Uncompromising Empiricism Once Again : Big Data and the Case of Numerical Taxonomy

Big Data are not rough data or mere tools. They foreshadow epistemological and ethical changes and raise issues as to the nature of knowledge and the categorization of reality. In this paper, we aim to analyse the so-called “data science” from a historical and epistemological perspective. We will first consider how, in the name of a new and uncompromising empiricism, a reconfiguration of the traditional view of causality in science is claimed on the basis of data analytics. Secondly, we will investigate a historical antecedent of the current debate by comparing these claims with the ones raised by Numerical Taxonomy, or Phenetics, in the second half of the twentieth century. This will allow us to highlight how the opposition between knowledge-driven science and data-driven science resurfaces in the debate about Big Data science.