On the Uses and Abuses of Psychological Research in Moral Philosophy: The Case of Moral Progress

Recent decades have seen a steep increase in the numbers of moral philosophers who cite, engage with and make serious use of empirical research from psychology and its neighbouring disciplines. These are welcome developments; however, they also come with considerable new challenges. The First Challenge is to make the case that these developments are not fundamentally wrong-headed. Moral philosophy, so goes the objection, operates in normative space. Moral philosophers are in the business of figuring out how humans ought to live, think and act in the world. Yet it is impossible to derive an ‘ought’ statement from an ‘is’ statement, and so empirical premises do not in fact (and could not) play any significant role in moral philosophy. The moral philosophers who aim to incorporate more be behavioral science into their own work are therefore either misguided, incompetent or both. To addresses the First Challenge, Walter Sinnott-Armstrong and I argue that, as a matter of fact, lots of work in moral philosophy relies on assumptions about how people think, feel and act (chapter 3). Therefore, there are many lessons that disciplines like psychology, neuroscience, experimental economics and cognitive anthropology have to teach moral philosophers. We illustrate our argument using the case study of the reliability of moral judgments and decisions. The Second Challenge is to bridge the gap between empirical research and philosophical argument. It is one thing to recognize that an argument relies on assumptions about how people think, feel or act, and so needs empirical support. It is another thing to then make a convincing empirically-informed case for these assumptions. To achieve this, there are a series of inferential gaps between these assumptions and the (quantitative) research that need to be bridged. In my thesis (chapters 4-6), I focus on four of these inferential gaps: construct validity; effect size; generalizability; and causal inference from observational data. I further illustrate, explain and discuss each gap with the help of two case studies from the literature on the causes of large-scale moral progress. In the final chapter, I summarize the main points of the thesis. I then argue that three of the four challenges discussed throughout the thesis (construct validity; generalizability; effect size) are not just of concern in the literature on the causes of moral progress, but likely affect empirically-informed arguments in many other areas of philosophy, as well. To make this case, for each challenge, I highlight a series of additional examples where empirically-informed arguments failed (or have been argued to fail) due to not meeting that challenge. I then argue that these examples and other examples like them are to be expected given the types of questions philosophers seek to answer and the approaches they tend to use to answer these questions. I conclude that if this picture is approximately correct, then empirically-informed philosophy faces a crisis.