« Infinite Regresses: Vicious and Benign | Main | World Philosophy Day »

Wednesday, November 19, 2008

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

You say: "A vicious infinite regress is an infinite regress that arises in the context of an attempted philosophical explanation when the explanation given permits the question that was to be answered to arise at successively higher levels ad infinitum."

Perhaps you're right, but please notice that the view is controversial. Specifically, your take on viciousness renders all infinite explanatory chains (in philosophy) unacceptable. I should say that I entertained a very similar conception of viciousness until I read Klein (2003).

Are you familiar with the Gillett/ Klein controversy? Gillett argues that IV Regresses (IV=in virtue of) are vicious. Klein, in response, denies that there's something wrong with infinite explanatory chains per se.

See esp. Klein (2003: 729): "Put another way, Gillett is correct that 'the question consequently arises how it could ever come to pass that any member of the chain has the property H?' But the IV Regress is not designed to answer that question. It is designed to answer the quite different question 'How does it come to pass that each member of the chain - taken individually - has property H?'"

So, now the question is: what's so bad about infinite explanatory chains?

At the end you say: "The regress is vicious because it sets an impossible standard of explanation: if to explain an event one must explain every event in its causal ancestry, then no event can be explained."

I very much agree that we should look for "standards of explanation", but perhaps the "then" in your last sentence is a bit too quick.

Jan Willem,

Thank you very much for the excellent comments. I can see that you have gone deeper into this central metaphilosophical topic than I have. You say, "Specifically, your take on viciousness renders all infinite explanatory chains (in philosophy) unacceptable." You are right! And therefore I need to think harder about how to characterize viciousness. For I am not sure I want to say that ALL infinite explanatory chains (in philosophy) are unacceptable.

I haven't read the Gillett and Klein articles, but now I see that I must. To the library!

Perhaps you could help me by giving me an example of an infinite explanatory chain that is acceptable.

I wrote, ". . . if to explain an event one must explain every event in its causal ancestry, then no event can be explained." This seems right to me assuming that there are infinitely many events in the explanandum event's causal ancestry. For surely it is impossible for any given finite person to explain infinitely many events. Depending on how we construe ecents, there might be 2-to-the-aleph-nought of them.

Or am I missing something?

Jan Willem sent by e-mail these references:

Gillett, C. 2003. Infinitism Redux? A Response to Klein. Philosophy and Phenomenological Research 66: 709-17.
Klein, P. 2003. When Infinite Regresses Are Not Vicious. Philosophy and Phenomenological Research 66: 718-29.

"Perhaps you could help me by giving me an example of an infinite explanatory chain that is acceptable."

Well, the problem is that chains turn out acceptable or not depending on the criteria we use. It's not, I take it, the other way around: that we somehow intuitively know which chains are acceptable and which not, and then adjust our criteria which support exactly the desired outcomes.

Let me explain this in terms of your example. You say that "To explain an event one must explain its causes" generates a infinite regress which is unacceptable. Klein (2003: 728-9) holds the very same regress to be harmless. What's going on?

I think it's a disagreement about standards of explanation. By your standards, explanations are to be executed by a person in a finite amount of time (recap your "it is impossible for any given finite person"). Klein, however, doesn't talk about this; he just holds that if you're looking for an explanation for whatever event in the regress, you can find your answer.

So, my only point was that your conclusions do not follow without those standards. Perhaps the thing to do is to make explicit all standards of explanation, and argue like this: "assuming standards X, explanatory (super)tasks Y are impossible." I should admit that I didn't do anything like this yet. But I think it's the way to go.

Let's briefly return to B's regress. In the regress, the explanation of why R relates depends upon the explanation of why EX1 relates, of why EX2 relates, of why EX3 relates, and so on. One may argue, as I'm inclined to do, that this is a sufficient explanation of why R relates. So, the claim (or standard) is that we don't need to go all the way down to finish our explanation. Suppose you ask me why EX548890 relates, then I have a general rule which provides you the explanation: EXn relates in virtue of EXn+1.

Now if you happen to disagree with me, then I'd say you disagree with my standards of explanation. What do you think? (Sorry for this indirect answer to your question about the example.)

