Monday, December 04, 2006

Chalmers on "Modal Rationalism"

In my undergraduate/graduate seminar this quarter, we read David Chalmers's influential book, The Conscious Mind, and now we’re reading some of the subsequent criticism and discussion of it, and Chalmers’s replies.  A common theme in many replies -- and my sense, too, is that there’s something fishy about reflecting from the armchair about what we can conceive and reaching conclusions on that basis about the fundamental structure of the universe.

Chalmers’s response to this (in Philosophy & Phenomenological Research, 1999, p. 490; see also "Does Conceivability Entail Possibility?") is interesting. He insists -- correctly, I think -- that either you’re doing empirical exploration or you’re engaged in a “rationalist” enterprise centering on ideas such as “consistency, entailment, and ideal conceivability” -- that either you're doing scientific research or you’re exploring our concepts. In particular, he attacks the idea that something might be conceptually possible but still metaphysically impossible: Metaphysics just is about our conceptual space. This is hard for what Chalmers calls “Type B materialists” to swallow: They want to embrace materialism as a “metaphysical” thesis and at the same time allow that it’s conceivable that materialism is false, that it's conceivable (for example) that “zombies” -- beings physically identical to us but with no conscious experience -- exist.

As I said, I’m inclined to agree with Chalmers about this and disagree with the majority of his critics. But I think this only raises even more sharply the fundamental concern that seems to me to be driving them.  If Chalmers’s (and all of our) “metaphysics” is just exploration of our concepts, how can we claim to discover anything about the fundamental structure of the universe thereby -- anything about anything other than our concepts? Metaphysics seems then to become a branch of psychology.

Now, actually, I’m quite happy with that, but I’m not sure Chalmers should be, and it isn’t the tenor of The Conscious Mind as I read it. And if materialism is true, then I’d say it’s not -- or shouldn’t be -- construed as a metaphysical thesis at all, but rather as a scientific thesis, a claim only about the “laws of nature”, and not a claim about Kripkean “a posteriori metaphysical necessity” or the like.


Anonymous said...

Hi Eric,

Is it metaphysically necessary that concepts are in the minds/brains and hence the exploration of necessary or possible relations between concepts to be merely psychology (and tell us nothing about the world)?
Or to ask otherwise why would the issue of relation between the mind and the world fall outside of this contemplation on relations between concepts (and talk about conceivability and necessity of those relations), when "mind", "world", and "concepts" after all are all concepts?

So I think that the conclusion that metaphysics is psychology is made within the specific metaphysical view (want it or not) in which certain relation between mind and world is assumed.
Anyway most people today when talking about "concepts" start from the assumption that those (concepts) are some things in the mind. Because of this, I tend to use the term "notion" instead of "concept" on my blog some times, just to avoid this unnoticed metaphysical/theoretical import which is common today.

Unknown said...

I'm sympathetic with both with Chalmer's worries about type-B materialism and with your suggestion that metaphysics is beholden to psychology. However, I don't think Chalmers would agree the second and I don't think it's that straightforward. Chalmers has developed a philosophical system in which semantics, epistemology, and metaphysics are all connected. Most fundamentally, he takes epistemic rationality as a primitive and as involving (I think) a non-naturalizable type of normativity. If this is correct, then rational reflection (rather than empirical investigation) is how we come to understand the intensions of our concepts. Once we understand the intensions of our concepts, then we are in a positiona to answer metaphysical questions. Hence, we can answer metaphysical questions from the armchair provided that we are rational and possess the relevant concepts. I want to get off board early on, by rejecting non-naturalizable normativity. But lots of philosophers accept such normativity (often implicitly, I suspect).

Richard Y Chappell said...

"Metaphysics seems then to become a branch of psychology."

That seems questionable. The modal rationalist doesn't care about the contingent psychological makeup of human beings. Rather, metaphysics is taken to be tied to the rational domain (as Mike points out), and hence concerns idealizations that are not contingent at all.

There may be a close tie between metaphysics and the psychology of ideally rational agents. But the latter is not an empirical topic. (No such agents actually exist, after all!) It's normative instead: closer to logic than psychology.

Richard Y Chappell said...

Oops, I forgot to address the central concern:

"If Chalmers’s (and all of our) “metaphysics” is just exploration of our concepts, how can we claim to discover anything about the fundamental structure of the universe thereby -- anything about anything other than our concepts?"

