'metaphysical', concept of
reality. This latter concept
occurs only in traditional philosophy, not in the empirical sciences.'
And in sect. 175:
'We have determined what constructional (empirically ascertainable)
conditions must be fulfilled
in order for an object to be called real in the customary usage of the
empirical sciences. In
addition to this "constructional" or "empirical"problem of reality, the
question may arise whether
or not we must ascribe reality in a special sense to these empirically
real objects.... [M]ost
commonly, [this special sense] is characterized as *independence from the
cognizing
consciousness*. Thus we have to differentiate two meanings of the word
"reality".... calling the
one "empirical reality" and the other "metaphysical reality".... The
second concept of reality
indicates the point where the schools of realism, idealism, and
phenomenalism part company....'
It should be abundantly clear that the ONLY sense in which Carnap admits
he is a realist is
empirical realism, since it is the ONLY sense which he admits as
occurring internally and hence
meaningfully within his linguistic frameworks. It should be abundantly
clear that metaphysical
realism is what he is rejecting as literally nonsensical in "Empiricism,
Semantics, and Ontology,"
and what he considers to be covertly nothing more than disguised
pragmatic linguistic proposals
which are neither true nor false.
>>[Rodrigo:] I am discussing ontological dependence. Since you spell it
out above, I
>>know that you mean the same by "logical dependence", though I will
>>continue to use "ontological" instead. Just to be sure, one object is
>>ontologically dependent on another if the first cannot exist unless the
>>second also exists.
>
>***** [Jan:] Yes, we mean the same thing where objects are concerned.
But for me
>there can also be logical dependences among statements which describe
>logical fictions. But counterfactually, if the fictions were objects,
then
>those logical dependences would also be ontological.
"You've lost me, but we can return to it when relevant."
I just mean that for the 1914-18 Russell, "If there are elephants, then
there are mammals; there
are elephants; therefore there are mammals," would express a logical
connection, but if he is right
that elephants and mammals are logical fictions, then there would be no
ontological connection
because elephants and mammals would be ontologically nothing. Similarly
for "All elephants are
essentially mammals," if that helps. It's an unimportant point for our
discussion.
"I would love to read such an essay. If you or anyone can remember which
it
is, please let me know."
It wasn't a whole essay but something much briefer. I had suspected it
was just one sentence or two at most in _Word and Object_ or possibly
_The Philosophy of W. V. Quine_, but I didn't want to send anyone on a
wild goose chase. The closest texts I could find on a casual search of
_Word and Object_ were p. 5, "We cannot strip....saving the data," and
chapter 1, last two paragraphs, but I'm not convinced either of these is
what I was remembering. So I really am not sure where I might have read
it. It may well be not exactly as I remember it, and my memory is not
very exact on the point.
" I also am not sure how Quine would put these
things, but I do know a little about how he treats subjunctive
conditionals. In "Three Grades of Modal Involvement", Quine discusses
three different forms of modal discourse."
I use this essay in my second book, _Bertrand Russell on Modality and
Logical Relevance_, to
argue that all seven of the modal logics I impute to Russell are second
grade at the very least, and
a few things are third grade. My discussion of that starts on p. 81, and
my conclusion is on pp.
95-96, of that book.
"I don't remember ever reading the expression "immanent
within our referential system/theory/apparatus". What do you mean by
it?"
"Immanently" is Quine's term. To quote my "Quine: Whither Empirical
Equivalence?"
Immanentism is the thesis that all truths and all facts are immanent in
scientific theory.
This closely parallels the theory of naturalism. Immanentism is given to
us by Tarski (Quine
1987:
316).
> May I also speculate that when you say that all objects are
> conceptualized, and also that objects and their conceptualizations are
> mutually independent, that for you all objects merely happen to be
> conceptualized (or referentialized) in the actual world, since we make
> sweeping assertions about all objects, but that they need not be
> conceptualized in order to exist?
I'm not sure. I don't think so.
Would it help if I mention that I take the nature of concepts to be in
the
domain of cognitive science? I don't consider them intensional objects
at
all. They are, roughly speaking, whatever it is that a mind acquires
upon
learning to recognize objects as so-and-so and/or think about objects as
so-and-so.
It helps, especially if you can assign them a spatial location inside our
bodies. Then they will be
wholly and really distinct from rocks, insofar as there are no rocks in
our bodies. I assume you're
speaking casually about minds and not admitting them as a category.
"The two go hand in hand. If intensional objects afforded the simplest
theory then, naturalism would admit them. However, it would be
impossible
to even formulate such a theory about intensional objects, let alone a
simple one, without identity conditions for them."
Why couldn't the simplest theory be that they are given as having
intrinsic identity conditions?
"I'm not sure I understand the problem. How does ignorance of membership
preclude anything? I hereby assert that there will be as many apples and
oranges on your desk tomorrow. I may be wrong, of course, since I speak
from ignorance. But where otherwise have I faulted?"
Well, you wouldn't know what sets you were talking about. You would have
no way of telling
what the set of apples is. You've never enumerated the membership, and
that's the only way you
can identify an extensional set. In your terms, it would be impossible to
formulate a theory about
the equinumerosity of apples and oranges without providing identity
conditions for the sets in
question.
"Some sets exist in space-time and others don't. I would say that the
only
sets that exist in space-time are sets of points in space-time, call them
space-time regions. I would exclude all other sets, including sets of
space-time regions."
Since to exist in space-time is to exist and have an extension, either
your points have extensions
or they don't exist in space-time. How can they exist and have space-time
locations, yet not exist
in space-time? Can a spatial point exist and have a spatial location, yet
not exist in space? Do
points have infinitesimal extensions?
Can regions which exist in space-time be identical with sets of points
which do not exist in space-time? Can a set exist in space-time even if
none of its members does? (Most people would say the opposite, that a set
of
apples cannot exist in space-time even though apples do!) I'm well aware
of the fallacies of
composition and division, and arguably you might tough it out as strictly
not a problem, but it
seems unduly paradoxical. And you seem to be running directly into Zeno's
paradoxes as
opposed to using modern theory of limits. But I admire that, since
doesn't theory of limits duck
the compositional problem for those who do compose (lines and) regions
out of points?
Best wishes,
Jan
http://www.members.tripod.com/~Jan_Dejnozka/index.html includes vitae,
abstracts
of publications, philosophy book announcements, book corrections,
retrospects on
books, unpublished papers, brief lists of philosophical, musical,
literary, artistic,
and cinematic favorites, and a family photo and art gallery.
________________________________________________________________
GET INTERNET ACCESS FROM JUNO!
Juno offers FREE or PREMIUM Internet access for less!
Join Juno today! For your FREE software, visit:
http://dl.www.juno.com/get/tagj.
------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/_/_/980828514/
---------------------------------------------------------------------_->
Rodrigo Vanegas
>Thank you very much for your thoughtful remarks.
And thank you for your challenging responses.
I am tempted to reply to just these two paragraphs. Because I enjoy the
point by point discussion, I will reply to all of your post anyway, but I
do want to say that this here is surely the fulcrum of our difference. By
"natural realism" I meant what you describe as "naturalism", an empirical
use of the word "real", not metaphysical realism. Perhaps I called it a
kind of realism because I disagree that naturalism does not "assert
anything about what there is". On the contrary, I think that that is all
that it does assert. That is all that can be done with the word
"real". It is not merely a linguistic or epistemological thesis. (I am
disagreeing with Carnap here.)
As for Quine's "robust realism", now that you have made it clear you mean
to intepret him as having asserted a realism above and beyond his
naturalism, I think I would like to see some textual evidence, for it is
not the Quine that I know and support. His "unswerving belief that
physical objects" exist is insufficient since it follows directly from his
naturalism. Let me be clear: I do not doubt that he asserted belief in an
external world, in physical objects, in classes, and more. I only doubt
that these assertions were more than just naturalism.
>I'm also perplexed by how you admit that Quine is a robust realist, but
>appear to brush that off as being somehow unimportant or irrelevant to his
>being a metaphysical realist. Ignoring his classes, that IS his
>metaphysical realism. What else could it be? Nor is it a "moderate"
>realism, though it is a modified realism in my technical sense. What is
>moderate about an unswerving commitment to a world full of physical objects?
What is moderate is that the commitment flows from mere naturalism and does
not extend beyond it.
>If you feel that naturalism is inconsistent with metaphysical realism, and
>that you accept the former but not the latter, I am also perplexed on how
>you could accept "lumps before texts" and be arguing that Bruce was wrong.
Because I believe that naturalism does not stop short of making assertions
about what there is, and Bruce is defending a mildly anti-realist line.
>I would have thought that you would be pleased to have naturalism and
>metaphysical realism working together in your philosophy, much as they do
>in Quine's.
No, I regard opposition to metaphysics to be one of the hallmarks of
Quine's philosophy.
>>"I'll give you just two details which may suffice to distinguish my
>>position from what you understand as metaphysical realism. First,
>>natural realism is an empirical claim, not an a priori one. It follows
>>from physics, not metaphysics. Second, it does not admit of intensional
>>objects, such as properties or relations."
>
>The two details you mention don't distinguish your position from
>metaphysical realism at all. As to the first detail, there is no need to
>argue for metaphysical realism on an a priori basis. This is especially so
>for science-based versions of metaphysical realism. That's the whole point
>of arguing that science can tell us, at least in some ways, how the world
>really is. Even that argument itself would not be a priori for Quine, for
>whom nothing is a priori, and for whom everything rational is at bottom
>synthetic, a posteriori, and empirical. Why would you think that
>metaphysical realists are somehow required to argue for what they think is
>real on a priori grounds?
Just a long-standing impression I've had. The idea of a posteriori
metaphysics is rather new to me. It still isn't rolling off the toungue...
>As to the second detail, why would you think that metaphysical realists
>are somehow required to admit intensional entities?
That's what I thought (and still do think) metaphysical realism, as
compared to realism simpliciter, is all about. Maybe you can tell me
otherwise.
What is the difference between realism and metaphysical realism? I usually
take "realism" to mean "mind-independent being", and "metaphysical realism"
to mean "mind-independent being of universals". If you have different
interpretations, I'd like to hear them.
>>"This follows directly from causality being a relation between events."
>
>I thought you rejected all relations construed as intensional! Is
>causality not an essentially intensional relation? (I assume you do not
>wish to assay it as mere regularity, thus conflating it with accidental
>generalizations.) What do you mean by "intensional" when you reject
>intensional entities? Causality is certainly intensional in Quine's sense.
>That is, it's not truth-functional. Nor do I see how you could "construe"
>it as truth-functional.
