Thursday, September 01, 2005

Decision Theory as Casuistry

By 'casuistry' I mean casuistry in the older technical sense (the application of general moral principles to cases) rather than in the more modern colloquial sense.

It is commonly recognized that Pascal's Wager is the first clear example of a decision-theoretic argument. Decision theory as we know it really gets started with David Bernoulli in the next century; but the rudiments are in place in Pascal's exposition, so much so that, setting aside a few ambiguities, it is very easy to put the Wager into decision-theoretic terms.* I think it is also commonly recognized by Pascal scholars that Pascal's Wager is influenced by (Pascal's interpretation of) Jesuit casuistry.** It is not exclusively Jesuit, obviously. The root source is a very old response to a very old argument.

Thomas More somewhere tells a story in which two men meet, one of whom is engaged in physical mortifications. The other man asks him, "Why are you doing that?"

The man of mortifications replies, saying, "I am mortifying my body so that it will not tempt me with sin; sin is the way to hell."

The other man laughed at him, saying, "And what a fool you are if there is no hell."

To which the man of mortifications replied, "And if there is a hell, sir, what sort of fool are you?"

Many variations of this sort of argument can be given. There is nothing intrinsically religious about the argument form; anything that gets laughed at this way (philosophy, virtuous living, etc.) is amenable to this sort of treatment. Pascal's argument is more sophisticated, in that it is not merely a retort but an analysis. But the spirit is the same.

Pascal was very critical of Jesuit casuistry (it is attacked in the biting and witty satire of the Provincial Letters). But it's also clear that Pascal incorporates elements by means of which the Jesuits had refined the above retort; he just avoids what he sees as Jesuit errors. The Jesuits in their casuistry had taken thought for how to attract the 'libertine' to moral action and love of God. While Pascal thinks that in doing so they effectively began to treat libertinism as morally acceptable, he appears to be trying to attract the same crowd to the same thing. It's not surprising, then, that the Wager involves cleaning up a type of argument used by the Jesuits.

The roots of decision theory, then, are in casuistry, in the old sense. And indeed, it can easily be seen that decision theory is simply an abstract formalization of certain aspects of casuistry; it is the application of certain kinds of general principles to cases. Like casuistry in the old sense, this is a valuable thing; like casuistry in the old sense, it can, if not watched, degenerate into casuistical argument in the modern sense. And it faces the same problems of any casuistry: it cannot translate into actual choice and policy unless we take, at least implicitly, a position on the principle of application to cases. And that's a puzzling issue that needs more study than it is usually given.

* Which is not to say that I would agree with someone who claimed that the Wager just is an implicit decision matrix; certain features of the Wager are very amenable to decision-theoretic analysis, but the Wager is actually an interaction with a certain kind of position in a certain kind of context, on certain principles that are given justifications by Pascal (and which will not show up on a decision matrix because any decision matrix will presuppose them). Further, as I've discussed elsewhere, I don't think the argument is an argument for God's existence but an argument giving an account of the practical rationality of believing there is a God even if we can't be certain of it. In this sense, it should be read with (for instance) Augustine's argument on the utility of believing (i.e., that it is unreasonable to demand that no one believe anything that is not certain); it is, as it were, a meta-argument about why arguments like Augustine's are in conformity with practical reason.

**For a useful introductory discussion, see Jon Elster's "Pascal and decision theory" in The Cambridge Companion to Pascal, Nicholas Hammond, ed. Cambridge (New York: 2003) 53-74.

No comments:

Post a Comment

Please understand that this weblog runs on a third-party comment system, not on Blogger's comment system. If you have come by way of a mobile device and can see this message, you may have landed on the Blogger comment page, or the third party commenting system has not yet completely loaded; your comments will only be shown on this page and not on the page most people will see, and it is much more likely that your comment will be missed.