I work mainly in philosophy of science. My primary interest is currently in how knowledge is produced in field sciences, i.e., in sciences outside the confines of the laboratory and unable to conduct shielded experiments. What is the relation, for instance, between the development of abstract economic models, and progress in terms of better predictions and explanations of messy real-world events? Roughly speaking, I argue for more emphasis on local empirical work and less on development of generalised theory. I am also interested in the role of prediction. When can we rest content with after-the-fact explanation, and when should we insist ins…
I work mainly in philosophy of science. My primary interest is currently in how knowledge is produced in field sciences, i.e., in sciences outside the confines of the laboratory and unable to conduct shielded experiments. What is the relation, for instance, between the development of abstract economic models, and progress in terms of better predictions and explanations of messy real-world events? Roughly speaking, I argue for more emphasis on local empirical work and less on development of generalised theory. I am also interested in the role of prediction. When can we rest content with after-the-fact explanation, and when should we insist instead on accurate prediction? What kind of knowledge do we really have if we can't predict well?
I am currently working on a book, 'Science in a Fragile World', under contract at OUP. In it, I examine how to investigate a world in which laws and causal relations are 'fragile', in other words are intermittent and unpredictable. I argue that:
1) Once we venture outside the confines of laboratories and engineered artefacts, much of our world is like this, including many of the questions that we care about most: war, environmental damage, pandemics, elections, and more.
2) These questions need to be investigated by historian-like detailed case studies. Unfortunately, familiar methods such as experiments are less useful, because their results can no longer be relied on to generalise. There is a premium on accurate prediction, even though (in fact because) in messy cases it is difficult. Roughly speaking, I argue for more emphasis on local empirical work and less on development of abstract theory.
Although I work mainly in philosophy of science, I have also written extensively on related themes in metaphysics, especially the notions of causation and causal explanation. These various strands connect when, for instance, analysing the use of statistical techniques to measure causation, or the use of experiments in psychology. I have also applied some of this causal training to debates around several other philosophical issues, including scientific progress, harm, innateness, and free will.
