The question whether intelligence is, or ought to be, a craft or a science has heavily preoccupied some political scientists for more than a decade. Individual Central Intelligence Agency analysts have investigated the utility of the technique by applying it to an earlier "intelligence situation". Students of intelligence methodology were looking in vain to the exact sciences for guidance. The 1950s and the 1960s were a time of great optimism among political scientists and students of international relations in the United States; hopes prevailed that the borders of knowledge about human behavior could be greatly expanded. The most promising analogy for the student of intelligence seems to be the one between political and clinical judgment. Prediction is impossible without at least some measure of extrapolation, and because extrapolation does not work without at least some continuity, prediction becomes most difficult when it is most needed-at a time of rapid or radical change.