Since the publication of “Behavioral studies of obedience” in 1963, and then of “Obedience to Authority” in 1974, the experiments conducted by Stanley Milgram at Yale in the early 1960s has provoked many lively debates. The opening of his archives by Yale University (Blass 2002), the partial replication of the experiment (Burger 2009), interviews with former “guinea pigs” or collaborators (Perry 2012), as well as the more general context of the replicability crisis in experimental psychology (Ritchie 2020) have triggered a revival of these debates. On the basis of these new data, several aspects of Milgram’s experiments have been reexamined: their experimental protocol, which does not exactly match Milgram's account of it (Gibson 2019); the decision not to include in the interpretation interviews in which subjects claimed not to have been fooled by the device designed to deceive them (Perry, Brannigan & alii 2018); the lessons to be learned from the behavior observed (Burger, Girgis & Manning 2011; Reicher, Haslam & Smith 2012); the usefulness of these findings in explaining the behavior of ordinary perpetrators during genocides, and in particular the Shoah (Russell & Gregory 2015; Roth 2022); the moral legitimacy of the device itself (Perry 2012). Moreover, all of these investigations converge on the question of whether the crisis of reproducibility affects experiments on authority and, if so, to what extent. As Stuart Ritchie's (2020) cautious treatment of the Milgram case in his review of the reproducibility crisis shows, this question remains open. The fact remains that debates around one of the founding and paradigmatic works of experimental social psychology ultimately call into question the very value of this discipline or at least the direction it should take. Augustine Brannigan goes so far as to call for “the end of experimental social psychology” (Brannigan 2017, p. 141) in this regard. Finally, both Milgram and the psychologists and philosophers following in his wake interpreted the astonishment that his results aroused as a symptom of adherence to a mistaken anthropology. The situationist research program thus aimed to understand what this error is, where it comes from, and how to rectify it (Ross & Nisbett 1991; Doris 2005; Sabini & Silver 2005; Roth 2022). The subsequent interrogation of Milgram's results thus also calls into question the value of this philosophical program. The authority experiments were based on epistemological and ontological presuppositions that conditioned their design and interpretation, and that have thus far rarely been drawn out and directly discussed. For example, by designating his task as an “experiment on authority”, Milgram presupposed the existence of a concept of authority that would apply to the experiment in question, but about which he remained confused, because he was theoretically eclectic, relying as much on Hannah Arendt as on cybernetics (Milgram 1974). Stephen Gibson (2019) thus looked for a concept that could adequately describe what was actually taking place during the experiment and suggested that the Foucauldian concept of power was more suitable than the concept of authority. To take another example, Milgram's explicit goal was not only to reproduce in the laboratory an ordinary relationship of subordination, but to simulate the administrative and social structure that made possible the implementation of the extermination of the European Jews (Milgram 1963; Milgram 1974; Blass 2002). Against the prejudice according to which historical events are non-repeatable singularities, he thus presupposed the possibility of solving experimentally some of the explanatory problems that plague historians. In order to contribute productively to current debates on the value of Milgram's experiments, and more generally of social psychology, this special issue tries to identify the presuppositions, theoretical consequences, and justifications of the experiments.