Monday, October 16, 2006

Metaphysics, What?

Philosophers, I suppose, sometimes do metaphysics. No, let me put it more cautiously. Philosophers engage in certain practices, which they sometimes call "metaphysics". I can tell fairly well what sorts of practices will be labeled in this way -- e.g., much of David Lewis's work and the ensuing discussions, analytic philosophy of mind as driven by thought experiments, discussions of "personal identity". But is this really metaphysics? What the heck is metaphysics, anyway?

Here's one view. Let's call it the "mystical view" -- because really it is rather mystical, though many hard-nosed, atheistic philosophers seem implicitly (or even explicitly) to accept it. Metaphysics is the discovery, by a priori armchair reflection without depending upon anything empirical, of necessary truths of the universe -- truths such as that causes must precede effects, and that a functional duplicate of me must necessarily have (or will not necessarily have) conscious experience. Such facts are supposed hold true regardless of our concepts, to be independent of our (contingent) ways of thinking about things. We tap into them not by looking at the world but rather by... well, that's the mystical part. How, exactly, do we learn about the outside universe (not just our own minds) without looking at it? Those philosophers who have gamely tried to explain the process in question -- George Bealer and Laurence BonJour, for example -- have tied themselves in such knots, been forced to wave their hands at such absolutely crucial junctures, and if I may be frank have failed so utterly as to make the hopelessness of their project even more evident after having read them than one might have thought beforehand.

Here's another view of what's going on. Call this the "no metaphysics" view. What philosophers learn from their armchairs, without looking at the world, are facts not about remote possible worlds accessible in no other way, or facts about the deep metaphysical structure of the universe, but rather facts about their own minds -- facts, especially, about their concepts. What else would one learn about, sitting in one's armchair? We learn that our concept of "cause" is a concept involving the temporal priority of the cause to the effect, our concept of a person is thus-and-such, etc.

But of course learning about our concepts is learning not metaphysical truths in the sense that philosophers ordinarily mean the phrase but rather learning contingent empirical facts about how we think. The concepts so delivered may be revisable in the face of empirical evidence (see Friday's post). And furthermore, they are empirically, psychologically explorable: There's more than one way to learn about "our" concepts. Philosophers in the armchair might not be getting the story right, or they may be an unrepresentative sample.

The philosophical practices labelled "metaphysics", then, have two uses, as I see it, neither of which is the discovering of metaphysical truths: (1.) They provide a kind of evidence about how people (a certain type of people, with certain habits of reflection and standards of inquiry) happen to conceptualize things; and (2.) (more interestingly, to me) they provide recommendations about how we should conceptualize things. If construed in this way, such recommendations should be evaluated pragmatically, in terms of their usefulness in organizing our way of thinking about matters of concern to us.

Getting clear about the pragmatic standard of evaluation can, I think, help us sort through and evaluate competing "metaphysical" claims about personal identity, causation, and the like. So, for example, in my work on belief, which could easily be misconstrued as metaphysics, I advocate a broad dispositional approach as giving us the best tool for talking about and characterizing the kinds of case that interest me most in believing -- what I call the "in-between" cases of gradual learning and forgetting, self-deception, confusion, ambivalence, irrationality, and failure to think things through.

15 comments:

Tanasije Gjorgoski said...

Hi Eric,
If you think there is something weird about coming to truths by armchair reflection, then you should think mathematics. They (mathematicians) don't measure the sides of triangles in the world, to falsify the Pythagorean theorem. If they do say that Pythagorean theorem is valid only in the euclidean space, they provide argument for it, and accept each other arguments without being worried about empirical measurements.
BTW, some time ago, I wrote a post called 'Metaphysics Manifesto'. You might want to check it, as I tried to give account (basically Hegelian,as far as I can tell) having on mind objections like yours.

Brad C said...

Hi Eric,

I agree that the term 'metaphysics' is not that helpful. I always tell my students to think about its entymology - metaphysics is stuff like that which Aristotle thought should be taught after natural philosophy, and which is thus discussed in the book that comes after the Physics in his corpus.

I like your distinction between the two types of inquiry it might be used to pick out.

