The Nobel Prize stands above all other awards in science in terms of both scientific prestige and public recognition.  This paper examines the determinants of awarding the Nobel Prize in Physics. Using a novel dataset of candidate nominations, publications, and social networks from 1901 to 1966 – the most recent year available under the Nobel statutes – we analyze the contributions of social closure, cultural closure, and cumulative advantage to the selection of laureates by the Royal Swedish Academy of Sciences.  The results show that cultural closure – the preference of judges for certain specialties within physics over others – plays an especially important role in prize decisions.  Our findings are consistent with the existence of a strong and globally recognized hierarchy of research topics within physics, which has consequences for the development of scientific research over the long term.