Is causality essentially intensional? Well, it usually is, but I don't
know if it is essentially so. I think one can just consider laws to be
those regularities that are deepest to the core of a given theory, without
recourse to intensions. This alone, admittedly vaguely, would support the
counterfactual contexts expected of causal reasoning.
>>"To exist in space-time, is just to exist and also to have a
>>spatio-temporal extension. Properties and relations, construed
>>intensionally, do not exist, and so do not have spatio-temporal
>>extension. (Construed extensionally, as sets, they do exist, but do so
>>unremarkably.)"
>
>Are asymmetric relations construed extensionally ordered sets? Do you
>admit ordered sets? If so, how do you assay the ordering relations-- as
>further ordered sets needing further ordering relations?
Yes, that's right.
>>"Rocks can exist even when there are no conceptualizations of rocks."
>
>This is pretty much what Butchvarov and I mean by metaphysical realism,
>which is very close to the ordinary use of "real."
In my view, there is still an ambiguity. Do rocks exist independently of
conceptualizations because their being rests on the independently existing
universal "rock"? Or are they only ontologically independent, still
dependending on an existing conceptualization in the world in which the
rock's existence is asserted? The latter is the view I am recently
arguing. The former is my understanding of metaphysical realism.
On this point we may not be disagreeing. At this very moment, it's not
clear to me whether there is a need for both distinctions. It may come
back to me later.
>> >Fifth, compounding the difficulty is whether Rodrigo's expression "the
>> >object conceptualized" means just the rock, or whether it means "the rock
>> >conceptualized" as opposed to the rock itself.
>>
>>"I don't go in for Kantian notions of things-in-themselves as opposed to
>>things-as-they-seem. There are just things. I used the qualifier
>>`conceptualized' rhetorically to emphasize that I was not neglecting the
>>fact that all objects are conceptualized. I consider it redundant."
>
>I had nothing specifically Kantian in mind. Please feel free substitute
>"the rock" for "the rock itself." The "itself" was merely emphatic. My
>question remains.
It sounded Kantian to me because you contrasted "the rock conceptualized"
with "the rock itself". The phenomenal and the noumenal, respectively. I
don't acknowledge an opposition so it's difficult to answer your question.
>>'The difference between Carnap and Quine, as I see it, is that for Quine
>>all questions, including those of ontology, are internal, in Carnap's
Okay, we are closer. I would very much like to see the sordid quotations
you are referring to. If you formerly posted them to Analytic, please tell
me the message number and I will go look.
It turns out that I have some textual support for my view as well. I
should thank you for pointing me in the direction of the source of this
quote, since I wouldn't have thought to look there without your
mention. It is the second paragraph of "On Carnap's Views on Ontology".
>"When I inquire into the ontological commitments of a given doctrine or
>body of theory, I am merely asking what, according to that theory, there
>is. I might say in passing, though it is no substantial point of
>disagreement, that Carnap does not much like my terminology here. Now if
>he had a better use for this fine old word `ontology,' I should be
>inclined to cast about for another word for my own meaning. But the fact
>is, I believe, that he disapproves of my giving meaning to a word which
>belongs to traditional metaphysics and should therefore be
>meaningless. Now my ethics of terminology demand, on occasion, the
>avoidance of a word for given purposes when the word has been pre-empted
>in a prior meaning; meaningless words, however, are precisely the words
>which I feel freest to specify meanings for. But actually my adoption of
>the word `ontology' for the purpose described is not as arbitrary as I
>make it sound. Though no champion of traditional metaphysics, I suspect
>that the sense in which I use this crusty old word has been nuclear to its
>usage all along."
Quine, not "a champion of traditional metaphysics", thought there was "no
substantial point of disagreement" on this subject, except that Carnap did
not "like [Quine's] terminology". I think my reading is not only plausible
and sympathetic, it is also Quine's own reading.
It is true that Carnap's analytic and internal ontological statements are
devoid of factual content, but so what? Why is semantic content not good
enough? The standard should be only that he believed things exist
independently of mind. The standard was met.
I suppose it comes to matter of taste, then. If you are willing to say
that Quine's "unswerving commitment to physical objects" is a priori "for
all practical purposes" even as you admit that strictly speaking it was not
a priori, then I cannot see why you won't grant Carnap's unswerving
commitment to science as also a priori "for all practical purposes" since
among his a posteriori commitments physics was front and center.
>>"A question: Did Quine ever write about ontological hierarchies, with
>>some things more real than others? It strikes me as unlike him. What is
>>there? Everything. He was very matter of fact about the simplicity of being."
>>
>>"That logico-mathematical objects are real only by analogy does ring a
>>bell. Maybe I do remember reading this. However, again I ask whether
>>this reification by analogy results in a lesser status of being. I don't
>>think so. Rather, he was just detailing the process by which, starting
>>with hard-line nominalism, one yet must accept some abstract entities."
>
>Once again, there's nothing like a sordid quotation to slay a beautiful
>interpretation.
>
>I note in my book that he thinks hierarchically only for a 21 year period,
>1960-81, and that he expressly abandons such thinking from 1981 on. It is
>just as much his repudiation of his earlier hierarchic thinking as his
>earlier thinking itself that provides textual support. When you say it is
>unlike him, it is indeed unlike post-81 Quine. I give five quotations of
>what seem to be arguments for hierarchy in my book, p. 267. I give three
>quotations and a (fourth) cite of what seems to be his repudiation of his
>previous hierarchism, pp. 267-68. The gist of the repudiation is a
>repudiation of the robustness of physical objects. That is, in 1987 he is
>not leveling the field by making classes or sets robust, but by
>anemicizing physical objects. I think there were strong earlier tendencies
>in that direction, but they overwhelmed the conflicting considerations
>favoring hierarchy only sometime after 1981. If you look up my nine
>citations in context, I think you'll find my interpretation very natural.
I will look for them when I turn to the book. Thank you.
>There are two things to warn against. First, it's a big mistake to think
>that Quine does not admit degrees of reality because he uses the same
>existential quantifier for every object and applies the doctrine that to
>be is to be the value of a variable to every object equally. That is
>totally irrelevant. I have already shown that Frege makes physical objects
>more real than classes, and that has nothing to do with the fact that he
>uses the same existential quantifier for all objects and applies his 'no
>entity without identity' doctrine of being to every object equally. The
>very idea is a categorial confusion. The concepts of existence and reality
>are not even on the same logical level. The former is a second-level
>concept predicated of first-level concepts, and cannot be significantly
>predicated of objects precisely because all objects exist. The latter is a
>first-level concept predicated of objects precisely because it does apply
>only to certain paradigmatic objects. Perhaps it is no coincidence that
>Quine studied and revered Frege so highly, but if I am right, exactly the
>same Fregean type-levels would apply to Quine's concepts of existence and
>reality.
This is all news to me. Since you are only saying that Quine sided with
Frege, maybe you can point me toward wherever Frege made the distinction
between existence and reality. If you could show where Quine also accepts
this difference that would be great too. If you would prefer that I just
get it from your book, then it will have to wait until then.
Whatever the truth about Frege and Quine may be, I don't myself acknowledge
any difference.
>The second big mistake is to think that just because you can use arbitrary
>proxy functions to cook up indefinitely many empirically equivalent
>ontologies, reducing physical objects to numbers or vice versa, that Quine
>has no robust realism of physical objects in which physical objects "are
>the things, first and foremost."
I agree that Quine is a robust realist. I disagree that this is a form of
metaphysical realism. "First and foremost" along what ordering? I see no
evidence that he means that physical objects are more real.
>"Quine himself throws a wet blanket on the idea that that is relevant at
>all in _Theory and Things_. You need to look at first page 9, "It is a
>wrong question...," then p. 21, "But I also expressed..." which comes
>AFTER all the proxy function. swticharounds of translation, and then above
>all at the immediately following paragraph:
>
>>"The semantical considerations that seemed to undermine all this [robust
>>realism of physical objects] were concerned not with reality but with
>>analyzing methodology and evidence. They belong not to ontology but to
>>the methodology of ontology, and thus to epistemology. These
>>consdiderations showed that I could indeed turn my back on external
>>things and ride the proxy functions to something strange and different
>>without doing violence to any evidence [and he had just presented a whole
>>series of those including rewriting physical object talk as pure set
>>talk]. But all ascription of REALITY must come rather from within one's
>>theory of the world; it is incoherent otherwise [precisely because of all
>>these empirically equivalent and mainly absurd translation manuals]...."
>
>Then at the bottom of the same page 21, he even repeats the point:
>
>>"We are free to switch, without doing violence to any evidence.... But it
>>is a confusion to suppose that we can stand aloof and recognize all the
>>alternative ontologies as true in their several ways, all the envisaged
>>worlds as REAL. It is a confusion of TRUTH with evidential support. Truth
>>is IMMANENT, and there is no higher...."
As I read him, he was just restating Carnap's objection to external
ontological statements. They are true immanently, as opposed to
transcendentally. He doesn't repudiate these evidence-supported
alternative theories because they are absurd. To wit, it isn't the
theories themselves that seem absurd but their manuals by which we
translate from them into our home language.
There is yet no evidence either of metaphysical evidence nor of a
distinction between reality and existence.
>This is not to mention that the very term "robust realism" is repeatedly
>applied to physical objects only. Classes or sets are never called
>robustly real. What more evidence could you possibly want? Why, the fellow
>keeps calling physical objects his robust realism! He always calls
>physical objects robustly real, and he never calls anything else robustly
>real! He's handing his two-tiered reality view to you on a silver platter!
>The silver platter is filled with physical objects and has a large placard
>taped to it announcing THIS IS MY ROBUST REALISM.! He's shouting it from
>the rooftops! How could anyone possibly miss it? What's more, the fellow
>is flatly telling you that the indefinitely many systematic inversions of
>ontology through translation manuals or proxy functions, winding up even
>in replacing physical objects with a pure set ontology, has absolutely
>nothing to do with it!
I find myself interpreting him differently. Though I can offer no textual
support, I believe that the robustness of his robust realism is just his
relative commitment to physical objects in the face of possible theory
change. In other words, were he faced with "recalcitrant evidence" that
forced a change somewhere in his theory, he would be relatively disinclined
to disturb his ontological commitment to physical objects.
>And yet many people have missed it. This seems to me to be another
>widespread case of selective reading of the great analysts. To my mind,
>reading Quine's giving of primacy of reality to physical objects out of
>Quine's ontology as if it were never there is on a par with reading
>Wittgenstein's many realist aspects out of _Investigations_ as if they
>were never there.
Selective reading or different interpretations of full readings? Let the
impartial third party decide.
>> >Also note that in assuming abstract objects over and above the
>> >physical objects, Quine goes on to make it clear that abstract objects
>> >less paradigmatically real than physical objects. _Theories and
>> >Things_, also in my book. ADD CITE. This is an ontological statement
>> >about what is in the world that would be unintelligible and
>> >nonsensical to Carnap.
