Title: Ontic injustice in modelling : uprooting the value-free ideal
Abstract:
Causal modelling is sometimes taken to be a value-neutral activity, given that many of the inference procedures are formalized, and how it looks like the variables are measuring numerical parameters that couldn’t feasibly count as value-influenced. I will offer two main lines of reasoning to show that this is false, and that multiple distinctive kinds of injustice can be perpetrated with ahistorical and supposedly value-free modelling, especially in choice of variables. Recent work on ontic injustice with respect to gender categories concerns the way in which the very categories or conceptual divisions that we use can wrong the people in those categories, even if the categories seem, objectively, to apply. I will show how this is a broader problem than just for socially constructed kinds such as gender. Especially in health or biological sciences, the way we construct causal variables and select them for systems of variables can be loci of ontic injustice. I propose that the root of this is attempts to follow the value-free ideal (VFI). I show that it is meaningless to posit that there is any genuinely value-free way to do these modelling tasks; instead, there are value-opaque ways where the appearance of value-freedom is spurious, instead reflecting merely that the values involved match closely to those of the modeller(s). Furthermore, even if we grant that someone might be in principle able to achieve the VFI, it is morally otiose to even aim at it. To decide to follow the VFI in choosing variables for modelling is to tell those who will be negatively affected by any failure to achieve the ideal version of the VFI that consequences to them don’t matter. I conclude with some recent cases to show that we should instead draw on principles from transformative justice in many cases of modelling where the VFI had been supposed to be valid; it improves the modelling and outcomes for use of the model more than trying and failing to achieve to VFI.
Bio:
Holly K. Andersen is a philosopher of science who works on issues related to causation, explanation, and pragmatism. Much of her work concerns causal modelling and explanation, from a metaphysical or foundational perspective and from an applied, case-study driven perspective. Her PhD is from the History and Philosophy of Science Department at the University of Pittsburgh, and she has been in the Philosophy Department at SFU for 16 years.
