To understand a [case of delirium] properly and to be able to apply the most appropriate treatment, the doctor needs to know what its point of departure was' (Durkheim 1912a:10; t.6). Durkheim's linking of the study of religions with the practice of the alieniste shows above all that the believer's point of view does not have explanatory value when it comes to religious phenomena. Does this necessarily mean that religious beliefs are mere misunderstandings or illusions? On the contrary, Durkheim thought that they had objective meaning, and even suggested that thought categories themselves had a religious origin (ibid.: 3). To discover this meaning, what was needed was simply a more reliable point of view than the believer's. However, having recourse to the scientific viewpoint still left him open to a petitio principii, since he also held that the objectivity of abstract thought was genetically derived from that of religious representations, i.e. from the demonstrandum. If we answer, as Durkheim in fact did, (1) that religious representations obey practical necessities only, not theoretical ones, we would be sidestepping the problem, since representations are always representations of something; (2) they always have a cognitive or speculative aspect. The comparison with thealieniste contains another paradox: whereas discovering the objective cause of a case of madness makes it possible to measure the great distance between pathology and a normal relation to reality, the doctor must also apply therapy, which is a means of making the phenomenon itself disappear. The search initiated by the postulate of objectivity (of religion, of collective representations) would thus seem to lead to the discovery that there is no objectivity at all.