Understanding Epistemological Debates in the Humanities and Social Sciences Can Aid in Model Development: Modeling Interpretive and Explanatory Theories

Abstract

When embarking on a new model, a programmer working with scholars in the humanities is often tasked with helping a likely non-programmer(s) with critical decisions concerning how to set about modeling the theory at hand. I argue that, in these early stages of development, the goals of the researcher and epistemological considerations are of paramount importance to the development of valid computational models. In order to start this discussion with a real-world example, this chapter outlines a mistake, made by myself, in a critical stage early on in the modelling process. Specifically, using early discussions with the theorist, I suggested modeling the theory as an agent-based model. After some critical reflection after substantial development, I came to the conclusion that the theory is better modelled as a system dynamics model. In the chapter, I reflect on what drove me to make the original mistake, what caused me to realize the error, and what the result of correcting the error was. I share this mistake in this chapter for two reasons: (1) so that others in similar situations might not fall into the same trappings and (2) to open up a dialogue concerning epistemology of the social sciences and humanities insofar as it relates to modelling and simulation. My general conclusion is that the thinking received by the social scientist and humanities scholar should be fully flushed out at early stages of model development, as their strength is attention to theoretical nuance. This is of utmost importance to model development, which if unaddressed should still cause issues later during model validation and verification.