>>
>>"I am not convinced that it would be nonsense. As for physical objects
>>being paradigmatically real, yes, they set the standard by example, I
>>would say. But is this talk of ontological foundations, or just guides
>>by which to judge our theories?"
>
>It would be nonsense as an external question. I think you are right that
>it would not be nonsense as an internal question. I am not clear on your
>distinction between ontological foundations and guides by which to judge
>our theories.
By "ontological foundations" I meant "ontological categories at the first
tier", and by "guides by which wo judge our theories" I meant "high
relative commitment in the face of recalcitrant evidence". I better
explain myself in these other terms elsewhere in this post.
>>"To clarify, I meant not that his commitments should be doubted. Carnap
>>never called himself a phenomenalist, only a methodological
>>phenomenalist. Never a solipsist, only a methodological solipsist. It is
>>critical that we understand what this qualifier was meant to express. My
>>interpretation is that, a priori, he acted epistemologically as a
>>phenomenalist or solipsist would, but that a posteriori, his commitments
>>would change. Implicitly, instead of a phenomenalism, he would conclude
>>with physicalism, and instead of solipsism, he would believe in other
>>minds. The reason he didn't outright adopt these positions
>>metaphysically was basically his opposition to metaphysics. He thought
>>it was meaningless to adopt any position a priori. He only adopted,
>>provisionally, phenomenalism and solipsism as epistemological starting
>>points if you will, because they seemed ideally free of metaphysical
>>baggage and because he had to start somewhere."
>
>Your observation of a disconnect between Carnap's methodological
>commitments and his actual beliefs seems not only to support my
>interpretation, but to suggest that our interpretations are not so far
>apart as you might think.
I don't think they're not connected, just not identical. His
methodological commitments lead to his beliefs. It just seems to me that
if we are going to call him an anti-realist, then it is his beliefs that
should count for or against him, not his methodological commitments.
>>"Quine's innovation was to start in the middle, thereby obviating the
>>necessity for "methodological" commitments. Quine's "method", if you
>>will, was just as physicalist and realist as his conclusion, but not
>>because he had the a priori metaphysical commitments that Carnap lacked,
>>rather it was because he didn't start with the a priori, preferring to
>>deny its availability altogether."
>
>What Quine means by starting in the middle is starting by positing
>medium-sized dry goods, _Word and Object_, p. 4. This is precisely the
>initial (and revisable) commitment his methodology requires. I agree with
>the second sentence, except that I would say instead that Carnap's a
>priori commitment was to an anti-metaphysical verificationism intimately
>related to an a priori analytic-synthetic distinction.
Yes, that's right. We don't disagree about where they began. We only
disagree about where they ended. You think Carnap was stuck with
relativistic anti-realism, and I think that he ended, alongside Quine, with
physicalism.
>>"Quine, on the other hand, handily disavows traditional metaphysics as
>>easily as he holds to "robust realism". Only, it must be noticed that
>>for Quine, unlike for Carnap's targetted metaphysicians, this realism is
>>supported by physics."
>
>I never disagreed. I say all four great analysts disavow traditional
>metaphysics on page 1 of my book.
Not everything I write to you is intended to disagree with you. :-)
It is these degrees of linguistic and empirical content that I am referring
to. Following Quine's example, since "analytic" is now an obselete term, I
freely adopt it to new usage. And again, like Quine, I feel that the
adoption is not so arbitrary, that, in fact, "the sense in which I use this
crusty old word has been nuclear to its usage all along".
>As to the preposterousness claim, let me embroider on that. If "There are
>physical objects" is analytic a priori, then we should be able to give a
>formal a priori proof of the existence of physical objects.
Why so? Rather than a proof, it requires only the adoption of a linguistic
framework.
>To me, it sound like pulling an ontological rabbit out of a
>logico-linguistic hat. But of course, it is only an internal rabbit, not
>an external rabbit. In both traditional and pre-philosophical terms, it's
>a sham, and that suggests that Carnap has strayed far from common sense in
>a way that traditional ontology has not.
My view is that he has only strayed from traditional ontology and that this
was a good thing, in every respect.
>>"I agree, citing the very same essay, that the a/s distinction is
>>underwriting their differences. I only disagree that Carnap is anti-realist."
>
>As long as there is a disconnect between what you call his realism and his
>philosophical views, he is philosophically a radical relativist. And
>again, I don't think you're seeing all the consequences of Carnap's
>acceptance and Quine's rejection of the analytic-synthetic distinction. To
>make it as visible as I can, for Quine "God exists," "Numbers exist,"
>"Classes exist," and "Relations exist" are all internal only, all have
>some factual, empirical content, and all are a posteriori. They all have a
>minimal empirical relationship to the world, and are all assertions about
>the world to some minimal degree. Furthermore, for Quine these statements
>I just made about "God exists" and "Numbers exist" are themselves a
>posteriori and empirical, and have some minimal factual content and are
>about the world. For Carnap, if the existential statements are construed
>externally, they are cognitive nonsense if taken literally, and they are
>covertly linguistic proposals-- all of this being an a priori and
>analytically true philosophical position. If they are construed
>internally, they are analytically true and empty of all factual and
>empirical content (I assume that we are concerned only with the
>ontological argument for God), and this is itself an a priori and
>analytically true philosophical position which is devoid of factual
>content and not about the world. I find this ample reason for finding
>Quine a metaphysical realist and Carnap not.
Again, I find it evidence only that Quine was a realist, not a metaphysical
realist. And although it is not evidence that Carnap was a realist, he is
realist enough on account of his treatment of internal ontological questions.
>And let's look at what you call Carnap's realism. I assume you're
>referring to his distinction between empirical realism and metaphysical
>realism in the _Aufbau_. This is basically just the same as the
>distinction between existential statements construed internally and
>construed externally.
Yes, that's what I'm referring to.
>Empirical realism is what the pre-philosophical scientist would deem real
>as opposed to an illusion, hallucination, dream, mirage, decoy, or fake.
>It seems related to Austin's theory of the ordinary use of "real" as well
>as at least somewhat to Kant's conception of empirical reality.
>Metaphysical realism is as opposed to other metaphysical positions such as
>idealism or solipsism. Thus calling Carnap an empirical realist is
>certainly correct, but adds nothing to the discussion.
Well, that altogether depends on how you want to frame the discussion. We
are discussing not only metaphysical realism, after all, but realism as
opposed to relativistic anti-realism. The central questions are (1) Was
Quine a metaphysical realist as well as an empirical realist? and (2) Was
Carnap a relativistic anti-realist?
As I see it, Carnap's empirical realism, which was not substantially
different from Quine's, was enough to oppose him to any relativistic
anti-realism. To idealism and solipsism, Quine and Carnap were both
opposed in substantially different ways. Quine thought them demonstrably
false and Carnap thought the doctrines meaningless.
>We've already discussed the distinction using other terms. To say that
>Quine collapses the internal-external distinction so as to make all
>existential statements internal yet also genuinely factual and about the
>world, is just to say that Quine collapses the empirical
>realism-metaphysical realism distinction so as to make all existential
>statements empirically realist, yet also genuinely factual and about the
>world.
I'm not so sure that he did collapse this distinction. Rather he just did
away with the metaphysical side. The model here is not the collapsing of
the analytic/synthetic distinction, but the doing away with the a
priori. Just as Quine isn't said to have collapsed the distinction between
the a priori and the a posteriori, I think it should not be said that he
collapsed the distinction between metaphysical and empirical
realism. Rather, he repudiated the former and made "ontology" a matter of
the latter.
>Carnap says in _Aufbau_, sect. 170:
Hey, how did you get your hands on a copy of the _Aufbau_? I have tried
finding a copy several times, with no success.
>It should be abundantly clear that the ONLY sense in which Carnap admits
>he is a realist is empirical realism, since it is the ONLY sense which he
>admits as occurring internally and hence meaningfully within his
>linguistic frameworks. It should be abundantly clear that metaphysical
>realism is what he is rejecting as literally nonsensical in "Empiricism,
>Semantics, and Ontology," and what he considers to be covertly nothing
>more than disguised pragmatic linguistic proposals which are neither true
>nor false.
Yes, of course. No argument on this point. I just happen to think that
empirical realism is quite enough antidote against relativistic
anti-realism, not to mention against metaphysical realism.
>> >>[Rodrigo:] I am discussing ontological dependence. Since you spell it
>> out above, I
>> >>know that you mean the same by "logical dependence", though I will
>> >>continue to use "ontological" instead. Just to be sure, one object is
>> >>ontologically dependent on another if the first cannot exist unless the
>> >>second also exists.
>> >
>> >***** [Jan:] Yes, we mean the same thing where objects are concerned.
>> But for me
>> >there can also be logical dependences among statements which describe
>> >logical fictions. But counterfactually, if the fictions were objects, then
>> >those logical dependences would also be ontological.
>>
>>"You've lost me, but we can return to it when relevant."
>
>I just mean that for the 1914-18 Russell, "If there are elephants, then
>there are mammals; there are elephants; therefore there are mammals,"
>would express a logical connection, but if he is right that elephants and
>mammals are logical fictions, then there would be no ontological
>connection because elephants and mammals would be ontologically nothing.
>Similarly for "All elephants are essentially mammals," if that helps. It's
>an unimportant point for our discussion.
Ok, I get it.
>>"I don't remember ever reading the expression "immanent within our
>>referential system/theory/apparatus". What do you mean by it?"
>
>"Immanently" is Quine's term. To quote my "Quine: Whither Empirical
>Equivalence?"
>
>Immanentism is the thesis that all truths and all facts are immanent in
>scientific theory. This closely parallels the theory of naturalism.
>Immanentism is given to us by Tarski (Quine 1987: 316).
Yes, I remember it now. It is immanence as opposed to transcendence.
>> > May I also speculate that when you say that all objects are
>> > conceptualized, and also that objects and their conceptualizations are
>> > mutually independent, that for you all objects merely happen to be
>> > conceptualized (or referentialized) in the actual world, since we make
>> > sweeping assertions about all objects, but that they need not be
>> > conceptualized in order to exist?
>>
>>I'm not sure. I don't think so.
>>
>>Would it help if I mention that I take the nature of concepts to be in
>>the domain of cognitive science? I don't consider them intensional
>>objects at all. They are, roughly speaking, whatever it is that a mind
>>acquires upon learning to recognize objects as so-and-so and/or think
>>about objects as so-and-so.
>
>It helps, especially if you can assign them a spatial location inside our
>bodies. Then they will be wholly and really distinct from rocks, insofar
>as there are no rocks in our bodies. I assume you're speaking casually
>about minds and not admitting them as a category.
