| In science, as in everyday life, multiple pieces of evidence that are diverse in character, or that arrive from different quarters, sometimes work together, pointing to the same conclusion. We tend to find this particularly convincing. But is our intuitive reaction correct? Does a notion of kinds of evidence, or of diversity in the character of the evidence, really have a place in rational inductive inference, or is it only, so to speak, the total quantity of evidence that matters? One label for the phenomenon of multiple sources of evidence working together for greater effect is "consilience". In this dissertation, I investigate the epistemology of consilience from two directions: first, by reconstructing the reasoning involved in an important scientific argument from Darwin's Origin that ostensibly appeals to consilience, and second, by attempting to formally model the epistemological value of consilience in probabilistic terms. |