As per my previous comment, we should bear in mind that the metaphysician is still studying a non-contingent domain. They're interested in the logical space of rationally coherent concepts, not just those that we happen to currently have a grasp of. It's coherent concepts in general that the metaphysician is interested in; the fact that they're our concepts, is merely an unfortunate precondition for being able to examine them!

Still, your pressing question remains, and can be restated: why think that the rational domain is any sort of guide to how reality is in itself?

I'm not sure how best to answer that. But a couple of options spring to mind:

Response #1: "Why not?" We're usually happy to dismiss radical skepticism, e.g. about the problem of induction. Is this so different? The burden seems to lie with the skeptic about rationality.

Alternatively, #2: go deflationist. On this view, metaphysical truths aren't really anything over and above the most rationally coherent or "ideal belief", in any case.

(That doesn't make it trivial, of course, since our actual concepts are often less than perfectly coherent!)

Anonymous said...

The whole project depends on there being a distinction between conceptual truths and empirical truths. However, there's this rumor going around that there is no such distinction.

Unknown said...

Pete suggests that there's no distinction between conceptual and empirical truths. Quinean naturalists agree with that, Chalmers does not. It is interesting, however, that his position on the distinction is neo-Carnapian. He thinks that we pragmatically decide upon framework within which there is a relatively clear distinction between conceptual and empirical truths.

This brings me to Richards responses. On the first point, it seems to me that general normative skepticism undermines arguments for inductive or other types of epistemological skepticism. On the second point, Chalmers himself takes a deflationary attitude towards many metaphysical disputes (e.g. the special composition question). And, as you point out, this doesn't mean that he thinks metaphysical questions are trivial, since they depend on idealized reasoning. Of course, the normative skeptic will find such idealized rationality mysterious.

Alejandro said...

Eric has nailed prefectly my major intuitive problem with Chalmers. Let me try to have a go at answering Richard's second comment.

Answer 1) strikes me as a non sequitur, although I am probably missing something. If metaphyisics is about exploring which fundamental concepts can cohere with which others and in which ways, then it cannot by itself say anything about which self-consistent set of concepts should we use to describe reality. This is not comparable to skepticism, but rather a simple consequence of this view of metaphysics. For example, Chalmers could (optimistically) show that our ordinary concept of "consiousness" is logically inconsistent with physicalism, but this does not amount to much. The physicalist could say "well, let's use instead a different concept of consiousness which is not".

Dualists see this as eliminating consciousness altogether, but it is not: it is describing the same undeniable phenomenon in a different way. Things do not force upon us ways of describing them. This brings to answer 2), deflationism, which I sympathize with but which I regard as fatal to dualism, because it implies a certain amount of pragmatism, and pragmatically most scientists would be much happier with a change in our concept of consciousness than with admitting it as a fundamental and irreducible one akin to space and time, as Chalmers proposes.

Brad C said...

Hi Eric,

I have not had a chance to read much Chalmers yet but...

(1) I am a bit puzzled by the suggestion (at the end of the post) that materialism can be understood as a scientific thesis. You seem (earlier in the post) to associate scientific research with empirical exploration; I am unclear how we could empirically test (or otherwise "explore") the truth of materialism.

(2) You ask what understanding our conceptual scheme tells us about "the fundamental structure of the universe". First, our armchair investigation might, in Kantian fashion, tell us something about the concepts we must deploy in order to, e.g., have experience, make empirical judgments, have empirical knowledge, or act. I am not sure these types of arguments will get much, but it is a live option (I think).

Second, your claim that metaphysics is a branch of psychology seems contentious because it implies that it is a branch of *empirical* psychology - and that is not a claim you are making (right?) and Chalmers would reject it (I guess). Why not just stick with "elucidating and cleaning up our conceptual scheme"?

Eric Schwitzgebel said...

Hi, everybody -- what a lot of great comments! I'll get to them soon!

Eric Schwitzgebel said...

Interesting challange, Tanasije. Since I don't know what metaphysics is -- or to the extent I do, I think it's either an incoherent enterprise or one solely devoted to spelling out the relationships between psychologically explorable entities (our concepts) -- I wouldn't describe my own position as metaphysical; but I agree that someone else more sympathetic to substantive metaphysics might describe it that way.

