Abstract
Factive scientific understanding is the thesis that scientific theories and models provide understanding insofar as they are based on facts. Because science heavily relies on various simplifications, it has been argued that the facticity condition is too strong and should be abandoned (Elgin 2007, Potochnik 2015).
In this paper I present a general model of a metabolic pathway regulation by feedback inhibition to argue that even highly simplified models that contain various distortions can provide factive understanding. However, there is a number of issues that need to be addressed first.
For instance, the core of the disagreement over the facticity condition for understanding revolves around the notion of idealization. Here, I show that the widely used distinction between idealizations and abstractions faces difficulties when applied to the model of a metabolic pathway regulation.
Some of the key assumptions involved in the model concern the type of inhibition and the role of concentrations. Contra Love and Nathan (2015) I suggest to view these assumptions as a special sort of abstraction, as vertical abstraction (see also Mäki 1992).
Usually, it is the idealizations that are considered problematic for the factivist position because idealizations are thought to introduce distortions into the model, something abstractions do not do. However, I show that here abstractions distort key difference-makers (i.e. type of inhibition and the role of concentration), much like idealizations do elsewhere.
This seemingly further supports the nonfactivist view, since if abstractions may involve distortions then not only idealized models but abstract models as well cannot provide factive understanding. I argue that this is not the case here.
The diagrammatic model of a metabolic pathway regulation does provide factive understanding insofar as it captures the causal organization of an actual pathway, notwithstanding the distortions. I further motivate my view by drawing an analogy with the way in which Bokulich (2014) presents an alternative view of the notions of how-possibly and how-actually models.
The conclusion is that, at least in some instances, highly simplified models which contain key distortions can nevertheless provide factive understanding, provided we correctly specify the locus of truth.