On the role of antiparticles in Dirac phenomenology