>>Klein, however, doesn't talk about this; he just holds that if you're looking for an explanation for whatever event in the regress, you can find your answer.<< This is unclear. Suppose there is a crop failure and we want to explain it. A possible explanation is that spring came late and there wasn't enough rain. Now one could stop right there and have a decent explanation of the crop failure: one would not have to explain why there was not enough rain, etc. etc. If this is all Klein is saying, then I agree.

But the theory I was examining what that to explain any event, one has to explain all its causes. How could that not entail a vicious infinite regress? That every event has a cause does not entail a vicious infinite regress: the regress it entails is benign. But explanation, even causal explanation, is different from causation.

I'll have to read Klein's article.

Back to Bradley. For you, the fact Rab is infinitely complex, and the actual infinity of exemplification relations is what cements all the constituents together. But what is the difference between the set or sum of constituents, on the one hand, and the fact, on the other? Can't the constituents exist without the fact existing? I would say yes whether the constituents have a finite or an infinite cardinality. So, as I see it, the mere fact that there is an actual infinity of exemplification relations in Rab does not suffice to unify Rab's constituents. The ontological 'gluer' or unifier cannot be a constituent or an infinite series of constituents.

I think it is important to realize that Bradley's regress is properly understood only within the context of the problem of the unity of a complex, no matter what sort of complex we are talking about. A complex is one, and yet it is many, and the problem is to understand how such a structure is possible.

"But the theory I was examining what that to explain any event, one has to explain all its causes."

Your point is very important, and I'm not sure you will find a clear answer in Klein (2003). The only thing he says is that your "one has to explain all its causes" should be read as "one has to explain all its causes taken individually (one by one)" and not "one has to explain all its causes taken together (as a whole chain)".

I have to think more about all this. As to Bradley, I should say that I'm not happy with binding relations, i.e. posits that are supposed to exist for the sole reason of gluing things together. But I don't think my personal views matter here; my interests lie in the question whether we have conclusive reasons (and which) for holding the regress to be vicious. So, it's interesting to see that Francesco puts forward a completely different criterion!

As said, I have to rethink our discussion and all the points and worries you have raised & look foward to return to it!

Jan,

I wonder if you would counternance all five of these epistemic possibilities with respect to the Bradley regress:

1. There is no regress at all. The business of a relation is to relate, and in the fact Rab, R relates a and b. And that's it! Blanshard's view.

2. There is a one-step regress: 'Rab' does not perspicuously display the fact; 'EXRab' does. But there is no regress of exemplification relations.

3. There is an infinite regress, but it is benign and non-explanatory. This would be Orilia's external regress.

4. There is an infinite regress which is both benign and explanatory. This would be your internal regress. But this commits you to actual infinities.

5. There is an infinite regress, it is internal to the fact, and it is vicious. This is my view.

Bill,
Here are my responses:

1. There is no regress at all. The business of a relation is to relate, and in the fact Rab, R relates a and b. And that's it! Blanshard's view.

> OK, if it's R's business to relate, we get the unity.

2. There is a one-step regress: 'Rab' does not perspicuously display the fact; 'EXRab' does. But there is no regress of exemplification relations.

> OK, if it's EX's business to relate, we get the unity.

3. There is an infinite regress, but it is benign and non-explanatory. This would be Orilia's external regress.

> I'm not sure Francesco would be happy with a non-explanatory regress. Anyway, if all those facts are harmless consequences of aRb, rather than items that contribute to the unity of aRb, we shouldn't bother too much about them.

4. There is an infinite regress which is both benign and explanatory. This would be your internal regress. But this commits you to actual infinities.

> OK. I think the regress is explanatory from a non-temporal perspective, that is, if we don't have to go all the way down in a finite amount of time. (As said, I'm working on this.)

5. There is an infinite regress, it is internal to the fact, and it is vicious. This is my view.

> OK. The regress is non-explanatory from a temporal perspective.

#

So, explanatory power doesn't seem to be enough to settle the matter. We need more criteria of theory choice, so that we can choose among explanatory succesful options. Or perhaps we have to put more restrictions on the criterion of explanatory power and/or on what is to be explained.

How would you proceed?

> How would you proceed?

PS: I mean in this methodological matter. I've read you ground-breaking papers, of course, and am familiar with your own solution to the unity problem. But now I wonder what you would do if different theories can solve the very same problem.

The comments to this entry are closed.

My Photo
Blog powered by Typepad
Member since 10/2008

Categories

Categories

March 2021

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31      
Blog powered by Typepad