Taxonomy for the non-material experiment

The article examines the epistemological relations between the classical laboratory experiment, the thought experiment, and the computational experiment. In the context of the modern history of the philosophy and methodology of science from positivism to the so-called experimental turn and contemporary discussions of immaterial experiments, the question of the epistemological similarities and differences between material, thought, and computational experiments is raised, as well as the methodological specificity of the experiment as a concrete scientific method and a generic concept for this controversial, but de facto used taxonomy. A common feature of all quasi-experimental methods in scientific knowledge is their semiotic function as a means of ensuring objectivity, giving meaning to the formal structures of knowledge. The first section examines the so-called “experimental turn” in the philosophy of science, associated with the works of the Stanford School and the transition from understanding the experiment as “simply” armed observation to its interpretation as a practice of active intervention in reality and the producing of facts. The moment of “spontaneous realism” in experimental science and the presence, as noted by Ian Hacking, of a “life of their own” for experimental practices and the facts reproduced in them are emphasized. The second section is devoted to the epistemology of thought experiments. The arguments in favor of denying thought experiments’ “experimental nature” and recognizing them as a type of theoretical models that deal exclusively with logical consequences and logical integrity (consistency) of a theory are critically examined. Using the example of the EPR paradox and related plots in history of physics, the ability of thought experiments to create new knowledge and “live a life of their own” is emphasized, i.e., an ability to reproduce in different theoretical contexts and to give different results rather than only those supposedly fixed once and for all by their logical structure. The third section emphasizes that computational experiments and digital simulations are similar to thought experiments in their “immateriality”, but differ in the cognitive infrastructure used and in the transparency of obtaining results. While a thought experiment relies on the work of the imagination and provides the immediate clarity of obtaining a result, a computer simulation uses an “external” computational infrastructure and, due to the high complexity of models and calculations, makes the origin of specific observed results opaque to the researcher, which makes simulations closer to classical laboratory experiments. At the same time, the ability of modern computer simulations to model empirically non-existent objects, giving them observability, and to produce different results in different iterations emphasizes their methodological “experimentality” as sources of new quasi-empirical data. In conclusion, it is noted that a productive solution to the “taxonomic confusion” is the recognition of the essential epistemological kinship of material, thought and computational experiments, despite the exact degree of closeness has yet to be clarified. The presented material is intended for the lecture part of the courses Philosophy and Methodology of Science and Modeling, Forecasting and Expertise in Scientific Activity (taught, respectively, in semesters 3–4 and 7–8 to undergraduate students of Lomonosov Moscow State University’s Faculty of Philosophy), and is completely given by the authors within the framework of the course Experimental Practices in the Methodology of Social Sciences (Master's degree program of the Faculty of Philosophy).

methodology of science, experiment, experimental turn, philosophy of experimentation, thought experiment, computational experiment