Abstract
Suppose I am now making plans for next summer’s vacation. I can spend a week in Rome or on the Riviera, but not both. Either choice would be excellent, but after weighing various pros and cons, I decide that for my purposes Rome would be better. If I am rational, then, I must choose Rome. It is an assumption of standard decision theory that rationality requires maximizing: trying to get the maximum amount of whatever form of value we are after (usually construed as “utility”). An alternative has been proposed, under the heading of “satisficing” – being satisfied with what suffices, as it were, or settling for an option that is “good enough” – but this may seem rational only when there are costs to determining which option is best that diminish the value of choosing it, to the point where the choice of a less good option really amounts to maximizing. Where there is no serious cost, even in time and effort, to getting hold of something better – a vacation in Rome rather than the Riviera – how could it be rational to turn it down? If we pay attention to the temporal standpoint from which a choice is made, though, satisficing makes good sense. It accords with our common appeal to thresholds: adequate levels of satisfaction or value, such that getting above them is not necessary, though it might be nice. Once we have reached a threshold, it is rational – meaning rationally permissible – to stop. Pushing further toward the best may also be permissible but is not rationally required, where we already have a good enough option in hand. So if offered a chance to move to Rome while already settled happily on the Riviera, I would not be irrational to turn the offer down, even if I grant that accepting it would make my vacation even better.