I fear you'll later take me to task for something unexpected if I now admit
minds as a category. So, instead, I'll admit only that there are minds. I
don't know if this is what you are asking, but since it is just about all I
want to admit, it may be answer enough to your question just the same.
As for concepts, as I said, they are entities in the theory of cognitive
science, and as such, are locatable in the mind, which is itself locatable,
roughly, in the brain. The same goes for conceptualizations, which as I
write about them I begin to think are just concepts. Is there a difference
here?
>>"The two go hand in hand. If intensional objects afforded the simplest
>>theory then, naturalism would admit them. However, it would be
>>impossible to even formulate such a theory about intensional objects, let
>>alone a simple one, without identity conditions for them."
>
>Why couldn't the simplest theory be that they are given as having
>intrinsic identity conditions?
That would be a simple theory, yes, but what is the evidence for it? You'd
have to say much more for me to comment intelligently.
>>"I'm not sure I understand the problem. How does ignorance of membership
>>preclude anything? I hereby assert that there will be as many apples and
>>oranges on your desk tomorrow. I may be wrong, of course, since I speak
>>from ignorance. But where otherwise have I faulted?"
>
>Well, you wouldn't know what sets you were talking about. You would have
>no way of telling what the set of apples is. You've never enumerated the
>membership, and that's the only way you can identify an extensional set.
>In your terms, it would be impossible to formulate a theory about the
>equinumerosity of apples and oranges without providing identity conditions
>for the sets in question.
We may have very different views here. I think I do know what sets I was
talking about. I did not know their members, but I knew the sets. Knowing
a set's members could hardly be a requirement for being able to specify the
set. I can know the set of restaurants in Boston, and though I've been to
nearly all of them, I know I've missed a few and am ignorant of a few
more. Yet surely I know the set.
>>"Some sets exist in space-time and others don't. I would say that the
>>only sets that exist in space-time are sets of points in space-time, call
>>them space-time regions. I would exclude all other sets, including sets
>>of space-time regions."
>
>Since to exist in space-time is to exist and have an extension, either
>your points have extensions or they don't exist in space-time. How can
>they exist and have space- time locations, yet not exist in space-time?
>Can a spatial point exist and have a spatial location, yet not exist in space?
According to the theory I devised ad hoc in answer to your question, only
space-time regions, that is, sets of space-time points, have extension and
hence exist in space-time. So according to this theory, space-time points
do not exist in space-time and do not have locations, though the singleton
sets containing space-time points readily substitute for these points. To
have a location is just to *be* a space-time region. Locations themselves,
if pressed to reify them, could be just space-time regions themselves.
Not to confuse matters further, but perhaps I can revise my theory, mostly
on point of terminology (or is it terminology of "point"?) Instead of
saying that space-time regions are sets of space-time points, let me say
that they are sets of space-time coordinates. In this new scheme,
space-time points are just singleton space-time regions, sets of one
member, a space-time coordinate. There will be less of a temptation to
think that coordinates have space-time locations as they are rung higher on
the ladder of abstraction.
>Do points have infinitesimal extensions? Can regions which exist in
>space-time be identical with sets of points which do not exist in
>space-time? Can a set exist in space-time even if none of its members
>does? (Most people would say the opposite, that a set of apples cannot
>exist in space-time even though apples do!) I'm well aware of the
>fallacies of composition and division, and arguably you might tough it out
>as strictly not a problem, but it seems unduly paradoxical. And you seem
>to be running directly into Zeno's paradoxes as opposed to using modern
>theory of limits. But I admire that, since doesn't theory of limits duck
>the compositional problem for those who do compose (lines and) regions out
>of points?
Continuing with my ad hoc revision, I would say that points do not have
infinitesimal extensions. Their extensions are of size zero, which isn't
to say that they do not have extensions. Regions which exist in space-time
can be identical with sets of coordinates which do not exist in
space-time. Regions are one thing and coordinates are another thing. A
set can exist in space-time even if none of its members does. I agree that
a set of apples does not exist in space time even though the apples do.
I am ignorant of the connections to Zeno's paradoxes, the relevance of
limit theory, and the fallacies of composition and division. I can already
imagine that I am going to find them "strictly not a problem", though as we
all know, in the practice of philosophy, if an intelligent philosopher
thinks its a problem, then there is a problem *at least* with the
philosopher, and I like to be there to offer help.
Rodrigo <Van...@yahoo.com>
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/_/_/980931942/
---------------------------------------------------------------------_->
Jan Dejnozka
Thank you very much for your last. This will have to be my last.
To put Quine-Carnap in perspective, even if everything you say about
Carnap being a realist the same as Quine is true, it would only
strengthen my contention that the analytic tradition is full of modified
realists by adding yet another name to the list. Protagoras would remain
as an actual example of a radical relativist, and so would many followers
of Wittgenstein and Quine.
I am tempted to reply to just these two paragraphs. Because I enjoy the
point by point discussion, I will reply to all of your post anyway, but I
do want to say that this here is surely the fulcrum of our difference.
By
"natural realism" I meant what you describe as "naturalism", an empirical
use of the word "real", not metaphysical realism. Perhaps I called it a
kind of realism because I disagree that naturalism does not "assert
anything about what there is". On the contrary, I think that that is all
that it does assert. That is all that can be done with the word
"real". It is not merely a linguistic or epistemological thesis. (I am
disagreeing with Carnap here.)
Sorry, I'm confused by your language, specifically by:
By "natural realism" I meant what you describe as "naturalism", an
empirical use of the word "real", not metaphysical realism.
Grammatically it looks like a run-on sentence. You sem to be saying that
what I describe as "naturalism" is an empirical use of the world "real."
That makes no sense because naturalism is a thesis, not a use of a word.
If the thesis of naturalism defines any use of any word, it defines the
use of "naturalism," not the use of "real."
As for Quine's "robust realism", now that you have made it clear you mean
to intepret him as having asserted a realism above and beyond his
naturalism, I think I would like to see some textual evidence, for it is
not the Quine that I know and support. His "unswerving belief that
physical objects" exist is insufficient since it follows directly from
his
naturalism. Let me be clear: I do not doubt that he asserted belief in
an
external world, in physical objects, in classes, and more. I only doubt
that these assertions were more than just naturalism.
I never said it was above and beyond his naturalism. I said it was
consistent with his naturalism. More than that, I said his robust realism
works together with his naturalism. It's asserted internally or
immanently within scientific theory, though only from one's own home
language. I don't think it would even make sense for him to assert it is
above and beyond his naturalism. And I provided five textual cites in my
last message. Again, they're on p. 267 of my book.
What is moderate is that the commitment flows from mere naturalism and
does
not extend beyond it.
In that case, let's drop "moderate" and keep "naturalistic." It's very
confusing otherwise. A thesis is not moderate just because it is
naturalistic. There are theses ranging from bold to trivial which can
turn out to be be equally well supported by naturalistic means. In fact,
sometimes more radical theses turn out to be better grounded
scientifically than many moderate theses. Not all the best physics is
moderate.
>If you feel that naturalism is inconsistent with metaphysical realism,
and
>that you accept the former but not the latter, I am also perplexed on
how
>you could accept "lumps before texts" and be arguing that Bruce was
wrong.
Because I believe that naturalism does not stop short of making
assertions
about what there is, and Bruce is defending a mildly anti-realist line.
Assertions about what there is mind-independently are by definition
assertions of metaphysical realism for Carnap, me, Butchvarov, and the
whole philosophical tradition. That's the gist, at least, of metaphysical
realism.
>I would have thought that you would be pleased to have naturalism and
>metaphysical realism working together in your philosophy, much as they
do
>in Quine's.
No, I regard opposition to metaphysics to be one of the hallmarks of
Quine's philosophy.
For Carnap, Butchvarov, me, and the philosphical tradition, the gist of
metaphysical realism is merely the general claim that there are some
mind-independent existents. Would you say opposition to that is one of
the hallmarks of Quine's philosophy? Please don't confuse metaphysical
realism with substance metaphysics. Both are traditional, but substance
metaphysics is only a rather general species of metaphysical realism.
What is the difference between realism and metaphysical realism? I
usually
take "realism" to mean "mind-independent being", and "metaphysical
realism"
to mean "mind-independent being of universals". If you have different
interpretations, I'd like to hear them.
I already gave and accepted Butchvarov's rough working definition of
"realism." That's what I
mean more fully by "realism" in the sense of "metaphysical realism." It's
also what Carnap expressly means
by "metaphysical realism," if you abbreviate it to "mind-independent
being." I quoted Carnap's
definition of "metaphysical realism," and it's the same as your
definition of "realism." Your "realism" is precisely the same as the sort
of metaphysical realism which Carnap rejects as a literally nonsensical
and external issue. And you agree with me that Quine can and does assert
the existence of mind-independent beings. And still you profess not to
see why Carnap is so different from Quine!
If you read my quotation of Carnap's definition of "metaphysical
realism," I'm confident you'll see it's basically the same as what you
take "realism" to mean. And as you know, Carnap rejects metaphysical
realism as by definition an external question. You do to see at least
that Quine is a realist-- what Carnap, Butchvarov, I, and the whole
metaphysical tradition would consider a metaphysical realist (not: a
substance realist, though I argue in my book he is very close to that--
in precisely this sense which Carnap rejects. But this has nothing to do
with the problem of universals. And you are boggling at the terminology
"metaphysical" realism.
As I explained in an earlier post, this usage of "realism" or what
is the same, "metaphysical realism," as admitting mind-independent
entities is to be distinguished from the usage of "realism" in the
problem of universals. The difference of usage is simply that the former
is general and the latter is specific to the problem of universals.
Realism-- or metaphysical realism-- is the thesis there are
mind-independent beings. Realism-- or metaphysical realism-- with respect
to universals is the thesis that properties or features of things are
mind-independent beings. Through the past thousand years of discussion of
the problem of universals, it hasn't made a bit of difference whether you
call the realist position realism or metaphysical realism. Most authors
simply say "realism." It's redundant to add "metaphysical" concerning the
problem of universals, or, for that matter, concerning the general
problem.
You need not be a metaphysical realist with respect to universals in
order to be a metaphysical realist. To be a metaphysical realist, you
merely need to admit some mind-independent entities or other. I suspect
your terminology might come from reading a few analytical philosophers.
My terminology comes from the history of philosophy.
>>"This follows directly from causality being a relation between events."
>
>I thought you rejected all relations construed as intensional! Is
>causality not an essentially intensional relation? (I assume you do not
>wish to assay it as mere regularity, thus conflating it with accidental
>generalizations.) What do you mean by "intensional" when you reject
>intensional entities? Causality is certainly intensional in Quine's
sense.
