The problem of group rationality can be fairly simply stated: Under what conditions is a group of scientists rational? There are a few scattered answers, some not so well known, some not so well reasoned. Our task is to examine some of these theories of group rationality to show why that problem is a marvelous puzzle, why that puzzle is yet unsolved, and why it needs solving. Hitherto, philosophers of science have dealt with the following problems: Under what conditions is a theory scientific? (Popper: when the theory is falsifiable.) Under what conditions is a theory making scientific progress? (Lakatos: when the theory is a successful research program.) Under what conditions is a decision to accept a scientific theory rational? (Bayes: when the prior probability of the scientific theory multiplied by the likelihood of the theory, the total divided by the probability of the evidence, is high.) Finally, under what conditions is a scientist generally rational; in other words, what is individual rationality? There is a paucity of theories on this; perhaps attempts at solving the problem of group rationality will stimulate interest in this area by making it evident that this last question lies intriguingly beneath the topsoil.

In 1983, the problem of group rationality was a new problem. By now, more than twenty years later, several philosophers have been engaged in the task of solving it; but in solving it, they have sometimes assumed that it is a problem of a different stripe.