Elsewhere, I portrayed the problem of rational choice as the problem of what to do when certain kinds of information are not available.  The more information that is not available, the less clear it is what it is rational to do.  In one of the most widely accepted accounts of rationality, it is assumed that one knows a great deal:  the alternative actions open to you, the various possible consequences of those actions, the value of those consequences and the probabilities that each consequence will occur.  One can then do an expected utility calculation to decide which action is best.  The classic sort of case where we face this sort of choice is casino gambling, for example, when we play Roulette or Poker.  

 But casino gambling actually involves more information than we have in every day life and the less information that is  available, the less clear it is what it is rational to do.  We often struggle to apply expected utility theory to cases outside of casino gambling, as when we do certain kinds of  'cost-benefit analyses'.  

Because of its connection to the work of the important political philosopher John Rawls, philosophers have been particularly interested in the case in  which we know pretty well what our alternatives are, and we know pretty well the possible consequences of each alternative action, but we simply have no idea what probabilities to assign to those consequences.  I talk some folk into a nice Montana vacation and when they arrive, I suggest a hike in the local Gallatin mountains.  But I warn them the mountains are swarming with bears.  I offer to sell them my very own invention, a hyper plasmid electocorpuscular bear repelling system. It is admittedly pricey at $1,500 and heavy, a real bear to carry.  I inform them of the terrible things that will occur to them if them if they meet a bear in the mountains unprotected by the device.  But since I am honest, I inform them that if they do not meet any bears, the heavy weight of the device, and the $1,500 price tag, will seem a tad high.  They ask me how likely it is that they will meet a bear.  I inform them that some days it is strikingly unlikely that they will see a bear and other days it is a virtual certainty, with everything in between.  No one can know in advance what the odds are this day.  They must choose whether or not purchase and carry my device in the fact of this deep ignorance.  So what is it rational for an intrepid dayhiker in the Gallatin mountains to do?  

Their choices are reasonably clear, hike with the device or hike without it.  The various possible outcomes are also reasonably clear.  If they hike with the device and meet a bear, they will proceed with their hike, tired for the extra weight, but with some nice 'encounter with nature' snaps.  If they hike with the device and do not meet any bears, they will have an enjoyable hike but be more tired after slogging up 2000 feet over a four mile distance, the hike I picked out for them.  However, if they do not bring the device and meet no bears, that four mile, 2000 foot trip will seem, well, like a four mile, 2000 foot trip, but they will survive well enough.  But if they do not bring the device, and meet a bear, they are destined fertilize the forest floor as bear droppings.  If they just had the probability that they would run up against a bear, they would have a shot at some sort of expected utility calculation.  But they lack this crucial bit of the puzzle.  

Case I P1 P2 P3 Total
Act 1 10 -2 5 13
Act 2 2 -4 3 1
Make a Free Website with Yola.