Experimenting with computing

We distinguish two kinds of experimental activity: post-theory and exploratory. Post-theory experiment enjoys computer support that is well-aligned to the classical theory of computation. Exploratory experiment, in contrast, arguably demands a broader conception of computing. Empirical Modelling (EM) is proposed as a more appropriate conceptual framework in which to provide computational support for exploratory experiment. In the process, it promises to provide integrated computational support for both exploratory and post-theory experiment. We first sketch the motivation for EM and illustrate its potential for supporting experimentation, then briefly highlight the semantic challenge it poses and the philosophical implications.