Thought Experiments and Computer Simulations

Abstract. The main purpose of this paper is to investigate some important aspects of the relationship between thought experiment (hereafter TE) and computer simulation (hereafter CS), from the point of view of real experiment (RE). In the first part of this paper, I shall pass in critical review four important approaches concerning the relationship between TE and CS. None of these approaches, though containing some important insights, has succeeded in distinguishing between CS and TE, on the one hand, and REs, on the other. Neither have they succeeded in distinguishing TEs and REs (Section 1-4). In Section 5, the paper briefly outlines an account of CSs as compared with TEs that takes REs as a central reference point. From the perspective of the analysis of the empirico-experimental intensions of the concepts of TE, CS, and RE\u2014considering their empirical content and actual performance within a discipline\u2014, the attempts to find a distinction in logical kind between TEs, CSs and REs breaks down: for every particular characteristic of one of these notions there is a corresponding characteristic in the two others. From an epistemological-transcendental point of view, the only difference in kind between TEs and CSs consists in the fact that any simulation, even a computer one, involves a kind of real execution, one that is not merely psychological or conceptual. In TEs the subject operates concretely by using mental concepts in the first person; in contrast, real experiments and simulations involve an \u2018external\u2019 realisation. As shown in Section 6, this manifests itself in the in the higher degree of complexity often found in CSs as compared with TEs