# The Anatomy of a Decision Part II

Ken’s charge of irrationality can be understood at two levels:
1. At the first level, Ken thought I had ignored what decision theory was telling me.
2. At the second level, Ken thought my desire to cycle from coast to coast itself was irrational.

To start, we need to distinguish between what decision theory tells us to do and the inputs we apply the theory to. Perhaps an analogy is in order. The distinction is between what a calculator calculates, and what we enter into the calculator. If we want to know what 3 + 2 is, but we enter 3 + 1…we’re going to get the wrong answer. But this is no fault of the calculator; the calculator is doing what it is supposed to do by giving us the answer 4. The problem is with our inputs. We can’t expect the calculator to tell us what the correct answer is if we enter the wrong numbers.

So Ken’s charge was this: if I input the wrong kinds of things (irrational preferences, for instance), I can’t expect decision theory to tell me anything useful. Further, if I had inputed the correct types of things (rational preferences, for instance), decision theory would tell me to do something other than cycle coast to coast.

Although Ken’s charge was aimed at two levels, the issue boils down to what makes preferences rational. Note that the issue isn’t what makes a preference ranking rational. That’s well established in the literature: a preference ordering is rational if it is transitive. (If I prefer grapes to bananas, and bananas to oranges, I prefer grapes to oranges.)  Rather, the issue is what makes a preference itself rational.

Perhaps we need to reword the question to this: What makes a desire rational? At first blush, it seems odd to question whether desires are rational. Aren’t they just the types of things that we have? I desire water when I’m thirsty, food when I’m hungry, and sleep when I’m tired. These are natural bodily appetites, and it seems silly to analyze them as rational or otherwise.  But maybe this is too quick. If I were to say instead that I desire turpentine when I’m thirsty, wouldn’t you have grounds to say that that is an irrational desire: it won’t get me to my desired goal (to no longer be thirsty), but rather dead. (Of course if I’m dead, I’ll no longer be thirsty…but that’s quite a different point.)

So is a desire rational just when it helps us attain some specified goal?