Following-up on a previous neuroethics seminar about the integration of social neuroscience and the traditional social sciences, Dan Dohan and Kate Rankin led a discussion contrasting research in the natural and social sciences. Throughout our conversation, we drew from arguments presented in “Proving or Improving: On Health Care Research as a Form of Self-Reflection” by Annemarie Mol and “The Heart of the Matter. About Good Nursing and Telecare” by Jeannette Pols.

In response to Mol, who critiques clinical trials for merely addressing the biological aspects of disease without incorporating individuals’ lived experiences with illness, we questioned in what ways clinical research methods and qualitative methods might be viewed as complementary. We considered that there might be different standards of success in different inquiries; while natural scientists often seek to understand truth through quantitative evidence and generalizable data, many social scientists accept a probabilistic notion of truth and embrace the complexity of human health within broader social contexts. In discussing the utility of statistical significance within scientific research as a means of identifying studies with meaningful conclusions and determining the safety and efficacy of clinical treatments in larger populations, practices such as p-hacking suggest that a bare reliance on p-values in the competitive journal publication process may invite a narrow view of what is truly significant in science.

In the end, there may be more congruence between approaches to truth in the social sciences and in neuroscience than we originally supposed. While social scientists often regard their investigations of complex social phenomena as necessarily incomplete, neuroscientists also often regard their models (e.g., of the basal ganglia) as merely useful yet simplified tools for distilling complicated information.