Monday, June 12, 2006

Formal Proof of 'ought' implies 'can'

I have been looking into this problem since posting on ephphatha's blog philochristos.
I must admit I have found nothing satisfactory so far. Deontic logic seems beset with problems. Can anybody help?

8 comments:

ephphatha said...

Did you see this?

Psiomniac said...

Ephphatha,
Thanks, I hadn't spotted that one. I have read it all and find it plausible. Now if only you can solve all the problems you raise!

ephphatha said...

Well, I've got my daughter this weekend, so I can't critique the whole article, but I'll give you one short thought on it.

His argument uses the transitive property:

If a then b
if b then c
therefore if a then c

1. If an agent has an obligation to @ then an agent has a reason to @
2. If an agent has a reason to @ then @ is a potential action of the agent.
3. If @ is a potential action of the agent, then the agent can @
4. Therefore, if an agent has an obligation to @, then the agent can @.

This is a valid argument. The only problem I have with it is premise 2. Premise two IS the ought implies can principle, and sticking it in the argument causes the whole argument to beg the question.

Look at the first premise. I agree that if you have a moral obligation to do something, then you have a reason to do it. But the reason is THAT you have a moral obligation to do it. So it's a moral reason. That means "obligation" and "reason" mean exactly the same thing.

Since "obligation" and "reason" mean the same thing, the second premise can be substituted to say "If an agent has an obligation to do something, then it's possible for the agent to do it."

Notice that in the third premise, "potential action of the agent" and "the agent can do @" both mean exactly the same thing. It's another tautology just like the first premise. So, again, you can substitute in the second premise:

If an agent has an obligation to do something, then the agent can do it.

So really, the second premise IS the ought implies can principle. So the author hasn't proven anything. He's just assumed his conclusion in one of the premises.

Unfortunately, this author focuses more on objections to the ought implies can principle than he does defending his argument FOR the ought implies can principle.

Psiomniac said...

Thanks for that ephphatha, I see what you are getting at although I am not sure I agree yet. I need to think about it some more.

Psiomniac said...

Ephphatha,
I wondered what the author of the paper would think of your objection so I emailed it to him saying that I had encountered it online and he gave this response:

Thanks for your message and for your interest in my paper. There are several things to say in response, but I will limit myself to just one point. I don't think that "obligation" and "reason" mean the same thing. If I promise to help you, then I have both a reason to help you and an obligation to help you. But my reason to help you is not that I have an obligation to help you; it's rather that I have promised to help you. (And my obligation to help you derives from that reason.) So the second premise is not the same as the conclusion.

I think he has a point there.

ephphatha said...

psiomniac,

If his reason is not the same as his obligation, then he's got another problem to solve. His new problem is how to defend the first premise--If an agent has an obligation to do something, then the agent has a reason to do it. If the reason does not lie in the obligation, then how does it follow merely from having an obligation that you also have a reason?

And even if I grant that the second premise is not the ought implies can principle, his second premise seems even less obvious than the ought implies can principle. The ought implies can principle has a strong appeal to common sense. Most people think it's intuitively obvious and doesn't require any proof. But his second premise (if it's not the ought implies can principle) seems far less obvious. How does it follow that because I have a resaon to do something that I therefore am able to do it? It seems to me that several counter-examples exist. Suppose I make a promise and, unforseen occurences prevent me from fulfilling it? If I make a promise, then I have a reason, but it doesn't follow that I'm able to do it.

Now the author may argue that in that case, I don't have an obligation either, but then he'd be begging the question again.

Psiomniac said...

ephphatha,
Good response. I had a think about this. On reflection, let obligation to @ be o@ and having a reason to @ be r@.
Then P1 becomes o@->r@.
Now if we consider the truth table for this, the above is false iff o@^¬r@ is true. So it is false that o@->r@ if we can have an obligation to @ and yet have no reason to @. This seems fair to me.
Following from that you say that P2 is less obvious than OIC itself. But that is why the author, acknowledging this explicitly, included counterexample 4.3. You criticized him for focusing on objections then posed some objections that he had addressed!
Can I take it that you think OIC is true anyway so we can have a cup of tea?

ephphatha said...

Oh yes, I fully subscribe to ought implies can, and I'd love a cup of tea.