Thursday, January 24, 2013

Reasons and Persons: What is a self-defeating theory?

Reasons and Persons is separated into four parts. The first part is about self-defeating theories.

Before getting into self-defeating theories, let's make it clear what a theory is. Whereas in science a theory is an explanation that ties together observations into a more general idea, in Reasons and Persons a theory is a set of principles that state how people ought to act. For example, utilitarianism constitutes an ethical theory that states all people ought to act so as to maximize the happiness of all people. This use of the word theory, denoting a principle or practice rather than an explanation, is more like its use in music theory than its use in the theory of gravity.

Ethical theories, like any theories, may succeed or fail. What causes an ethical theory to fail? If we were scientific about the matter, we might try setting up social experiments whereby we would have groups of people live their lives according to different theories, and then we would observe which theories work best. However, such an experiment would be impossible to set adequate controls for, making such experimentation impractical. Moreover, being philosophers, we're more apt to settle things from the comfort of our armchairs, by reasoning through the details.

A common way theories fail is by being measured according to the values of another theory. For example, a utilitarian, who believes we each ought to act so as to maximize everyone's happiness, would conflict with someone who believes the most important principle is that we each worship God. Though there might be points of compatibility between the two theories, where worshiping God would also maximize happiness, there would also be inevitable points of conflict—a dilemma of having to either worship God or maximize happiness, but not both. No matter how rare such points of conflict would be, any one conflict would be enough to cause each theory to fail in terms of the other.

Failure in the terms of another theory isn't ipso facto failure. If two theories conflict with each other—let's call those theories A and B—it could be that A fails according to B because B is wrong. This wouldn't make A wrong; at most it might make adherents of B wrong in their belief of A being wrong. It could turn out that A is either right or wrong, but in either case B would be useless for measuring A.

Is there a better, more objective way to measure a theory? Yes, there is. A theory always fails, regardless how other theories measure it, if that theory is self-defeating. A self-defeating theory fails in its own terms.

How can a theory be self-defeating? Here's an example. Take the following theory about self-interest: that (1) each of us ought to act so as to bring about outcomes that are best for ourselves (and without regard to others' circumstances) and (2) we each ought to never deny ourselves from fulfilling our desires. It turns out that this theory fails in its own terms. Here's a hypothetical scenario described by Parfit that causes the theory to fail.

Suppose that I am driving at midnight through some desert. My car breaks down. You are a stranger, and the only other driver in this desert. I manage to stop you, and I offer you a great reward if you drive me to my home. I cannot pay you now, but I promise to do so when we reach my home. Suppose next that I am transparent, unable to deceive others. I cannot lie convincingly. Either a blush, or my tone of voice, always gives me away. Suppose, finally, that I know myself to be never self-denying. If you drive me to my home, it would be worse for me if [I] pay you the promised reward. Since I know that I never do what will be worse for me, I know that I would break my promise. Given my inability to lie convincingly, you know this too. You do not believe my promise. I am stranded in the desert throughout the night. This happens to me because I am never self-denying. It would have been better for me if I was trustworthy, disposed to keep my promises even when doing so will be worse for me. You would then have driven me home.

(I've previously written about this scenario in a previous post.)

Some of you who may recognize the above scenario as a colorful instance of a social dilemma. Specifically, the narrator's reward and punishment payback is the same as with a prisoner's dilemma, though the stranger's payback is different from a prisoner's dilemma, so this scenario isn't a true prisoner's dilemma.

Be honest Lie
Give ride 3, 4 4, 1
Don't help 1, 3 (tie) 2, 3 (tie)

Here's how to read the table. The narrator chooses the column, and the stranger chooses the row. The pair of numbers in each cells denote the payback for both the narrator and stranger, with the first number being the payback for the narrator and the second number being the payback for the stranger. A higher number is better for that person, but irrelevant to the other person.

Regardless of the choice the stranger makes, the narrator is always better off lying, and stiffing the stranger, as measured by the narrator's own principle of never being self-denying. The stranger has a more difficult choice because he's better off helping the narrator only if the narrator is honest. Otherwise the stranger is better off driving off without helping. However, because of the narrator's transparency, the stranger knows the narrator will choose to lie, and thus the stranger will choose not to help out. This leads to the outcome in the lower-right cell marked in bold, which is the Nash equilibrium for this scenario.

A consequence of all this is that in order for the narrator to create an outcome that's best for the narrator, the narrator must do what's not in his immediate self-interest: the narrator must be honest with the stranger. More to the point, the narrator ought to convince himself that he is better off denying himself the immediate fulfillment of his own desires, thus giving him reason to reject his own ethical theory.

This is a peculiar result, with a possible consequence that goes beyond the mere success or failure of this one ethical theory. I'll elaborate on this in subsequent weeks.

Related posts:

No comments: