Abstract

The question of what constitutes science has been debated for a long time in philosophy of science, where some authors even consider the debate pointless (e.g. Auroux, 2000). However, increased regulations in research ethics in Canada have made the debate a concrete preoccupation. Research ethics regulators have had to define research (e.g. TCPS, 1998; 2010) and consequently, in the course of their work, ethics committees in Canadian universities are compelled to reflect on whether projects submitted to them represent scientific research. Citing real cases, this paper presents three types of projects in social sciences that appear not to meet commonly accepted criteria for scientificity, all the while representing a considerable number of projects involving human participants regularly examined by ethics committees. The typical pattern of a purely qualitative study based on a very small number of participants is described, highlighting the frequent absence of hypothesis testing, hypothesis generation, variables, or systematicity. The paper raises again the question of applicability and/or generalizability of data as a criterion for determining the scientific nature of projects. It also suggests that social sciences, like natural sciences, can be either scientific or not, if this determination is based not on the topic that is investigated but on the way researchers investigate it.