The notion of morality appears to have been introduced into quantum theory by Wigner, as reported by Goldberger and Watson. The question at issue is the famous ‘reduction of the wave packet’. There are, ultimately, no mechanical arguments for this process, and the arguments that are actually used may well be called moral. This is a popular account of the subject. Very practical people not interested in logical questions should not read it. It is a pleasure for us to dedicate the paper to Professor Weisskopf, for whom intense interest in the latest developments of detail has not dulled concern with fundamentals.

Suppose that some quantity F is measured on a quantum mechanical system, and a result f obtained. Assume that immediate repetition of the measurement must give the same result. Then, after the first measurement, the system must be in an eigenstate of F with eigenvalue f. In general, the measurement will be ‘incomplete’, i.e., there will be more than one eigenstate with the observed eigenvalue, so that the latter does not suffice to specify completely the state resulting from the measurement. Let the relevant set of eigenstates be donoted by ϕfg. The extra index g may be regarded as the eigenvalue of a second observable G that commutes with F and so can be measured at the same time.