It is argued that a generalization of Heisenberg's principle of indeterminacy is possible in the social sciences. The empirical grounds for this contention lie with interference phenomena induced by transference distortions that may occur when human beings investigate the behaviour of one another.