I guess I do think of concepts as something mental. And as I see it the fact that "mind" and "world" and "concepts" are concepts doesn't mean that we discover substantive truths about the mind and world by reflecting on those concepts -- any more than the fact that "gravity" and "apple" are concepts means that we discover substantive truths about gravity and apples by reflecting on those concepts, except of course to the extent our concepts reflect substantive truths already acquired in different ways.

Does that make sense? What do you think?

Eric Schwitzgebel said...

Mike, what you say about Chalmers sounds like it might be right. I'm not quite sure how to read him on these issues. But the question in my mind is: What are "metaphysical" questions? If they're merely the exploration of our concepts, as Chalmers seems to say, I don't see how they get us to new substantive truths about the structure of the universe; but he does seem to think that they do. So that's where I'm puzzled.

Chalmers is a very smart guy and top-notch philosopher. I'm sure he's thought about this -- maybe even written about it somewhere that I haven't seen. I bet he has something smart to say about it! He'll be spending a few days at UCR in April; maybe I can push him on these issues a bit at that time.

Eric Schwitzgebel said...

Thanks, Richard, for your very well-articulated thoughts on the question! You say:

"It's coherent concepts in general that the metaphysician is interested in; the fact that they're our concepts, is merely an unfortunate precondition for being able to examine them!"

I think that's a very good and interesting point. What follows from it, in my view, is that to the extent metaphysics is substantive at all, it's substantive in the way math and logic are substantive. And what way is that? All three reveal patterns of relationship in axiomatic structures, considered very abstractly. There is, in my view, no one right and true logic or arithmetic or metaphysics of identity. In that sense, they are mere ungrounded structures in the air. It's only looking empirically at the world that we see a Euclidean geometry as approximately right for objects on the surface of the earth, and a two-valued or a fuzzy logic as the best tool to model a particular domain, or some particular sense of "person" as the most useful conception in accord with our values.

This view is, I take it, approximately Carnapian.

So yes, Pete, I accept a roughly Carnapian division between analytic and synthetic. I think our everyday concepts are not cleanly divisible into the two categories, but I see no reason why we can't keep analytic and synthetic truths more cleanly separable through the artifices of philosophy. I take this view to be in harmony with some of Chalmers' recent "neo-Carnapian" remarks, as Mike points out.

And Alejandro, I take this to be roughly in the spirit of your reply, too -- though I do not regard it as necessarily bad news for dualism. It may turn out to be useful to have concepts of consciousness that differ from our concepts of material things, even if there is no empirical divergence. But in my view, that's a very pallid dualism. And even empirically, while the evidence in favor of materialism seems to be very good right now, such things as near-death experiences and paranormal phenomena imply that the case is not quite closed (as I discussed in a post a couple months ago).

Eric Schwitzgebel said...

And finally, Brad -- thanks for your usual insightful and incisive comments! The concluding paragraph of my previous comment perhaps addresses your first point.

On your second, Kantian, point: I'm kind of inclined to think that armchair Kantianism doesn't get us much, since there are so many alternative possible axiomatic structures. Now it may quite well be that we have to experience objects as colored, events as in three-dimensions of space and one of time, etc. -- but if so, we learn that fact about the structure of our minds at least as much by looking at the world (including by looking at what people with brain damage say) as by reflecting in the armchair.

On your third and final point: Perhaps it was overstatement to say that metaphysics is a branch of psychology. "Elucidating and cleaning up our conceptual scheme" is a more cautious way of putting it. And yet, to the extent we are dealing with our actual, everyday concepts, as opposed to merely possible concepts -- and philosophers haven't been very good at separating these types of inquiries (though Chalmers is maybe better than most) -- "metaphysics" does start to converge, I think, on psychology.

Anonymous said...

Thanks for this interesting discussion (which somehow I missed until now). My first reaction to Eric's post coincides with what Richard says in comment #3: on my view, modality is constitutively tied to ideal rationality, not to psychology. But this still leaves the question "If metaphysics is just about exploration of our concepts, how can it tell us anything about the fundamental structure of the universe". I take it this question applies to ideal rationality as well as psychology. Here, it's crucial to note that I don't say that metaphysics as a whole is tied to rationality -- just that modality is. Some metaphysical questions concern modality, but a lot don't.

There's a close connection between modal questions and questions about the fundamental structure of reality. One such connection is that fundamental truths (i.e. a description of the fundamental structure of reality) must necessitate all the truths. So if one can make the case that a certain set of truths don't necessitate all the truths, that shows that that set of truths doesn't exhaust the fundamental truths. Of course that is at the core of the argument against physicalism: microphysical truths don't necessitate all the truths, so they don't exhaust the fundamental truths. So modal analysis can help us in the adjudication of e.g. what is more fundamental than what.

