Abstract
It is known that a faithful representation of a finite group generates all its irreducible representations. A question about the validity of a more general theorem is raised and its role in the theory of phase transitions discussed.