Durkheim's Lost Argument (1895–1955): Critical Moves on Method and Truth

Abstract

Durkheim’s course of twenty lectures on pragmatism, given at the Sorbonne during the academic year 1913 to 1914, has been regularly reassessed, particularly since an apparently complete English translation (1983).1 Far from being marginal in Durkheim’s work, as claimed by Steven Lukes (1973), the lectures seem central for understanding Durkheim’s epistemology and methodology. This was initially set out in his two doctoral theses – the main one on the division of labour (1893) – then substantially reworked in later writings, particularly Les Formes élémentaires (1912). Unfortunately, we know the lectures only from a posthumous reconstruction by the faithful Durkheimian and sympathiser with Marxism, the philosopher Armand Cuvillier, who published Pragmatisme et sociologie in 1955, drawing on two anonymous sets of ‘student notes’ that later disappeared. It is thus diffi cult to know the scope and effect of Cuvillier’s own rewriting of these notes. Moreover, he made his reconstruction forty-two years after the actual presentation by Durkheim at the Sorbonne.2 The sociological context in France was by this time entirely different. The most prominent sociologists, such as Jean Stoetzel, were outspoken anti-Durkheimians in their demand for an empirical knowledge clearly severed from any philosophical foundation. The Durkheimians who tried to pursue the founder’s endeavour in the interwar period were dead. The very fi rst reviews of Cuvillier’s edition3 indicate that Durkheimianism seemed to belong to the intellectual past, at least since the death of Marcel Mauss in 1950. A third set of student notes on the lectures was recently discovered by one of us, as an extension of our joint project (Durkheim 2003). The dis covery allows a reassessment of Pragmatisme et sociologie and a better understanding of Cuvillier’s handling of his material. Without claiming to revolutionize the interpretation of Durkheim’s complex relationship with pragmatism and without necessarily disagreeing with the most recent scholarly work on the issue, we would like to shed some new light on the proper