Computational Power and Model Selection

Explore the complex relationship between computational power, background assumptions, and model selection in scientific discovery through this philosophical seminar. Examine the commonly cited example of Kepler's discovery of Mars' elliptical orbit and challenge the conventional narrative that suggests unlimited computational power would have prevented this breakthrough by allowing him to fit epicycles to his observations. Investigate how this scenario oversimplifies the role of computational resources in scientific methodology while ignoring the crucial influence of background assumptions beyond raw data. Analyze the intricate interplay between empirical data, computational capabilities, and theoretical frameworks that actually drives model selection in scientific practice. Consider how philosophical hindsight and sufficient computational power are often necessary to distinguish between alternative explanations and understand the assumptions underlying specific scientific choices. Reflect on the implications for how philosophers of science should approach and interpret stories about model selection, recognizing the limitations of oversimplified narratives about the relationship between computational power and scientific discovery.