>That is, it's not truth-functional. Nor do I see how you could
"construe"
>it as truth-functional.
Is causality essentially intensional? Well, it usually is, but I don't
know if it is essentially so. I think one can just consider laws to be
those regularities that are deepest to the core of a given theory,
without
recourse to intensions. This alone, admittedly vaguely, would support
the
counterfactual contexts expected of causal reasoning.
Most textbooks on causality or philosophy of science warn against
confusing causation with the
laws of a given theory. The theory also has to be true, so that its
generalizations are true. (Of
course, for Quine, one's own home theory is one's own vehicle for
asserting truth and making
reference, but one's own theory would also have to *be* true.) And even
that's not enough to
distinguish accidental true generalization from causation. As Hume well
knew, the heart of the
concept of cause is some sort of necessity or dependence. Abandon that
heart as false or
unintelligible, and you abandon causation. That's what the whole problem
of causation is. It's not
as if it could be sometimes intensional and sometimes not. The very idea
suggests unfamiliarity
with the problem. I have no idea of how a theory consisting only of
truth-functional
generalizations could imply a non-truth-functional counterfactual of the
sort causal
necessity requires. Vagueness is not the problem. Admittedly, Quine
thinks he can get around the
problem. But dispositional properties would seem intensional (no matter
what your metaphysical
position on theory of universals) and thus only to postpone the problem.
I say this not having read Quine on the topic in some years.
>Are asymmetric relations construed extensionally ordered sets? Do you
>admit ordered sets? If so, how do you assay the ordering relations-- as
>further ordered sets needing further ordering relations?
Yes, that's right.
Don't you see that as a vicious regress? Nothing will be ordered unless
there is a final ordered
pair to order the prior pairs. Ordering also seems to be an intensional
notion.
There's an excellent critique of Quine on ordered pairs in Jose
Benardete, _Metaphysics: The Logical Approach_ (Oxford, 1989). It
convinced me, anyway! Ordered pairs are in the index.
>>"Rocks can exist even when there are no conceptualizations of rocks."
>
>This is pretty much what Butchvarov and I mean by metaphysical realism,
>which is very close to the ordinary use of "real."
In my view, there is still an ambiguity. Do rocks exist independently of
conceptualizations because their being rests on the independently
existing
universal "rock"? Or are they only ontologically independent, still
dependending on an existing conceptualization in the world in which the
rock's existence is asserted? The latter is the view I am recently
arguing. The former is my understanding of metaphysical realism.
Unfortunately, your two alternatives are mixing up the problem of
universals with the problem of realism. But
fortunately, the mixup is easy to sort out, because the relationship
between the two problems is
an extremely simple one. Namely, the problem of universals is the problem
of realism as
instantiated to the very limited topic of the ontological status of
properties or features of things.
That is, the problem of realism is the general problem of whether
anything at all exists
mind-independently (or, to give a fuller account, whether anything at all
satisfies Butchvarov's
rough working definition. That is the traditional problem of realism,
redundantly called the
problem of metaphysical realism. So Carnap is rightly defining and
attacking the traditional
problem as nonsensical, and is even doing it under the correct label of
metaphysical realism. So
he knows his history of philosphy quite well. The general problem of
realism concerns very
general arguments such as private language arguments for realism, and for
the opposing side, the basic arguments for radical relativity.
Now, the problem of universals is the specific problem of whether
features or properties
of things exist mind-independently (or, to give a fuller account, whether
they satisfy
Butchvarov's definiton-- with the caveat that since they would *be*
properties, they need not all
*have* properties, on pain of a vicious regress of properties which no
realist with respect to
universals should be saddled with, as I'm sure Butchvarov is well aware).
This is as opposed to
other specific problems such as the ontological status of God, minds,
bodies, atoms, space,
time, causation (a very special case of what admittedly might be called a
relational property), and
so on. The problem of universals, redundantly also known as the problem
of realism with respect
to universals, concerns very specific arguments such as Plato's Third Man
argument, Bertrand
Russell's vicious regress of resemblances argument, and so on.
Now, a rock is not a feature or a property. Therefore asking whether
rocks exist and are
mind-independent is a completely different and logically independent
question from asking
whether their features or properties are mind-independent existents
called universals. To think
otherwise is to deny that you can consistently be a materialist (with
respect to bodies, it goes
without saying) and a nominalist (with respect to properties, it goes
without saying)! Conversely,
you can have people like Plato denying the reality of rocks because
everything subject to change
is fleeting, impermanent, and unreal, yet consistently affirming the
reality of properties as
supremely real and perfect forms.
Now let's sort things out one step further, so as to see the full
extent of the confusion.
Metaphysical realism is the general position that *something*--
anything-- exists mind-independently. Therefore, you can be a
metaphysical realist and consistently take any one of these four options
(though not all four options at the same time!):
1. affirm that *both* rocks and their properties are mind-independently
real (Frege, early
Russell)
2. affirm that rocks, but not their properties, are mind-independently
real (William of Ockham)
3. affirm that properties of rocks, but not rocks, are mind-independently
real (Plato)
4. affirm that *rocks* things nor their properties are mind-independently
real, but that something
*else* is mind-independently real, such as God and human souls (Berkeley,
Leibniz)
We would not wish to deny that Berkeley and Leibniz are metaphysical
realists. I have
already noted that they are more specifically substance realists! We must
remember that even an
idealist can be a substance realist, though of course an idealist cannot
admit *material*
substances. As I also already noted, all substance realists are
metaphysical realists, but not all
metaphysical realists are substance realists. I think we would agree that
Quine is a metaphysical
realist but not a substance realist, insofar as he admits rocks are
mind-independent but rejects
traditional substance theory-- notably because he rejects the major
Aristotelian theme of essence.
At least, I think we would agree if you adopted the usual terminology,
which I have been
explaining.
By the way, when Butchvarov defines something as real when it exists,
has features, is mind-independent, and is language-independent, he does
not mean that the features themselves must be real. Not only would that
make his definition circular, but it would imply that nominalists and
conceptualists with respect to features, that is, with respect to the
problem of universals, could not be realists with respect to anything
else, such as rocks or God, which is absurd. It would make William of
Ockham an atheist, among other things.
Okay, we are closer. I would very much like to see the sordid quotations
you are referring to. If you formerly posted them to Analytic, please
tell
me the message number and I will go look.
My book, p. 276 has the five citations indicating that physical objects
are more real than classes for Quine.
It turns out that I have some textual support for my view as well. I
should thank you for pointing me in the direction of the source of this
quote, since I wouldn't have thought to look there without your
mention. It is the second paragraph of "On Carnap's Views on Ontology".
Not if you look at the last sentence of that paragraph. I quoted that
sentence in an earlier post: "Though no champion of traditional
metaphysics, I suspect that the sense in which I use this crusty old word
["ontology"] has been nuclear to its usage all along." And in the context
of my extensive quotation from the rest of that article in the same
earlier post, I see no support for your view there. The entire second
paragraph follows:
Indeed, he was not a champion specifically of *substance* metaphysics. If
you keep reading to the end of the paragraph you see him championing
traditional ontology on a more highly generic level, which is my level.
It is true that Carnap's analytic and internal ontological statements are
devoid of factual content, but so what? Why is semantic content not good
enough? The standard should be only that he believed things exist
independently of mind. The standard was met.
I would need to understand your distinction between semantic and factual
content. For Quine, facts of
the matter devolve to differences in microphysical states. Theories are
factually different if they
make different predictions of microphysical states. Quine does not admit
a metaphysical
*category* of facts as, notably, Russell and early Wittgenstein do. Now,
semantic conditions are
what make a theory true (individual statements do not have individual
semantic conditions, thogh
observation statements *almost* do). But this ultimately devolves to
microphysical states, so
what's the difference for Quine?
And the standard was not met by Carnap. Carnap expressly rejected
the standard, which he called by
the name of metaphysical realism. I even quoted his definition! He
defines metaphysical realism as existing independently of mind, and he
rejects it as literally nonsensical. For independently of mind read
independently of human framework. Your position emerges as the position
that Carnap says that existence independently of framework can be
determined internally to framework. That sound contradictory to me, and
out of charity I would not saddle Carnap with a self-contradictory
position. You might as well say that for Carnap, external questions are
internal questions, if you are going to say that for him questions of
mind-independent existence are dependent on our framework..
I'm lost on your suggestion, but I suspect it doesn't matter, since the
real question is whether
both admit matters of fact, in the sense of facts which are
mind-independent. Quine does, and
Carnap rejects the very admission as literally nonsensical. When I say
Quine's commitment to
matters of fact might as well be a priori, I was referring to its
epistemological strength, but I also
hand in mind that the a priori-a poseriori distinction is irrelevant to
the admission of facts. One
who admits facts, either in the general sense of metaphysical realism
where "fact" is used in a
loose ordinary sense, or in the specific sense of a specific category of
metaphysically real facts,
where "fact" is used in a technical philosophical sense, can admit either
synthetic a priori or
synthetic a posteriori facts, or both. Please remember that I'm
discussing here, as consequences
of Quine's admission of facts (in this case microphysical states) as
(metaphysically) real, his
derivative abilities to believe and to provide strong evidence for
factual assertions.
This is all news to me. Since you are only saying that Quine sided with
Frege, maybe you can point me toward wherever Frege made the distinction
between existence and reality. If you could show where Quine also
accepts
this difference that would be great too. If you would prefer that I just
get it from your book, then it will have to wait until then.
You're not going to find it in a single text in Frege. Some texts discuss
reality, others existence. But you will find both concepts in the same
works, notably both major works, _Grundlagen_ and _Grundgesetze_. Perhaps
the most interesting discussion of what is real is at the end of "Der
Gedanke." The distinction is universally acknowledged by Frege scholars.
Please believe me when I say it's completely unquestioned and basic to
understanding Frege. I take Quine to have made much the same distinction
for a period of at least 21 years, though more unobtrusively.
For Frege, to be real (reell) or actual (wirklich) is to be causal. To
excerpt from my book, chapter 2:
When Frege fully presents his realism, he distinguishes between the real
and the merely objective within the domain of objects on the basis of
causal considerations. To be real is to have causal impacts (“The
Thought”) or at least causal capacities (BL 16). This is a traditional
metaphysical definition straight from Plato (Sophist 247; see Geach 1969:
65). Even the sophistical twist that imperceptible, merely objective
thoughts can ‘act’ by being grasped and judged as true, and thus have at
least a modicum of reality (“The Thought”), is from Plato (Sophist 248;
see Geach 1969: 66).
