Thursday, May 13, 2004

Dennett has an interesting, but invalid, argument for the compatibility of freedom & determinism. But I think I've come up with a modification which solves the problem.

Dennett's argument is that no-one knows whether determinism is true, but (according to incompatibalists) determinism implies not-freedom, which then implies not-morally-responsible. So, Dennet argues, the incompatibalist must think we never know whether people are morally responsible. But obviously we do, so the incompatibalist is wrong.

Despite it's initial plausibility, this argument is actually invalid, as our lecturer pointed out. if P implies Q, you CAN know Q without knowing P. For example, "If it rains tomorrow then 2+2=4" is true, but just because I don't know whether it rains tomorrow, does not mean that I don't know 2+2=4.

So it is logically possible to know whether we are morally responsible WITHOUT knowing whether determinism is true. So Dennett's argument falls apart.

But let's look at the logic more closely: If P implies Q, you can know Q without knowing P. That was Dennett's mistake. BUT, we can safely assert the reverse: If you know P implies Q, and you know P, then you DO also know Q. (This 'closure principle' is an important epistemologial principle. See my post on Nozick for more detail).

This suggests Dennett's argument might work if we run it in reverse:

My Argument:
1. No-one knows whether determinism is true.
2. We can (sometimes) know that S is a morally responsible agent.
3. If S is morally responsible then S is free.
4. (Incompatibalist premise) If S is free then determinism is false.
5. Therefore, we can (sometimes) know that determinism is false

I think the logic here is valid. If you can see any flaws, please comment & let me know!

These 5 statements are clearly inconsistent. But 1,2,3 all seem true, and 5 follows from 2,3 & 4 (via the closure principle, i.e. if X knows P, and knows that P implies Q, then X knows Q). Our best option seems to be to deny 4: That is, we should conclude that the incompatibalist is wrong.

1 comment:

Nice argument - it seems good to me anyway. What about denying 3 rather than 4? I find that the idea that moral responsibility implies freedom doesn't quite sit right with me. How is moral responsibility defined? Because I'm not sure and I think that if I had a clearer picture I might be convinced to accept 3 (and hence be forced to reject 4).
Patrick Kerr | Email | 13th May 04 - 8:41 pm | #

-------------

I dunno, 3 seems pretty solid to me. I'm not sure how to define moral responsibility at this stage (I'm more just using the intuitive idea we all have about it).

You may find 3 more convincing if expressed as its contrapositive (which is logically equivalent):

3') If S is not free, then S is not morally responsible.

That strikes me as pretty axiomatic. Whatever freedom is, we tend to regard it as necessary for moral responsibility. I don't have any further argument to back this up though, so if your intuitions differ from mine, you could reasonably deny this premise.

But I think the libertarian simply has to bite the bullet and deny either #1 (saying "I really do know that determinism is false!" - from what Dusan said in today's tutorial, it sounds like Kane might believe this) or #2.

If I had to choose, I'd probably deny #2, and say "well, yeah, I guess we really can't ever be sure if someone is truly responsible or not, since it might all be pre-determined".

But most people presumably wouldn't want to be forced into such a corner, so rejecting incompatibalism (#4) does seem the best option!
Richard | 13th May 04 - 11:22 pm | #

----------------

The reason I was wondering about moral responsibility is this: I understand it as being some kind of absolute term... in a way removed from the practicalities of when we actually do find someone responsible in real life. Ie the philosopher's normative account of responsibility as opposed to just summing the practicalities of responsibility up in a descriptive term. Knowing whether this understanding is right would allow me to decide whether i had any counterexamples up my sleeve.
Patrick Kerr | Email | 16th May 04 - 8:18 am | #

--------------

Update:
For the argument to be valid, actually requires that premises 3 & 4 be preceeded by "We know that:".