Data, Curve-Fitting, and Model-Building in the metaphysics of laws and causation

According to inductive metaphysicians, metaphysics is a non-exceptional enterprise in the sense that not only empirical data but also non-deductive methods that have been successfully employed in the empirical sciences can and should play a central role in metaphysics. Curve-fitting and model-building are among the most established methods used in the empirical sciences to deal with fallible data sets and to generate more error-robust theories. Assuming that also the data for metaphysical theorising are in principle fallible, it seems natural to employ curve-fitting and model-building in metaphysics too. However, within the empirical sciences, these methods rely on a fairly clear understanding of ‘data’. So, if they shall play a role in metaphysics, it must be clarified what the data for metaphysical curve-fitting and model-building are. Focussing, exemplary, on metaphysical theories of laws of nature and of causation we look at different candidates of what the data for these theories might be and examine the prospects of applying curve-fitting and model-building to these types of data.