Abstract

After the historiographic revolution in science studies of the 1960's, philosophers began to envision science as a product of historical and sociological forces and not as the result of the 'scientific method'. While the importance of experimentation is recognized in recent science studies, the full significance for its roles in scientific practice generally remains overlooked. Therefore, attempts to reconstruct narratives of scientific practice are often incomplete. I propose several features of experimentation, as well as an associated empirical methodology, for analyzing the roles of experiment in scientific practice. ;To realize the significance of experimentation for scientific practice requires a close inspection of the experiments designed and performed by scientists and of the data and observations obtained from those experiments. To that end, the experiments critical for the discovery of thrombin by Buchanan and Schmidt in the nineteenth century and for the discovery of heparin by McLean and Howell in the twentieth century are examined in detail. These case studies provide a tangible framework for engaging in a series of philosophical reflections on the roles of experiment in empirical science. ;From the analysis of the reconstruction of this historical narrative, I suggest that several characteristics of experimentation--controllability, reproducibility, plasticity, and fecundity--are important for understanding how scientific practice leads to the discovery of novel entities within the natural world and ultimately to scientific progress. In addition the notion of horizons for scientific practice, representing particularly the experimental and theoretical limits within which scientists ply their trade, is introduced to facilitate the analysis of scientific discovery. And the notion of progressive horizons, in which the practice of scientists intersects dynamically with 'the way nature is', is also introduced to analyze scientific progress. ;Finally, the notion of an empirical methodology, a notion that may be applicable more generally to other experimental sciences, is introduced to further the analysis of the roles of experimentation in the case studies presented herein. Briefly, the methodology indicates how experiments mediate between theories and natural phenomena, for experiments form the intersection between theory and nature. That mediation I denote as 'tampering' with the natural world, in which the measure of a natural phenomenon is taken through experimental activity