BL is _The Basic Laws of Arithmetic_, Furth trans. and ed., p. 16. Or
even in _Foundations, trans. Austin, p. 35, _ "I distinguish what I call
objective [Objective] from what is handleable or spatial or actual
[Wirklichen]." That's a very early version. Frege says in "The Thought,"
trans. A. and M. Quinton, "The world of the real is a world in which this
acts on that, changes it and again experiences reactions itself and is
changed by them. All this is a process in time. We will hardly recognize
what is timeless and unchangeable as real. Now is the thought changeable
or is it timeless? The thought that we express by the Pythagorean theorem
is surely timeless, eternal, unchangeable.... There is lacking here
something we observe throughout the order of nature: reciprocal action.
Thoughts are by no means unreal but their reality is of quite a different
kind from that of things [because they cannot be acted on, and act only
indirectly on us by being apprehended]." There's also a text that says,
"Something entirely and in every respect inactive would be unreal and
non-existent for us. Even the timeless, if it is to be anything for us,
must somehow be implicated with the temporal." One must not confuse this
use of "non-existent" with existence as a second level concept for
Frege.
Existence, the second-level concept denoted by the existential
quantifier, is Existenz.
Both real (concrete) and merely objective (abstract) objects such as
numbers exist for Frege. He uses the same second level concept of
existence as denoted by the existential quantifier to assert (Ex)Fx for
real objects and abstract objects alike. This is basic to understanding
Frege. Some cites are: _Basic Laws_, sect. 21 (not readable unless you
start from the beginning of the book), _Foundations_ sect. 53. Also see
the famous essay "On Concept and Object," where Frege says:
"I speak of properties asserted of a concept, and I allow that a concept
may fall under a higher one. I have called existence a property of a
concept.... Second-level concepts, which concepts fall under, are
essentially different from first-level concepts, which objects fall
under."
You need to read the whole essay. To exist, concepts need not be
classificatory. But to be significantly informative, concepts must be
classificatory. That is, it must be possible for them to fail to apply to
something. The concept real significantly applies to objects because some
are real (causal) and others are not. The concept existence does not
significantly apply to objects because necessarily, all objects exist
(there is no such thing as a merely possible object). Therefore, since
the concept existence is significant, it must be a second-level concept
predicated of first level concepts. Thus "There exists a horse" means,
the first-level concept horse falls within the second-level concept of
being realized (or instantiated). Or in Frege-Russell-Quine notation,
(Ex)Hx. But abstract objects exist too. "There exist numbers" means, the
first-level concept number falls within the second-level concept of being
realized or instantiated.
Frege makes existence a second-level concept as early as 1897
_Begriffsschrift_, p. 27 note 15, in Jean van Heijenoort, ed., _From
Frege to Goedel_. There Frege equates asserting there are houses with
asserting that there is at least one house. And as we all know, for Frege
assertions of number are assertions about concepts, not about objects.
That is the heart of his whole philosophy of arithmetic. In _Foundations
of Arithmetic_ Frege equates assertions of existence with "denial of the
number naught." That is, there exist horses means it is not the case that
there are no horses. Note you can say there is at least one real object
as easily as you can say there is at least one abstract object. You can
say there are not zero real objects as easily as you can say there are
not zero abstract objects.
Russell makes much the same distinction between real and existing as
Frege and Quine, only for him what is ordinarily said to be real is
assayed as a logical fiction. See chapter 4 of my first book, where what
corresponds to Frege's reell or wirklich (Frege also calls real objects
concrete objects, at least in translation) is Russell's second or
correlational sense of "real.".
Here are some extra cites from _Word and Object_, p. 221: "If we are
limning the true and ultimate structure of reality, the canonical scheme
for us is the one that knows... no propositional attitudes but [knows]
only the physical constitution and behavior of organisms." On p. 3, he
says that if we deny that physical objects are real, "[w]e should only be
depriving them of the very denotations to which they mainly owe such
sense as they make to us." Classes or sets need not apply for jobs here,
and this has nothing at all to do with Quine's existential quantifier,
which is predicated of predicates, not of objects. Exactly as in
_Theories and Things_ 21 years later, the fellow is telling us that
physical objects are real, and he's not saying that of anything else.
That makes a nice context for physical objects being "the things, first
and foremost."
I agree that Quine is a robust realist. I disagree that this is a form
of
metaphysical realism. "First and foremost" along what ordering? I see
no
evidence that he means that physical objects are more real.
You have to be sensitive to tradition. "First and foremost" is a perfect
description of the place of
substances in Aristotelian metaphysics. It's also basically the
description the tradition used. The
traditional terms are "primacy" and "primary." Quine admitted not knowing
much about history
of philosophy, but in what little he does say about traditional
philosophy, he doesn't go wrong
anywhere that I can think of. He was a pretty meticulous researcher.
I've already explained what I and Carnap and Butchvarov and the
tradition all mean by "metaphysical realism." Basically, it means
mind-independence. Your terminology seems not just deviant, but
idiosyncratic!
What is metaphysical evidence? Would the existence of angels be
metaphysical evidence of the existence of God?
The passage was cited only on reality. We all know (I hope) that he makes
existence a second-level predicate, namely the existential quantifier,
following Frege very closely by making that quantifier ontologically
committal. As with Frege, you're not going to find a single text making
the distinction, because the concepts are so different-- categorially
different, as I noted-- and the issues are wholly different. It would
have seemed otiose to Frege, Russell, and Qune alike to state that such
obviously different concepts are different from each other. They cannot
even be on the same type-level in simple theory of types! Frege has
simple theory of types, but of course not ramified theory of types. If he
had ramified types, he would have had Russell's solution to Russell's
paradox.
In my view, it's that, but it's not just that. It's that his scientific
theory has factual content. When he rejects
Carnap's distinction between external and internal, he doesn't just drop
the external and keep the
internal unchanged. The internal changes a great deal. It becomes
factual. It incorporates the
factual and mind-independent character of the rejected external side of
Carnap's distinction.
I think it's even worse to think that when Quine rejects Carnap's
core distinction between
analytic and synthetic, he just drops the analytic and keeps the
synthetic, because "synthetic" is
*defined* as not analytic. You *could* say everything is then synthetic,
in the negative sense
that nothing is analytic, and it's the concept of analyticity he finds
problematic. But I think that's wrong.
Rather, the entire distinction vanishes and what you are left with is
that everything is *factual*
and *a posteriori*, which are very different from *synthetic* Factual
means mind-independently
existing, a posteriori means not knowable independently of experience,
and synthetic means
having a propositional or statement structure not of the sort defined as
analytic.
By "ontological foundations" I meant "ontological categories at the first
tier", and by "guides by which wo judge our theories" I meant "high
relative commitment in the face of recalcitrant evidence". I better
explain myself in these other terms elsewhere in this post.
Sorry, still confused. By at the first tier, do you mean the category of
physical objects, following
my use of "two-tiered"? And can't a guide be a good or poor guide, so why
does the commitment
always have to be high? We very often decide which theories to accept on
the basis of weak or
mixed evidence.
>Your observation of a disconnect between Carnap's methodological
>commitments and his actual beliefs seems not only to support my
>interpretation, but to suggest that our interpretations are not so far
>apart as you might think.
I don't think they're not connected, just not identical. His
methodological commitments lead to his beliefs. It just seems to me that
if we are going to call him an anti-realist, then it is his beliefs that
should count for or against him, not his methodological commitments.
Then I think all you mean by saying he believes in realism is that he
believes some things are empirically real, as I quoted his definition of
"empirically real." Such beliefs will not be about what is factual in the
sense of
(metaphysical) realism, as I quoted his (very traditional) definition of
"metaphysically real." That is, they will not be beliefs that there are
mind-independent beings in the sense of traditional philosophical
realism. The
disconnect I originally saw between his methodology and any belief in
real facts (which is what I
thought you meant) remains as complete as ever. It is nothing less than,
and exactly the same as,
his disconnect between external and internal.
>>"Quine's innovation was to start in the middle, thereby obviating the
>>necessity for "methodological" commitments. Quine's "method", if you
>>will, was just as physicalist and realist as his conclusion, but not
>>because he had the a priori metaphysical commitments that Carnap
lacked,
>>rather it was because he didn't start with the a priori, preferring to
>>deny its availability altogether."
>
>What Quine means by starting in the middle is starting by positing
>medium-sized dry goods, _Word and Object_, p. 4. This is precisely the
>initial (and revisable) commitment his methodology requires. I agree
with
>the second sentence, except that I would say instead that Carnap's a
>priori commitment was to an anti-metaphysical verificationism intimately
>related to an a priori analytic-synthetic distinction.
Yes, that's right. We don't disagree about where they began. We only
disagree about where they ended. You think Carnap was stuck with
relativistic anti-realism, and I think that he ended, alongside Quine,
with
physicalism.
Quine's physicalism is factually real, i.e., metaphysically real. That's
imported into his
internalism from Carnap's externalism. And you can assert and believe it,
though only in
your own home language. Carnap's physicalism is merely empirically real,
in Carnap's sense.
Carnap would be the first to condemn the assertion that physical objects
are metaphysically real
as literal nonsense and a covert merely pragmatic linguistic proposal at
best.
Let me put the problem another way. Allow me to pose the following two
dilemmas for Carnap.
DILEMMA 1. Supposedly any statement can be construed as internal or
external. Now, he defines empirical reality as the [only] sort of reality
which can be and is asserted internally, the only sort science can
assert. He defines metaphysical reality as the sort of traditional
reality which is externally asserted, and finds it literally nonsense and
covertly linguistic proposals. Now consider S:
S. Stones are empirically real.
Can S be construed as either external or internal? In particular, can S
be construed as an external assertion? I hardly think so. That would
flatly contradict what S asserts, because empirical reality is defined as
the sort of reality which is handled internally.
DILEMMA 2. Now consider S*:
S*. Stones are metaphysically real (which Carnap defines as, stones exist
mind-independently).
Can S* be construed as either external or internal? In particular, can S
be construed as an internal assertion? I hardly think so. That would
flatly contradict what S* asserts, because metaphysical reality is by
definition something that cannot be handled internally.
I don't think Carnap perceived this two-part dilemma of semantic ascent.
Nor do I think the dilemma is serious. I think the solutions I indicated
are obvious, and I thikn Carnap would agree with them. But you are
running headlong into dilemma 2 when you say that any statement. for
Carnap can be construed internally or externally, in particular the
statement that stones are mind-independent. That's the one thing that by
his own definitions of "external" and "internal" he cannot construe
internally. In effect, you are taking the self-contradictory position
that for Carnap, S* can be construed internally. You might as well also
hold by parity of reason that for Carnap, S can be construed externally!
To sum up dilemma 2, there is no way out for Carnap's being able to
assert mind-independent facts internally. The very effort creates a
contradiction. As I've been saying all along, he cannot assert
metaphysically factual contents, either externally or internally, but
Quine can assert them internally.
