Talk by Alexandre Hocquet, Konrad Hinsen and Frèdèric Wieber (INM-6/IAS-6 Seminar)

Host: Markus Diesmann

The past, present, and future of reproducible computational research practices

Reproducibility has been considered one of the core values of science since the beginnings of organized research in the 17th century, even though it has never been obvious in practice. Whereas its role as an indicator of rigor and trustworthiness has remained the same, its technical and social embodiment has evolved together with scientific methods and institutions. Alexandre Hocquet and Frédéric Wieber will first present a short outline of this evolution and briefly assess the interest of focusing on software as a research object for historians of science. Konrad Hinsen will then focus on the aspects specific to computational research, which is characterized by the determinism of its fundamental technique on one hand, and by the complexity of its software on the other hand. The roles of reproducibility and replicability therefore differ from their traditional ones in experimental research. We will conclude with an outlook towards the emergent practices of Open Science, in which reproducibility is increasingly considered an important quality label.

Alexandre Hocquet and Frédéric Wieber are historians of science. Their main research interest is the role of software in computer simulations.

Konrad Hinsen is a computational biophysicist, with a special interest in computational reproducibility issues, and one of the founders of the ReScience C journal. A brief video of a dialogue of theirs, part of a MOOC on reproducible research is available on YouTube.

Alexandre Hocquet, Konrad Hinsen and Frédéric Wieber
Käte Hamburger Kolleg, RWTH Aachen,Center for Molecular Biophysics, CNRS, Orléans, Archives Henri Poincaré, Nancy