Question: You say that metaphysics 2 involves providing pragmatic recommendations about how we should conceptualize things. I want to ask two questions I put to J.D. Trout when he gave a talk here about epistemology and said (very roughly) that epistemic norms should be pragmatically anchored.

The first question is: What gounds the assessments of pragmatic value? What is (are) the end(s) against which judgments of usefulness are to be made?

This suggests the second question: Even if we take the pragmatic value line on concepts like belief, should (and can) we take it when investigating (non-epistemic) normative concepts - thinner ones like good for and useful for, and thicker ones like piety?

Eric Schwitzgebel said...

Thanks for your comments, tanasije and brad!

You're right, tanasije, that the view here needs to be connected to a view in philosophy of mathematics (and thanks for the link). My short answer here is that mathematical truths are truths about our concepts, and those concepts are only optionally deployed. For example, we have a choice between a Euclidean or a non-Euclidean geometrical structure, and we have a choice between a two-valued logic and a fuzzy logic, and we have a choice between a Boolean arithemetic and our ordinary arithmetic; and those choices are to be made pragmatically.

Your questions, too, are helpful, Brad. I'd suggest that pragmatic decisions are decisions of value in light of empirically discovered information about the world. To the question whether there's a "right" set of values, I charge circularity. Pace Kant, grounding value in something else is hopeless; but grounding a general value system in that very value system is circular. But this is no more than the usual skeptical paradox as it occurs in ethics -- a paradox that arises whenever there's a demand to ground something very broad and basic. (Well, that's a can of worms!)

On question two, let me simply say yes! Why not?

Tanasije Gjorgoski said...

Eric,
Where do those (mathematical or otherwise) concepts come from? I guess you are saying they come from our minds? That they are contingent boundaries we set to our experience?
Or?

Brad C said...

Hi Eric,

Here is my worry about the second question.

Say everyone agrees that pragmatic value is fixed by conduciveness to well-being. Some are subjectivists and others are objectivists about well-being. Some objectivists are hedonists, some are not. Etc. As you point out debates about which of these is right will draw on intuitions and empirical facts.

But now we ask about the concepts of morally right and piety. If we take the pragmatic line, then we say that we should adopt the conception of morally right and piety that is pragmatically justified.

But suppose someone else (e.g. Kant) claims that we should adopt the conception of morally right that best fits our intuitions about what is morally right, even if doing so leads us to be less well off. Adopting, and acting in accord with that conception, they tell us, may make us more worthy of being well-off even if it does not make us actually more well-off.

How do we back our appeal to pragmatic value as the ultimate court of appeal here?

kboughan said...

"I always tell my students to think about its *entymology*..."

Who knew metaphysics at root was all about insects? :) (Drop the nasal phoneme in the first syllable of that last word and the sentence will be somewhat less buggy.)

Eric Schwitzgebel said...

Ah, those metaphysical bugs!

Brad: You raise a good point. If I say we choose our concepts pragmatically, is there then no room for (knowingly) choosing concepts contrary to our interests? Indeed, I would like to be able to say that we can choose to mold our concepts in accord with the demands of morality even if those demands run contrary to our interests. This still fits my general picture: We use our values to guide decisions about how to develop our concepts, in light of empirical facts -- but you're right that the word "pragmatic" doesn't seem the best in such a chase.

Tanasije, yes, I would say the concepts come from our minds. Do you see anything problematic in that?

Clark Goble said...

I'm not sure mathematics is a good analogy. Peirce thought that mathematics was basically a world of pure possibility since it's based on rules we generate and what would be true if those were true. It seems to me, however, that metaphysics is much closer to science and empirical knowledge. The problem with metaphysics is that the evidence is usually extremely weak and there are few ways to adjudicate disagreements. But I personally (following Peirce largely) think it very useful. It's just that typically in metaphysics one has to put off the verification of what is discussed.

But certainly metaphysical claims can be falsified. Look at Newtonian materialism. There were lots of claims we'd have to call metaphysical claims. Yet, primarily because of empirical discoveries we reject these claims. We're still doing metaphysics.

I personally think the line between physics and metaphysics is pretty blurry.

Tanasije Gjorgoski said...

