I’ve gotten some nice responses on my previous blog on Self-Deception and Moral Dilemmas. I argued there that self-deception in the context of a moral dilemma has morally negative consequences, because it undermines our ability to minimize damage on whichever side of the dilemma we break a moral requirement.

Two major questions arose in the comments. First, what’s the definition of self-deception? Second, was Lance Armstrong self-deceived in thinking he could beat cancer and come back to win the Tour? The idea the second question suggests is that self-deception might be good insofar as it can help people have a positive outlook that facilitates overcoming the odds. As one commentator, calling himself Anaxagoras, put it, “self-deception can be transformative, and . . . believing in the irrational is what gets us through our day jobs, our lonely nights, and our limitations.” Of course, the answer to the second question depends on the answer to the first, so I’ll handle them in order. (Warning: the definition stuff is technical; feel free to skip the next three paragraphs to get to the discussion of Lance, which can still be pretty well understood without reading the definition.)

Here’s the definition I’ve come up with.

An agent is in a state of self-deception if and only if she holds a belief (i) that is contrary to what her epistemic norms in conjunction with what evidence she has would usually dictate and (ii) a desire for a certain state of affairs to obtain, or to have a certain belief, causally makes the difference what belief she holds in an epistemically illegitimate fashion.

“Epistemically illegitimate fashion” in the second clause means here illegitimate according to the usual epistemic norms of the agent. I relativize in this fashion because it isn’t psychologically interesting when an agent believes contrary to epistemic norms she doesn’t actually have; it’s when the agent’s own better standards of belief are subverted that you get a bizarre and interesting epistemic upheaval. Of course, it can’t be just any desire that plays the role of the deceptive element; it must have a content relation to the resulting self-deceptive belief. So I add a content restriction to complete the definition.

Content Thesis: in order for the definition of self-deception to be satisfied, the first-order content of the desire that brings about the belief must be identical to the content of the belief or its negation, or the higher-order content of the desire must be the content of the belief.

I won’t parse my definition at length (email me if you want a more thorough discussion). But there’s one important thing to notice. It’s not necessary for a belief to be false in order for it to be part of a state of self-deception. The reason falsity is not required is that, from a psychological perspective, it’s possible to be in the same mental state as someone who’s self-deceived, even though your belief comes out true accidentally. Take the following case: an abused spouse in denial counts as being in a state of self-deception about whether her husband will beat her again even if he gets hit by a bus the next day and never does beat her. Her belief that he wouldn't was still self-deceptive, even though it turned out accidentally true.

This is a good place to turn to the second question. Was Lance Armstrong self-deceived? If you require falsehood of the belief as a necessary condition on being self-deceived (as philosophers like Al Mele and Robert Audi do), then he wasn’t. But the better question is, I think: was he in a state of self-deception? It’s perfectly possible for him to have been in a state of self-deception with respect to the belief that he could come back and win, even though that belief ultimately was true. (Seven times!)

Let’s distinguish a few questions for the sake of precision.

Alpha: Was Lance in a state of self-deception in believing he could come back (and win)?

Beta: If he was in a state of self-deception, did that causally contribute to his success in coming back?

Gamma: If Lance’s hypothetical self-deception did in fact contribute to his later success, is that a good example to follow in that we should allow ourselves to become regularly self-deceived?

The first two questions are empirical and would best be answered after a lengthy personal interview with Lance himself. But the deeper ethical question is the third, Gamma, so I think it’s still a good idea to hazard some speculations about the first two with an eye to considering the third.

On Alpha, we need to distinguish between believing in the unusual and believing in the irrational; for it’s only the latter that is tantamount to self-deception. Lance believed in the unusual in believing he could come back, but it wasn’t irrational. Lance had already shown evidence of unusual abilities before his cancer; if he did overcome cancer, it would be not irrational to take prior success as evidence that he would still be capable of the unusual. Believing in your ability to achieve the unusual is only self-deceptive if you’re a usual person. Lance Armstrong clearly isn’t. So I answer Alpha in the negative.

On Beta, my answer to Alpha makes the issue moot. But let’s imagine in general what effect self-deception might have on athletic performance. The tempting thought is that it can enhance performance by increasing confidence. But convincing yourself that you have an ability increases confidence in a helpful way only if you actually have that ability. In short, it’s only helpful to convince yourself in a confidence-building way if it’s not really self-deceptive to do so. Convincing myself I can jump the ditch is only helpful if I actually have it in my legs to jump it. It won’t help me jump the Grand Canyon. You might object that there are many situations where it’s uncertain whether you can actually do something but it can’t hurt to try, and in those situations it’s helpful to be self-deceived. I don’t think so, because being self-deceived could decrease your awareness of what needs to be done to increase your chances. A documentary I saw on Michael Jordan made the point that he was originally thought to be a mediocre defender. He later won the NBA award for Defender of the Year. I believe that only by being honest with himself--not self-deceived--could Michael Jordan zero in on exactly what work needed to be done to make him a top defender. Self-deceptive overestimation of your abilities can cause you to do less work, not more. In situations of uncertainty, determination and self-honesty are in the recipe for success; self-deception isn't.

It should be clear by now what I think the answer to Gamma is. But let me just say one thing. Even if Lance Armstrong was in a state of self-deception and was helped by that, that’s the exception among self-deceivers, not the rule. For every Lance, there are 1,000 drunk drivers who think they’re sober, 1,000 abused spouses in denial, 1,000 dropouts who won’t face up to reality, and 1,000 bad relationships in which people won’t face up to their problems. Having something good happen because of self-deception is like winning the lottery: very unlikely, and there are much better routes to success. I conclude that it is still best to cultivate the kind of mind that is as little susceptible as possible to self-deception.