Knowing what you want

Time for some philosophy.

The Oxford philosopher John Broome, whom I think very highly off, has written a world-class paper ("Can a Humean be Moderate?") some time ago, in which he argues that the way we know what our preferences are is through our judgments of betterness. If you cannot figure out what you prefer, A or B, you should consult what you think is better and then you will know what you prefer. I will not go into his reasons for arguing for this. In the essay the reasons are mostly negative: there is serious problems for other possible candidates for the epistemology of preferences which this account does not face.

Obviously, if Broome's account is accepted a whole raft of other problems has to be answered. Among those are the by now standard objections against the idea that we desire (and ipso facto, prefer) under the guise of the good. If we don't always willingly and knowingly desire what is good, what makes Broome think that judgments about value will be reliable in identifying our preferences?

However, I thought of yet another problem which is neat and probably irrelevant. Basically, the problem has to do with changing preferences. Suppose today I prefer A to B, but tomorrow I will prefer B to A. Suppose I believe this to be the case. To make things more interesting: suppose there is nothing wrong (brain damage, irrationality, what have you) with me now, nor do I believe there will be something wrong with me in the future. Suppose also that this is not a case of weakness of the will or anticipated temptation.

Sounds implausible? Here is an example. When I was 16 I preferred David Bowie's music to any opera by Mozart. I also knew that this would change and that it was likely that I would reverse the order of appreciation between the two. Or think of how you felt before and after you had a child. Before I had mine I preferred a career over children. Now, I wish we had more children...

Anyway, you get the idea. Here is what is strange about Broome's epistemology of preference. If today I believe that A is better than B (because I believe that I prefer A to B), and I believe that tomorrow I will prefer B to A and I don't believe tomorrow (or today) I will be incompetent, irrational, hypnotized or tempted, it seems I have a problem. For I believe p and I also believe that I will believe that not-p. And I believe that I will have good grounds for either belief (otherwise I would be incompetent, etc, which we just ruled out). So, I have good reasons for believing p and good reasons for believing that I will come to believe not-p. This seems, irrational. (In fact, this intuition is cast into a principle by Bas van Fraassen: the Principle of Reflection which says that in such a case one of the two beliefs has got to go. Well it does not exactly say that, but the intuition is clear.) So Broome's epistemology of preference has the same problem: it cannot deal with anticipated changes in one's preferences.

There seem to be three possible responses:
1. Broome is wrong and we cannot reliably know our preferences through our value judgments.
2. It is not irrational the believe p and believe that you will believe not-p at the same time.
3. The case is impossible.

Since I like Broome's proposal, I would want to hang on to 1. Furthermore, with some restrictions I like the principle of reflection, so I would want to endorse a qualified version of 2. This leaves us with 3. However, it does not seem strange that you can predict that your preferences will change without this being a case of incompetence or what have you.

So, a real problem. I'll let you know what the solution is.

Once I have found it....