>rejects that there can be degrees of analyticity. It's a big part of his
>argument in "Empiricism, Semantics, and Ontology" that analytic internal
>questions are either true or false, and can admit of no degree, while
the
>same questions construed externally are neither true nor false, but are
>only pragmatically useful to some degree. As I see it, Quine rejects all
>talk of degrees of analyticity along with yes-or-no analyticity. He
admits
>instead degrees of linguistic content and converse degrees of empirical
>content.
It is these degrees of linguistic and empirical content that I am
referring
to. Following Quine's example, since "analytic" is now an obselete term,
I
freely adopt it to new usage. And again, like Quine, I feel that the
adoption is not so arbitrary, that, in fact, "the sense in which I use
this
crusty old word has been nuclear to its usage all along".
When you adopt a new usage, you have a responsibility to explain it at
the time of use, not after the fact. And on your new usage, you appear to
be agreeing with me that for Quine every internal assertion has some
degree of factual content, and by implication that that is not so for
Carnap's internal statements, which are on your view totally yes-analytic
and thus have zero factual content, meaning no factual content.
>As to the preposterousness claim, let me embroider on that. If "There
are
>physical objects" is analytic a priori, then we should be able to give a
>formal a priori proof of the existence of physical objects.
Why so? Rather than a proof, it requires only the adoption of a
linguistic
framework.
I think that for Carnap, analytic truths are true independently of
framework in the sense that they would be, in possible worlds talk, true
in all possible worlds, precisely because they have no factual content.
You make it sound as if "There are physical objects" is so obviously
analytic and a priori, it does not even need a proof to see it is true!
All you have to do is a lower-level activity of adopting a linguistic
framework that postulates that there are physical objects. As Russell
would say, that has all the advantage of theft over honest toil.
>Empirical realism is what the pre-philosophical scientist would deem
real
>as opposed to an illusion, hallucination, dream, mirage, decoy, or fake.
>It seems related to Austin's theory of the ordinary use of "real" as
well
>as at least somewhat to Kant's conception of empirical reality.
>Metaphysical realism is as opposed to other metaphysical positions such
as
>idealism or solipsism. Thus calling Carnap an empirical realist is
>certainly correct, but adds nothing to the discussion.
Well, that altogether depends on how you want to frame the discussion.
We
are discussing not only metaphysical realism, after all, but realism as
opposed to relativistic anti-realism. The central questions are (1) Was
Quine a metaphysical realist as well as an empirical realist? and (2) Was
Carnap a relativistic anti-realist?
If you'll recall the definitions I excerpted from my book, radical
relativity is defined as a metaphysical position opposed to metaphysical
realism (specifically, to modified realism, which is a form of
metaphysical realism.) If you want to discuss empirical or scientifically
ascertainable relativity as opposed to empirical or scientifically
ascertainable independence of mind, you are back to the humble level of
Whorf construed as making a mere scientific hypothesis. I have no problem
with it, and Quine has no problem with it, and for him and me the two
levels, metaphysical and scientific,can be made to coalesce inside
naturalism, with scientific evidence becoming relevant to the thesis of
metaphysical realism, but Carnap has his hands tied behind his back.
I think that by empirical realism, Carnap could be allowed to mean
causal mind-independence in a sense in which scientists could
investigate, say, whether differences in perceptions of colors are
causally influenced by differences in human language. But metahysical
realism concerns logical mind-independence. And when I posed dilemma 2, I
Griceanly uttered S* meaning that stones exist and are logically
mind-independent." You can't escape from the dilemma of trying to
construe S* as an internal statement by switching meanings from logical
to causal mind-independence! But Quine can and does assert S* from within
his home language, in my opinion.
By the way, for Quine, following Tarski, the simplest argument of all
that Quine would admit mind-independent objects that existed long before
languages and people did would be this. For Quine, "S*" is true if and
only if S*. And "S**" is true if and only if S**, where S** is
S** There existed objects long before people and languages existed.
Recall Quine says, still following Tarski, "There are rabbits" just says
that there are rabbits, whatever they are. The "whatever they are" is a
qualification due to referential inscrutability and translational
indeterminacy. But since these drop out and one can safely refer in one's
home language, by parity of reason the "whatever they are" should drop
out too. So when one says in one's home language, "objects which existed
long before languages or people did," there is no need to add "whatever
they are." Even if you did, you wuld only switch to other "slicings" of
the same objects. Their undetached parts and temporal slices would also
have existed long before languages and people did too. I thought of this
new point just a day or two ago.
As I see it, Carnap's empirical realism, which was not substantially
different from Quine's, was enough to oppose him to any relativistic
anti-realism. To idealism and solipsism, Quine and Carnap were both
opposed in substantially different ways. Quine thought them demonstrably
false and Carnap thought the doctrines meaningless.
>We've already discussed the distinction using other terms. To say that
>Quine collapses the internal-external distinction so as to make all
>existential statements internal yet also genuinely factual and about the
>world, is just to say that Quine collapses the empirical
>realism-metaphysical realism distinction so as to make all existential
>statements empirically realist, yet also genuinely factual and about the
>world.
I'm not so sure that he did collapse this distinction. Rather he just
did
away with the metaphysical side. The model here is not the collapsing of
the analytic/synthetic distinction, but the doing away with the a
priori. Just as Quine isn't said to have collapsed the distinction
between
the a priori and the a posteriori, I think it should not be said that he
collapsed the distinction between metaphysical and empirical
realism. Rather, he repudiated the former and made "ontology" a matter
of
the latter.
I think I address this earlier. He did not just do away with the
metaphysical side. He imported assertions of metaphysical realism into
the internal side. Again, we must not confuse analytic with a priori.
Rejecting the analytic-synthetic distinction is very different from
rejecting the a priori-a posteriori distinction.
>Carnap says in _Aufbau_, sect. 170:
Hey, how did you get your hands on a copy of the _Aufbau_? I have tried
finding a copy several times, with no success.
I don't have it in German. Did you try interlibrary loan? Or Amazon.com?
Try searching Rudolf Carnap or _The Logical Structure of the World_ (I
haven't done it). I have the Rolf A. George translation by University of
California Press, which includes _Pseudo-Problems in Philosophy_. I
bought the 1969 paperbound probably in the early 1970s for $5.95. (That
was highway robbery-- I bought the clothbound _Hegel's Science of Logic_,
trans. A. V. Miller, around 1973 for $15!) I pretty much stopped buying
philosophy books in 1985 when I had to buy Husserl's two-volume
clothbound _Logical Investigations_ for $50. Outrageous! And here we are
in 2001, with my clothbound officially selling for $81. It's just too
embarrassing. I'm glad people can get it now for $30.95 by ordering the
paperback. And it's all thanks to members of analytic.
>It should be abundantly clear that the ONLY sense in which Carnap admits
>he is a realist is empirical realism, since it is the ONLY sense which
he
>admits as occurring internally and hence meaningfully within his
>linguistic frameworks. It should be abundantly clear that metaphysical
>realism is what he is rejecting as literally nonsensical in "Empiricism,
>Semantics, and Ontology," and what he considers to be covertly nothing
>more than disguised pragmatic linguistic proposals which are neither
true
>nor false.
Yes, of course. No argument on this point. I just happen to think that
empirical realism is quite enough antidote against relativistic
anti-realism, not to mention against metaphysical realism.
It's a good antidote against empirical relativism, anyway. I see
empirical realism and metaphysical realism as Carnap defines them as
logically consistent. Of course, I'm no verificationist and not a
naturalist really, so I can get away with it very easily But I think
Quine can and does see them as consistent within naturalism, and would
affirm both of them.
>Well, you wouldn't know what sets you were talking about. You would have
>no way of telling what the set of apples is. You've never enumerated
the
>membership, and that's the only way you can identify an extensional set.
>In your terms, it would be impossible to formulate a theory about the
>equinumerosity of apples and oranges without providing identity
conditions
>for the sets in question.
We may have very different views here. I think I do know what sets I was
talking about. I did not know their members, but I knew the sets.
Knowing
a set's members could hardly be a requirement for being able to specify
the
set. I can know the set of restaurants in Boston, and though I've been
to
nearly all of them, I know I've missed a few and am ignorant of a few
more. Yet surely I know the set.
Quine speaks in just this way too. He says he cannot enumerate the
members of all the sets he needs, so he can just specify a membership
condition. The problem is terminology. This makes what he calls sets what
most people call classes, and having to specify membership this way is
precisely what defines "intensional" in one of the three classic meanings
of "intensional" Russell indicates in _Principia Mathematica_, pp. 39-40.
It's not Quine's favored use of "intensional" as not truth-functional,
which is also in _Principia_, p. 8, but it is a classic meaning of
"intensional." There's no way out of it for you, since you satisfied the
definition directly, following Quine! Perhaps you and Quine need to be
more careful of what you mean when you say you reject intensional
entities! There's a third meaning in _Principia_, too! See my second
book, page 7.
To have a location is just to *be* a space-time region.
Do apples then have no location? They are not space-time regions! Or if
they are, then at least some space-time regions are rolling about and are
quite tasty, not to mention grown on trees and sold in grocery stores!
Locations themselves,
if pressed to reify them, could be just space-time regions themselves.
Yes. What else could they be?
[The extensions of points] are of size zero, which isn't to say that
they do not have extensions.
I think that's self-contradictory, or else you have a strange new meaning
of spatial "extension." In fact, it hardly matters what you mean by
"extension." Zero extension is no extension in the same way zero gas is
no gas, zero apples is no apples, zero horses are no horses, and zero
progressive development is no progressive development. So it looks like
you have a strange new meaning for "zero" too! Most unFregean!
Rodrigo, yours has been the best discussion of my book I have ever had
with anyone. You've been more helpful than the professional reviewers,
and more plausible in your objections! If you had read my whole book in
advance, I think we'd be much more in agreement, especially
terminologically. Similarly if I had read more Carnap, whom I read just
enough to come to my interpretation. I think that all or nearly all of
our disagreement is either terminological or about sorting out logical
consequences of Quine's and Carnap's views. But you've made me aware
again of how very compressed my book is, especially my brief discussion
of Quine (not to mention the quickie treatment of Carnap!), and how much
there is to unpack in it. More importantly, it's just so refreshing to
hear someone else who has read much Quine and Carnap, but who has
developed a thoughtful and different point of view. But what I really
hope you will share more in the future is your own views. That's what
I'll be doing in my third book. Maybe we can compare our views a few
years from now! I'm not a naturalist per se, but I'm not against natural
science by any means! If you broaden "science" to include any organized
body of rational study, maybe even I would be an attenuated naturalist.
