For the relativist, the choice of theories in science is so massively underdetermined by empirical considerations -or for that matter any considerations internal to science- the only plausible explanation of why the theories that do get chosen are actually chosen is one which explains that process in "sociological" terms, i.e. by appeal to the interactions between the individuals who comprise the community of expert professionals who determine theory acceptance in these matters, and by appeal to the larger social context in which they practice their profession.