Journal of Statistical Physics, Vol. 62, Nos. 3/4, 1991. p. 841-861.
The meaning of truth in quantum mechanics is considered in order to respond to some objections raised by B. d'Espagnat against a logical interpretation of quantum mechanics recently proposed by the author. A complete answer is given. It is shown that not only can factual data be said to be true, but also some of their logical consequences, so that the definition of truth given by Heisenberg is both extended and refined. Some nontrue but "reliable" propositions may also be used, but they are somewhat arbitrary because of the complementarity principle. For instance, the propositions expressing wave packet reduction can be either true or reliable, according to the case under study. Separability is also discussed: as far as the true properties of an individual system are concerned, quantum mechanics is separable.