Quine is pretty much that broad in scope, if not in traditional
intension. I'll have to think about that! For me this is the problem of
the given. But most of all, thank you so much for your kindness,
patience, and hospitality to this solitary but faithful pilgrim on the
philosophical scene.
Best wishes,
Jan.
http://www.members.tripod.com/~Jan_Dejnozka/index.html includes vitae,
abstracts
of publications, philosophy book announcements, book corrections,
retrospects on
books, unpublished papers, brief lists of philosophical, musical,
literary, artistic,
and cinematic favorites, and a family photo and art gallery.
________________________________________________________________
GET INTERNET ACCESS FROM JUNO!
Juno offers FREE or PREMIUM Internet access for less!
Join Juno today! For your FREE software, visit:
http://dl.www.juno.com/get/tagj.
------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/_/_/981003640/
---------------------------------------------------------------------_->
Rodrigo Vanegas
>Dear Rodrigo,
>
>Thank you very much for your last. This will have to be my last.
That's too bad. We were only just getting started. Since it is you who is
leaving early, I will take the liberty of having the last word.
>To put Quine-Carnap in perspective, even if everything you say about
>Carnap being a realist the same as Quine is true, it would only
>strengthen my contention that the analytic tradition is full of modified
>realists by adding yet another name to the list. Protagoras would remain
>as an actual example of a radical relativist, and so would many followers
>of Wittgenstein and Quine.
Maybe so, and I don't doubt that a category which is all-inclusive, or
nearly all-inclusive, is still informative. It is worth noting, however,
what the information is, finally. If Carnap does fail to be a modified
realist, then you have said something interesting about the Carnap/Quine
debate, one of most central philosophical debates of the 20th century. On
the other hand, if Carnap *is* a modified realist, then what you have told
us is, at the end of the day, something about Protagoras, and perhaps
something about philosophers as a whole. Since what is insightful and
exciting about philosophers is where they disagree, my personal preference
would be to pass over your thesis. (Unless I had a special interest in
Protagoras, which I don't.)
Since you will not be returning, I will save the point-by-point reply for
the possible occasion of your changing your mind about unsubscribing to
Analytic. In its palce, I would like to discuss what I think are the
highlights of the discussion so far.
In my previous post I offered definitions for realism and metaphysical
realism and asked you to agree with them or offer your own alternatives. I
defined realism as the doctrine of mind-independent being, and metaphysical
realism as the doctrine of mind-independent being of universals. You
replied that as you use the terms, tne qualifier "metaphysical" is
redundant, and that universals are irrelevant. Very well, I will keep to
my usage outside of this discussion, but for now I will work with the
terminology you prefer.
To this classification scheme you also added the category of "empirical
realism". This term comes straight out of Carnap's distinction between
internal and external questions. Internal questions about what is real are
the questions of empirical realism. By contrast, external questions about
what is real are the questions of what Carnap calls "metaphysical
realism". He even goes on to define metaphysical realism as "independence
from the cognizing consciousness". Since I don't have the _Aufbau_ yet --
I ordered a used copy yesterday -- I cannot look to see what he wrote
immediately after the text you quoted. I would be interested to know. I
think my argument may depend on interpreting Carnap differently than you do
and I cannot do this without knowing what he thought was metaphysical, in
Carnap's sense of "metaphysical", about the notion of "independence from
the cognizing consciousness". For Carnap, a statement was "metaphysical"
if it cannot be expressed internally to a linguistic framework. I do not
immediately see why cognition, consciousness, and dependence are external
to any reasonable linguistic framework. My guess was that he had a problem
with "consciousness" since at that time, this concept had not been
reclaimed from the thought of less analytic philosophers, as it has been in
the past twenty years or so.
In my previous post I observed that the central questions of this
discussion are (1) Was Quine a metaphysical realist as well as an empirical
realist? and (2) Was Carnap a relativistic anti-realist?
The status of these questions is mixed. I can now agree that Quine was a
metaphysical realist, but only in the sense of "metaphysical realist" that
you have insisted upon. You also argue that this is the sense that has
been common currency predominantly over the history of philosophy. I'm not
so sure, but I also don't care to argue the point either way. What I will
say is that Quine's realism, whatever else you may want to call it, stands
in relation preceding philosophy in at least two interesting ways. First,
it is a posteriori. Second, it does not admit of universals (viz.,
intensions, properties, meanings, etc.) If this makes Quine consistent
with modified realism, then so much the better for your thesis.
As for the second question, my position remains the same. Carnap's
empirical realism is consistent with realism, that is, with you what you
are calling "metaphysical realism", notwithstanding his rejection of what
*he* means by "metaphysical realism". Now I must admit that his
association of what he calls "metaphysical realism" with "independence from
cognizing consciousness" bears a strong resemblance to realism as we have
been using the term. This puts a snag in my position I will not easily
resolve without further understanding why Carnap considered "independence
from cognizing consciousness" to be metaphysical. Nonetheless, I remain
confident that no serious problem is posed. Since I'll soon have the book
in hand, maybe I can get back to you then. There is a second reason why
empirical realism might not be considered a form of "metaphysical realism",
as you and I are using the term. You argued that even if Carnap was able
to admit to the truth of internal analytic statements like "there are
physical objects", these would not have factual content, or be true as
matters of fact. Instead they would be mere semantic truths, or be true as
matters of meaning. I agree that they would be semantic truths, not
factual truths, but I do not agree that they not then be truths about the
mind-independence of things. And if this makes Carnap consistent with
modified realism, then so much the worse for your thesis.
Well, that's all I want to add for now. I will write more if I get a reply
or when I receive Carnap's book, whichever comes first.
Rodrigo <Van...@yahoo.com>
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/_/_/981086660/
---------------------------------------------------------------------_->
Rodrigo Vanegas
system. -moderator]
Rodrigo <Van...@yahoo.com>
http://click.egroups.com/1/11231/0/_/_/_/981087073/
---------------------------------------------------------------------_->
Jan Dejnozka
Thank you very much for your last. This will have to be my last.
To put Quine-Carnap in perspective, even if everything you say about
Carnap being a realist the same as Quine is true, it would only
strengthen my contention that the analytic tradition is full of modified
realists by adding yet another name to the list. Protagoras would remain
as an actual example of a radical relativist, and so would many followers
of Wittgenstein and Quine.
I am tempted to reply to just these two paragraphs. Because I enjoy the
Yes, that's right.
opposed to relativistic anti-realism. The central questions are (1) Was
Quine a metaphysical realist as well as an empirical realist? and (2) Was
Carnap a relativistic anti-realist?
If you'll recall the definitions I excerpted from my book, radical
>world.
Best wishes,
Jan.
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/_/_/981087070/
---------------------------------------------------------------------_->
Rodrigo Vanegas
You have unsubscribed already, so if you read this it will be because it is
publically accessible on the website or because you have resubscribed. I
write for those who were following the discussion and for the record.
We left off with a point to be settled by a closer look at the
_Aufbau_. As it happens, I have managed to get my hands on a copy, and can
now read through the context of the quotes you selected for your
argument. What follows is the section in your post where you present your
argument with the quotes. After this I include my reply.
I do agree that Carnap admitted only to what he called empirical realism
and not metaphysical realism. It may be worth taking a closer look at what
he thought each of these entailed. As your quote from section 175 makes
clear, objects are metaphysical real, for Carnap in the _Aufbau_, if they
are "independent from the cognizing consciousness".
Elsewhere in our discussion we had come to settle on a particular kind of
independence, namely ontological independence, which I have defined in my
discussion with Bruce as a dependence of being. Xs are ontologically
dependent of Ys iff if there were Xs then there would be Ys. It doesn't
seem, it turns out, that Carnap had ontological dependence in mind. In
section 176 he explains further. (I wish you had checked for me, but I
know how these things are. Fortunately, I found the book.)
"Section 176: The Metaphysical Concept of Reality.
The concept of reality (in the sense of independence from the cognizing
consciousness) does not belong within (rational) science, but within
metaphysics. This is now to be demonstrated. For this purpose, we
investigate whether this concept can be constructed, i.e., whether it can
be expressed through objects of the most important types which we have
already considered, namely, the autopsychological, the physical, the
heteropsychological, and the cultural. At first sight, it might appear as
though this were possible. An object which I have recognized, that is, an
object which has been constructed on the basis of my experience, will have
to be called "independent of my consciousness" if its constitution does not
depend upon my will, i.e., if an act of volition which aims at a change of
the object does not result in such a change. But this does not agree with
the concept of reality as it is meant by realism and idealism (the former
ascribing it to, and the latter denying it of, physical bodies). For,
according to the definition which we have just attempted, a physical body
which I hold in my hand should not be called real, since (even in the
opinion of the realists) it changes if I carry out an appropriate act of
will; this would then be contradictory to the realistic position. On the
other hand, this definition requires that any physical thing which lies
outside of our technological reach, for example, a crater in the moon,
should be acknowledged as real, since (even in the opinion of idealism) it
does not change if I carry out an appropriate act of will; this then would
be contradictory to the position of idealism.
One could try in various other ways to give a definition of reality (in the
sense of independence of my consciousness) in such a way that the concept
becomes constructible. However, one can show in each such case that the
concept which is so defined does not agree with the concept as it is meant
by realism as well as by idealism. It must be noted that this holds, not
only of a constructional system which has the system form represented in
our outline, but for any experiential constructional system, even for a
system which does not proceed from an autopsychological basis, but from the
experiences of all subjects or from the physical. The (second) concept of
reality cannot be constructed in an experiential constructional system;
this characterizes it as a nonrational, metaphysical concept.
...
The definition of the concept of a thing-in-itself goes back to the concept
of reality (in the sense of independence from the cognizing
subject). Thus, in our conception, this concept, too, must be placed
within metaphysics, for metaphysics is the extrascientific domain of
theoretical form."
It appears, then, that Carnap would not, and probably did not, think much
of our notion ontological independence. It's not mentioned to be sure, but
he does say that he couldn't think of anything constructible that would
serve as an adequate definition of reality distinct from empirical
reality. The question for us is whether "ontological dependence", as we
have defined it, is constructible in Carnap's system. I can think of no
reason why it wouldn't be. He has a way of expressing the existence of
physical objects and of other minds (the "heteropsychological"). All that
remains is a touch of de dicto counterfactual syntax and presto,
"ontological dependence" is constructible. And if it is, then according to
section 176, it isn't really what realist and idealist metaphysicians are
talking about or Carnap was a metaphysician after all. Either way, the
result is that Carnap slips within the all-encopassing fold of "modified
realism" and your thesis contracts to a portrait of the remaining and
unique anti-modified-realist, Protagoras.
Rodrigo <Van...@yahoo.com>
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/_/_/981794101/
---------------------------------------------------------------------_->