Eric,
Well one problem would be objectivity. Take for example Pythagorean theorem. There are no multiple theorems in each person's mind. There is just one Pythagorean theorem, of which the mathematicians talk.

Tanasije Gjorgoski said...

Hi Clark,

I agree that metaphysical system can be shown wrong by empirical results, but I don't think it makes the analogy with math wrong one, as simmilar case can happen in math.
As an example take Goldbach's conjecture (that every even number bigger than 2 can be written as sum of two primes), and imagine that somebody gives convoluted and delicate proof for it, which is regarded as true from the mathematics community for several years. But then, someone finds an even integer which can't be written as sum of two primes. That is enough for the theorem to be shown false and its proof invalid (even without pointing to a mistake in the proof). But because mathematical theorem can be shown invalid by example, doesn't mean that the mathematics is wrong in its approach, and should transform to empirical science about numbers. For example in such turn of events mathematicians would propose theories instead of conjectures, and then math-experimenters would do lot of testing if the theory is good for lot of different numbers. I guess you agree that would be silly.
In simmilar way, metaphysical reasoning can be shown invalid by empirical results. But that shows the concrete metaphysical reasoning wrong, not the approach, and it doesn't mean that metaphysics should change to empirical science.

Eric Schwitzgebel said...

Thanks for the comments, Tanasije and Clark!

Let me just add a one more thought:

In my view, metaphysical claims cannot be proven false by empirical results in the way you suggest. Either they should be construed as claims about our concepts -- claims which may be true even if our concepts are completely inapt -- or they (as I would prefer) should be construed as recommendations. And recommendations are astute or foolish, helpful or unhelpful, laudable or vile, but not true or false.

Clark Goble said...

Just coming back given this was linked to in the carnival. I don't think a mathematical example is really empirical in the normal sense of the term - although I guess that would depend if you are a constructivist and of what sort.

If someone makes a mathematical claim I still think it best to conceive of it as a claim of possibilities given some structural limits. I can find counter-examples meaning that a possibility is inconsistent with my structural requirements. But that's not really an empirical fact. Just a demonstration of logical inconsistency.

One can always simply change the rules. To a degree you find this in the different sorts of foundations for mathematics and how they approach infinities. But one could also take logic and say the different sorts of logic we have. I suppose the different kinds of geometry that developed back in the day when we moved by Cartesian geometry would count as well.

That's always been the lesson of Godel I've taken. I know he took it as a plea for Platonism but to me it's really a plea to constantly expand beyond our axioms.

Having said all this I'm still pretty sympathetic to quasi-empirical methods in math. But I think that "quasi-empirical" is more about methadology that the grounds of mathematics.

Eric Schwitzgebel said...

Thanks for your further thoughts, Clark. I'm not sure whether you mean to disagree with this, but my present inclination is to think that the choice of a mathematical-logical framework is an empirically governed choice. You might say there are structural facts about the mathematical-logical frameworks that are independent of the empirically-driven choice, but I'm not sure those aren't just empirical facts about our concepts or pragmatically governed decisions about structure (at a meta-level). Need there be some mind-independent, non-empirical fact about the world left over after all this?

Timothy Scriven said...

I'm just an amateur but couldn't large parts of metaphysics be construed to be purely lexicographic and hence not open to empirical revision. For instance no empirical discovery could reveal that we have the meaning of the word "hat" wrong. This would give an account of metaphysics ( or parts of it at any rate) similar to your suggested account about metaphysics being a study of our concepts, except with reduced possibility for empirical falsification.

I've always thought that large parts of metaphysics ( and philosophy in general) are just an exotic sort of dictionary writing ( one reason for the thought experiments in philosophy; you’ll note that when you try to define certain complex words you engage in similar forms of counterfactual reasoning) in this , is there a word for this view?

Eric Schwitzgebel said...

Even that's empirical, I'm inclined to think, Timothy: the empirical exploration of what our terms mean, of what our concept of "hat" is. After all, you have to do research to write a dictionary. Or, solipsistically, if you only care about your own concepts, you still need to look to the world to see if the conceptual structure you build fits nicely with any actual phenomena....