But of course there's a lot more to metaphysics than this. No amount of a priori analysis of this sort will yield any a posteriori truths -- whether physical truths, phenomenal truths, or whatever. I wouldn't suggest that one should use a priori methods to settle truths of this sort. I'm pretty sure that most of the fundamental metaphysical truths about the actual world (whether microphysical, phenomenal, or whatever) will be a posteriori, and their investigation will require the sort of a posteriori methods we find in physics, phenomenology, and other empirical fields. But given the class of a posteriori truths (microphysical truth, phenomenal, and so on), a priori methods can help us settle the question of which among them is fundamental. I don't see a problem with the idea that this sort of question should be constitutively connected to the rational domain. And of course this is just the sort of question that's at the heart of the issue about physicalism.

A couple of people mentioned my neo-Carnapian deflationism about ontology. I think this is somewhat separable from the issue above. For example, this deflationism doesn't lead me to think that in general, the question of conceptual-vs-empirical (or a priori vs. a posteriori) truth is framework-relative. It's just that on my view, the truth (and apriority) of certain existence-involving statements is framework-relative -- e.g. statements about the existence of mereological sums. On this view, such statements can express different propositions (have different truth-conditions) when uttered in contexts where different frameworks are operative. However, this view is consistent with an underlying nondeflationary and nonrelativistic view of the truth and apriority of propositions, and of their modal status.

Maybe I'll give my paper on ontological anti-realism when I visit Riverside in April...

Eric Schwitzgebel said...

Thanks for your thoughtful and detailed reply, Dave!

First, one minor point: I'm not sure what metaphysical inquiries you have in mind that don't concern modality. Aren't metaphysical claims in general supposed to hold with some modal force stronger than natural necessity? And if an inquiry yields a judgment construed as having that type of modal force, doesn't it concern modality?

On the main point, my basic confusion/disagreement remains: How can we get at "the fundamental structure of reality" by armchair inquiry? Here's my alternative take: When we reflect on a priori necessities, we're reflecting on the structure of our concepts. We may thereby discover what is more fundamental than what in a certain conceptual scheme. But whether that scheme matches up well with the universe is a separate question, and not one we can answer by a priori methods.

So I'll grant you this: In our conceptual scheme, phenomenological properties are distinct from physical properties. I think your work brings this out nicely. But I know you want to argue for more than that -- you believe that conceptual distinction between phenomenologial and physical reflects a real distinction in the fabric of the universe, and not just in the fabric of our conceptual scheme. And that's what I can't see discovering through a priori arguments of the sort you develop.

I think this point is at the heart of many of the criticisms of your work -- but most critics can't develop the point entirely convincingly and self-consistently because they don't reject, in general, the existence of subtantive (i.e., not merely conceptual) metaphysical truths -- so they find themselves committed, as you point out in your replies to them, to awkward notions of strong metaphysical necessity and the like.

The arguments are in The Conscious Mind are so tight, I think, that the only plausible way to reject them is to reject the basic conception of philosophy on which they're built; and few philosophers are willing to do that. It means going back to something startlingly close to Carnap -- to, that is, a very deflationary view of metaphysics.

Tristan Haze said...

'We may ... discover what is more fundamental than what in a certain conceptual scheme. But whether that scheme matches up well with the universe is a separate question, and not one we can answer by a priori methods.'

This resonates very much with my view on the necessary aposteriori. It seems more and more philosophers are thinking along this sort of line.

'I think this point is at the heart of many of the criticisms of your work -- but most critics can't develop the point entirely convincingly and self-consistently because they don't reject, in general, the existence of substantive (i.e., not merely conceptual) metaphysical truth...'

I'm actually trying to do something like this now. Not a criticism of Chalmers, but a positive contribution. Look out for a book on modality which will appear in 2-3 years (health permitting!).

In the more immediate future I will post some stuff on this topic on In fact, the next post will be about the meaning of "meanings ain't in the head", which is obviously an intimately related issue.

Eric Schwitzgebel said...

Very cool, Tristan! Keep me posted.

Tristan Haze said...

'Keep me posted.'

To this end: Blog post, Three True Semantic Externalisms.