Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

[Link] The Practical Argument for Free Will

1 Post author: entirelyuseless 04 March 2017 04:58PM

Comments (39)

Comment author: entirelyuseless 04 March 2017 05:00:03PM 1 point [-]

Discussing the bad argument that you should believe that determinism is false, because if determinism is true, you can't determine whether or not you believe in determinism. So you should assume it is false, since only in that case can you do yourself any benefit or any harm.

Among other things I note that people are convinced by the bad argument for the same reason they are convinced to smoke in the Smoking Lesion.

Comment author: MrMind 06 March 2017 02:20:22PM 0 points [-]

Another possible rebuttal, even retaining the analytical meaning of the word choose, is to complete the demonstration appealing to the "if I can" in point 5 with:

6a - since you cannot choose free will, then it doesn't exist.

So, for the argument to be complete, a believer would have to add:

6b - I choose to believe in free will.

But then you have to show that you actually chose conclusion 6b instead of simply having reached it, which I don't think can be done.

Comment author: entirelyuseless 06 March 2017 03:13:41PM 0 points [-]

Richard uses "choose" in this way, but I don't. I don't think choosing means you necessarily have libertarian free will.

In any case I am not sure this is a great rebuttal, since someone can say, "maybe it's choosing, maybe it's not. But at least I am going to try to choose."

Comment author: MrMind 06 March 2017 04:57:47PM 0 points [-]

I don't think choosing means you necessarily have libertarian free will.

Neither do I, but since it's just words, we can be gracious and grant the usage in the limited context of the philosophical discussion. Although your formulation is way clearer.

"maybe it's choosing, maybe it's not. But at least I am going to try to choose."

Well, anyone can do as anyone pleases, but the argument rested on choosing, not reaching the conclusion, so: if it is, show it, if it's not, the argument is invalid.

Comment author: entirelyuseless 07 March 2017 03:10:35AM 0 points [-]

It is a practical argument. It is not meant to prove that people have libertarian free will. It is not even meant to show that it is probable. It is meant to show that "trying to convince yourself to believe that you have libertarian free will is a good idea."

That is why I brought up the badness of being wrong about that, because it is an argument about what is good to do, not an argument about what is true. This response applies to your other comment as well.

Comment author: MrMind 07 March 2017 08:19:47AM 0 points [-]

It is meant to show that "trying to convince yourself to believe that you have libertarian free will is a good idea."

I feel that the crux of the argument shows that choosing (in the analytical sense) libertarian free will is a good idea, but only if you can. It lacks the part where also reaching the conclusion of libertarian free will is a good idea.

Comment author: entirelyuseless 07 March 2017 02:15:00PM 0 points [-]

It doesn't need to say that reaching the conclusion is good. It just has to show that trying as hard as you can to reach that conclusion is good. The argument is meant to be that making that an attempt has a potential upside (succeeding when you actually have libertarian free will) and no potential downside (because if you succeed and the conclusion is false, it is not your fault.)

That's wrong because not being your fault does not mean it is not a downside, just as it is a downside of smoking in the smoking lesion, if you get cancer, even if it is not your fault.

Comment author: MrMind 08 March 2017 08:55:13AM 0 points [-]

and no potential downside (because if you succeed and the conclusion is false, it is not your fault.)

That is the part of the argument that is missing from the original formulation, and assuming it I think does a disservice to your analysis and the original argument too.

Comment author: entirelyuseless 08 March 2017 02:10:34PM 0 points [-]

It certainly does not do a disservice to the original argument, since it is the only way it would ever convince someone.

That said, obviously I disagree with that, since I think you should not smoke in the Smoking Lesion case.

Comment author: g_pepper 07 March 2017 03:11:58PM 0 points [-]

I think that the argument is not so much that if you succeed in incorrectly convincing yourself that you have (libertarian) free will, it is not your fault. Instead, I think the argument is that success in willfully convincing yourself that you have free will (or convincing yourself of anything else, for that matter) implies that you have free will. If you didn't have free will, then you did not really willfully convince yourself of anything - instead, your belief (or lack thereof) in free will is just something that happened.

Comment author: entirelyuseless 08 March 2017 02:11:46PM 0 points [-]

Sure, but the question is why you should try to convince yourself of libertarian free will, instead of trying to convince yourself of the opposite. If you succeed in the first case, it shows you are right, but if you succeed in the second, it shows you are wrong.

Comment author: g_pepper 08 March 2017 02:34:17PM 0 points [-]

Sure, but the question is why you should try to convince yourself of libertarian free will, instead of trying to convince yourself of the opposite.

It seems like you answered the question yourself when you said:

If you succeed in the first case, it shows you are right, but if you succeed in the second, it shows you are wrong.

Surely it is better to be right than wrong, right?

Comment author: MrMind 06 March 2017 02:11:10PM 0 points [-]

Although I upvoted the article for your analysis, I don't feel that the counter-argument based on goodness is the strongest one can make.
Imagine that it is indeed good to believe in the correct metaphysical setup: still, you cannot derive that you should practically believe in libertarian free will, unless you have a way to distinguish from the inside how it feels to have libertarian free will from having deterministic free will, something that is probably impossible, or that even when you have deterministic free will is better to believe that you have libertarian free will.

Comment author: entirelyuseless 06 March 2017 03:11:49PM 0 points [-]

If I understand you correctly, you are saying that it does not follow that you should believe in libertarian free will, unless you can tell by looking at yourself that you in fact have libertarian free will.

I agree. But that is because the argument does not work. The argument would say, "If I try to believe that I have libertarian free will, I might get some benefit, just in the case where I actually have it. But I cannot cause any harm by trying to believe that I have libertarian free will, even if I don't have it, because if I don't, I am not causing myself to believe this, but the initial conditions are causing me to believe it."

This is why I posted it in the context of the Smoking Lesion, where the argument against EDT is, "If I choose to smoke, I might get some benefit, namely smoking. But I cannot cause any harm by smoking, even if my choice is caused by the lesion, because if the lesion makes me smoke, I am not causing myself to smoke, but the initial conditions are causing me to smoke."

Comment author: MrMind 06 March 2017 05:00:04PM 0 points [-]

But I cannot cause any harm by trying to believe that I have libertarian free will, even if I don't have it, because if I don't, I am not causing myself to believe this, but the initial conditions are causing me to believe it."

True, but in that case you're not showing that your belief is justified, and the argument becomes just stating a preference.