The natural selection of conservative science

Abstract

Social epistemologists have argued that high risk, high reward science has an important role to play in scientific communities. Recently, though, it has also been argued that various scientific fields seem to be trending towards conservatism-the increasing production of what Kuhn (1970) would have called 'normal science'. This paper will explore a possible explanation for this sort of trend: that the process by which scientific research groups form, grow, and dissolve might be inherently hostile to high risk science. In particular, I employ a paradigm developed by Smaldino and McElreath (2016) that treats a scientific community as a population undergoing selection. As will become clear, perhaps counter-intuitively this sort of process in some ways promotes high risk, high reward science. But, as I will point out, high risk high reward science is, in general, the sort of thing that is hard to repeat. While more conservative scientists will be able to train students capable of continuing their successful projects, and so create thriving lineages, successful risky science may not be the sort of thing one can easily pass on. In such cases, the structure of scientific communities selects against high risk, high rewards projects. More generally, this paper makes clear that there are at least two processes to consider in thinking about how incentives shape scientific communities-the process by which individual scientists make choices about their careers and research, and the selective process governing the formation of new research groups.

Accomplishing breakthroughs in behavioural medicine research

“…A minimum of two levels are integral in considering how incentives shape scientific communities to focus on simple rather than complex approaches 6 . First is the process by which individual scientists make choices about the focus and level of complexity of their work.…”

Practical Methodological Reform Needs Good Theory

“…Lakatos. Philosophers of science are actively engaged in theoretical modeling of the processes active in the scientific enterprise, and tools such as cultural evolutionary models, network epistemology, and game theory illuminate many things that reformers should perhaps mind (Bicchieri, 2016;O'Connor, 2019a;O'Connor & Weatherall, 2018;Skyrms & Pemantle, 2009;Zollman, 2007). Integrating these perspectives, some tentative recommandations are possible.…”

“…It can make our forensic assays of the field more efficient and impactful (Field et al, 2019). Theory can even give us hints as to what cues we may (even inadvertently) be sending obvservers, perhaps undermining the types of communities most likely to actually solve the practical challenges science faces today (O'Connor, 2019a;O'Connor & Weatherall, 2018;Zollman, 2010).…”

Interdisciplinarity Can Aid the Spread of Better Methods Between Scientific Communities

“…Or this could stem from the differential success of students whose advisors use credit-producing methods. Previous models have shown that these sorts of selection processes can help explain failures of methodology and discovery in science (Smaldino and McElreath, 2016;O'Connor, 2019;Holman and Bruner, 2017;Stewart and Plotkin, 2020;Tiokhin et al, 2020).…”