The “Materiality Principle” in Studies of the Epistemology of Computer Simulations

Abstract

There is an argumentative strategy called the “materiality principle” (J. Duran) in the studies of epistemology of computer simulations. The strategy is about attributing epistemic characteristics to computer simulations on the basis of their ontological similarity with material experiments (the degree of “materiality”). This comparison is motivated in part by linguistic intuitions: computer modeling is often referred to as a “computational experiment” etc. It is reasonable then to consider computer simulations as a subspecies of scientific modeling, occupying an intermediate position between the theoretical and experimental level of scientific knowledge (as it is within the framework of modern approaches to the structure of scientific knowledge, namely semantics and pragmatics). Four different kinds of the implementation of the “materiality principle” are considered, arguing both in favor of the reliability of computer simulations and against it. Computer simulations blur the boundaries between material and virtual (computational) experimentation. The “materiality principle” can work as an argument both for and against this distinction. Two main arguments against the principle of materiality — the “opacity argument” and the “multiple realizability argument” — are hereby presented and strengthened. We conclude that the materiality principle is insufficient to substantiate the epistemological status of computer simulations.