This book examines several epistemological regimes in studies of numerical data over the last four centuries. It distinguishes these regimes and mobilises questions present in the philosophy of science, sociology and historical works throughout the 20th century. Attention is given to the skills of scholars and their methods, their assumptions, and the socio-historical conditions that made calculations and their interpretations possible. In doing so, questions posed as early as Émile Durkheim's and Ernst Cassirer's ones are revisited and the concept of symbolic form is put to the test in this particular survey, conducted over long period of time. Although distinct from a methodological and epistemological point of view, today these regimes may be found together in the toolbox of statisticians and those who comment on their conclusions. As such, the book is addressed to social scientists and historians and all those who are interested in numerical productions.

This book is a translation of an original French edition. The translation was done with the help of artificial intelligence (machine translation by the service DeepL.com). A subsequent human revision was done primarily in terms of content, so that the book will read stylistically differently from a conventional translation.

Autorentext

After a dissertation in Mathematical Statistics and another in History and Social Sciences, Éric Brian became a Professor at EHESS (École des Hautes Études en Sciences Sociales, Paris, France) and between 2002 and 2018 guest professor at the Institut für Philosphie at the University of Vienna, Austria. At EHESS, he is the current holder of the Chairs "History of Probabilities and Statistics" and "Historical sociology of symbolic instruments". Since 1995, he is the editor of Revue de Synthèse, a journal at the crossroads of History, Social Sciences and Philosophy created in1900.