It's not that journals need to be more careful. Institutions need to be set up so that we don't produce this problem.