Abstract

This paper critically examines Mitchell’s integrative pluralism. Integrative pluralism is the view that scientific explanations should primarily aim to integrate descriptions from different ontological levels. We contend that, while integrative pluralism is a fundamental strategy in contemporary science, there are specific reasons why one should not expect integration in the sense developed by Mitchell to be the optimal strategy and the one that scientists should always aim for. Drawing on some examples from contemporary biology, we argue that integration is sometimes neither epistemically desirable, nor ontologically achievable. We conclude that integrative pluralism should thus be limited to a specific class of complex systems but cannot be generalised as the preferable research strategy without further information about the epistemic practices of the scientific community or the ontology of the system under investigation.