Doing Research on Simulation Sciences? Questioning Methodologies and Disciplinarities

Computer Simulations as Scientific Instruments

“…They are just model and equation solvers. To others computer simulations are a new form of experiment that encompasses broad and heterogeneous set of practices (Winsberg, 2010;Gehring, 2017;Resch et al, 2017). In this latter view, computer simulations are broadly defined as experiments (Barberousse et al, 2009;Lenhard, 2004;Parker, 2009).…”

“…Their implementation, of course, requires experimentation and the gathering of supporting data, or even metadata-on both, the implementation itself and on its results-to justify our reliance on it (Von Neumann and Brucks, 1946;van de Velde, 1960;Simon, 1969;Metropolis, 1987;Fox-Keller, 2002;Lenhard, 2012, Winsberg, 2010Morrison, 2015;Kaminski, 2017). But this also means that experts in diverse communities without a common language must learn to communicate as they work together without a unifying framework (Gehring, 2017). These views are a direct response to the narrow understanding of computer simulations as extensions of formal methods.…”

“…As I briefly explained above in section 1.1, on the broad conception of computer simulations there are those that besides acknowledging the extra-formal elements of computer simulations also argue that the term refers to something even broader than just implemented models or experiments. As a result, some on the latter side of the debate understand computer simulations more as a set of experimental practices whose epistemic assessment ought to be construed taking into account sociological features such as community consensus among heterogeneous experts, engineering ingenuity, and other social implications surrounding the practitioner's world (Gehring, 2017;Gransche, 2017;Durán, 2017;. For this view, computer simulations are, in short, a practice.…”

“…Nevertheless, together with sociological analysis of computer simulation practices as well as insight from social (Gehring, 2017) and pragmatic epistemology (Hubig and Kaminski, 2017), the views presented above, are sometimes taken to be reasons to reject an understanding of computer simulations as instruments. The gist of their arguments is the following.…”

“…Although it is true, as many have noted (Resch, 2013;Resch et al, 2017;Gehring, 2017), that a fully articulated and unified definition of computer simulation is difficult to extract from a survey of the literature across diverse disciplines, an initial common-sense definition can be generated in a fairly straightforward manner when one regards them as instruments. In fact, this is one of the initial benefits of understanding computer simulations as instruments rather than as practices or extensions of formal methods.…”