From Alan Camina
Andreas Frew implies that a system of giving out research grants at random is a silly idea (Forum, Washington Diary, 27 January, p 50). I feel that random decision-making deserves a serious analysis and should not be dismissed in so cavalier a fashion.
The world is one of imperfect information, so while rational argument may reduce the possible courses of action, it rarely leads to a unique decision. We then have to invoke phrases such as “expensive” or “judgment” to resolve the problem.
Let me try to illustrate this. Some years ago, I asked my head of department for a decision. Some time passed so I went to ask what his decision was. He explained that he had difficulty in deciding because he did not have enough information. When pressed further, he admitted that he would never have enough information, so he made a decision, anyway.
The second example concerns research studentships in mathematics. There are certain topics in mathematics which are not studied much in Britain. Given a particular topic, there are two choices: one is to give it priority and try to revive the topic, the other is to ignore it and let it fade away. Both policies can be argued quite persuasively, but what criteria would be needed in order to make a rational decision and how could these criteria be evaluated?
Advertisement
Much of the modern craze for collecting data is to avoid having to admit that we are just using our judgment. We invent criteria which can be calculated in order to pretend that decisions are rationally based. Since the data are often meaningless, the process is equally meaningless.
