Does research integrity matter?

Abstract A recent strand of literature argues that the importance of epistemic virtues is overstated and that knowledge generation and transmission benefit from traits that can be characterised as epistemic vices. This goes against the stance of the research integrity literature, which takes it as given that epistemic vices are bad and should be combated. Arguments ‘in defence of vice’ utilise qualitative, historical cases of epistemic vice that contributed to desirable scientific outcomes or knowledge transmission, and formal models of scientific research in which virtues can be harmful and vices beneficial. A typical stance is that negative effects of individual-level vices will be addressed at the level of the scientific or research community. I argue that singular case studies and highly stylised formal models only demonstrate possibility results with no established relation to the actual dynamics of research or even to the universe of possible models. While historical evidence does serve as a counterexample to the claim that epistemic virtues are always beneficial and epistemic vices are always harmful, that extreme stance is not needed to endorse epistemic virtues and integrity. The harms of basic epistemic vices to the modern research enterprise are evident from considering their implications for components of the research process such as peer review. While economics is the inspiration for many formal modelling exercises, the literature neglects economic literature on market failures, the importance of social norms, and trust. Recent contributions have not provided an adequate basis for treating research integrity and traditional epistemic virtues as anything less than fundamental to the process of inquiry and knowledge generation.