December 14, 2005

Post on Conditionals and Stuff, but mostly about putting up some content

Posting has been slow lately, but here's a link to a comment I left on CT explaining why it's ordinarily absurd to say "Whether or not P, I believe P," and why it wasn't so absurd for Bush to say it.

I haven't thought through the exact circumstances under which "If P or Q, then R" is equivalent to "If P then R, and if Q then R," which is one of the key steps in my explanation. I don't remember whether it comes out true in relevance logic. In a Lewisian closest-possible-world semantics for counterfactuals I don't think it comes out true; if the P-worlds are closer than the Q-worlds, and R is true in the closest P-world but not in the closest Q-world, then "If P or Q then R" comes out true but "If Q then R" comes out false. Other counterfactual semantics, I don't know. There may not be any semantics on which this equivalence holds and on which "If not-P then P" is absurd (for material implication, the equivalence holds, but "If not-P then P" is equivalent to P and so not absurd).

But even if that is the case, I think I'd still stand behind the analysis. Though I'd be prepared to run away from it at the slightest hint of opposition.

Posted by Matt Weiner at December 14, 2005 02:02 PM

Reading the interview, I totally agree that there were two senses of 'needed' in play. But that aside, I'm unhappy with the analysis, though not for the reasons discussed in your posts. I think Bush's original quote is fine, even interpreting 'needed' uniformily throughout, and that the problem is introduced by 'raising' the 'I believe' operator. Bush says:

“Whether or not it [the war] needed to happen, I’m still convinced it needed to happen.”

This seems just a way of saying something like:

'I don't know whether the war needed to happen or not, but I still believe that it needed to happen.'

But that's perfectly assertible, surely. There's nothing paradoxical (even in a loose, Mooreanesque sense) with asserting that one doesn't know whether something is the case, but that one still believes that it is. But on the analysis, Bush says:

'I believe: whether or not the war needed to happen, it needed to happen.'

And it looks like that runs into trouble, for just the reasons you give. But generalizing the Cheney argument gets us into parallel difficulties with simpler cases (i.e. cases not raising the questions about conditionals you discuss in the posts). I say to someone whose credibility has been totally undermined in my eyes;

'Even if you've told me the truth, I believe you have lied.'

I didn't attribute an absurd belief to myself, as the Cheney move would predict:

'I believe: even if you've told me the truth, you have lied.' (Note: lied on this very occasion, not just at some time.)

Posted by: Aidan at December 14, 2005 09:36 PM

Quoting me:

"This seems just a way of saying something like:

'I don't know whether the war needed to happen or not, but I still believe that it needed to happen.'"

I should just say to be clear, I just mean that Bush's sentence is conveying something like this, not that I've given anything like an analysis of any sort of Bush's original sentence.

Hmm, I never thought I'd be defending Bush on anything............

Posted by: Aidan at December 14, 2005 09:44 PM

There is in fact some textual evidence that Bush was using two different senses of 'needed to happen'. (The first commenter on the CT thread made that point.) Conversely, I'm a bit reluctant to interpret Bush as reporting his belief states because I'm not convinced he distinguishes his belief states from objective facts.

In most circumstances, it seems to me, the Cheney move does hold for "If P, I believe Q" or "Whether or not P, I believe Q"; maybe we're reluctant to raise the belief operator when P (or not-P in the second case) explicitly contradicts Q, because the attributed belief would be absurd.

There will be those who say that you can't actually believe Q while entertaining doubt about it, but I don't find that view convincing.

Posted by: Matt Weiner at December 15, 2005 08:34 AM

Note also this.

Posted by: Matt Weiner at December 15, 2005 08:35 AM

I really didn't mean to deny anywhere that the textual evidence suggested two senses of 'needed to happen'. I opened my first post expressing total agreement on that matter, and everything else I wrote, including the second post, was in the spirit of investigating 'the paradoxicality of the statement if you don't see it as involving two senses of 'need', as you put it on CT.

I didn't get the point of your response on the Cheney move; I don't see why our reluctance here shouldn't carry the day. A student in an epistemology class confronted by the sceptical hypothesis says:

'Whether there is an external world or there isn't, I believe there is an external world.'

These kinds of sentences don't strike me as absurd. As with my case above, the absurdity only gets a foot in the door when we raise the belief operator, and reluctance to ascribe absurd beliefs on the basis of assertions which have natural non-absurd readings hardly seems misplaced. What better evidence could we have regarding the applicability of the Cheney move than these bad results? That it works in most circumstances?

Posted by: Aidan at December 15, 2005 01:22 PM

I really didn't mean to deny anywhere that the textual evidence suggested two senses of 'needed to happen'.

Sorry, I didn't mean to suggest you'd denied it; it's just that the evidence was offsite in my CT comment. And when I made my comment I had completely forgotten about the opening of your first comment. Why? Because I'm stupid.

As for the Cheney move, I guess this is my view: A conditional of the form "If P, I believe Q" will almost always be used to express some belief that is actually held, often in a conditional. It's odd to conditionally ascribe a belief to yourself. (Perhaps it will be possible to construct a case where transparency of belief fails and it isn't odd. For instance, "If 'zoob' means philosopher, then I believe I am a zoob." Would that work?)

Then the thing about "Whether or not P, I believe P" is it does ascribe a belief unconditionally, but that's because the antecedent of the conditional is vacuous. Similarly with "Whether the Democrat or the Republican wins the next election, I believe the U.S. will change its foreign policy"; the antecedent isn't vacuous, but the two alternatives are taken to be practically exhaustive. So then raising the belief operator is only necessary when you have a non-vacuous conditional.

Still, I think (intuitively) there's something funny about the Bush sentence (with one sense of 'need') that there isn't about your 'external world' example. Perhaps it's that the existence of the external world could be taken to be something that one has a right to have a belief about even when the evidence doesn't settle it; whereas the necessity of the war isn't that way. Perhaps it's that being convinced is incompatible with entertaining doubts.

Posted by: Matt Weiner at December 15, 2005 03:23 PM

Bush's sentence is tricky because it's not clear to me how to capture the contribution that 'still' in 'I'm still convinced'. I'll need to think about this some more, and preferably after I've finished my paper due tomorrow.

Btw, you might have noticed already, but I thought it was ironic that after all the discussion generated by this quote, Jonathan Ichikawa pointed to one I think is even better from the very same page:

I think this one's a lost cause.

Posted by: Aidan at December 15, 2005 03:48 PM


You are arguing rationally about the contextual rationality of a statement by an attestor known to be most comfortable with irrational assertions where convenient or reassuring. That's irrational.

In a universe of discourse that has been dominated by the Chewbacca Defense, whereof the logic-based community cannot speak, thereof the logic-based community must be silent.

Posted by: Ben at December 15, 2005 04:29 PM

I prefer to think of it as "Bringing our sophisticated analytical tools to bear on a hypothetical statement, having just given up on our President and anyone who still thinks he makes sense." Or showing off our exceeding geekiness. Or -- see the post title -- "putting up some content on my fallow blog."

I'm almost prepared to say that "I think we are welcomed. But it was not a peaceful welcome" could be read as "We are in general welcomed, but some few people are causing violence, so the welcome is not peaceful." Then I realize that, given how popular the occupation actually is, that would be even stupider than the stupid-word-game interpretation. And yet I still think it's what Bush meant.

Posted by: Matt Weiner at December 15, 2005 04:49 PM