It is shown that in addition to the usual difficulties related to causality, the theory of superluminal particles also exhibits paradoxical symmetry violations. In the second part of the paper a conventional paradox is revisited: causality violations at the macroscopic level follow from simple statistical arguments.