Internal Convergence: A(n Almost) Neglected Epistemic Virtue

Since the beginning of the 19th century, agreement between two independent theories in a particular prediction has been interpreted as strengthening both our confidence in the prediction and plausibility of the theories in question. This inter-theoretical coincidence has been considered an epistemic virtue in scientific practice as well as in philosophical reflections about it. However, there is an epistemic virtue at work in scientific practice that has received less attention from the community of philosophers of science, and which we aim to emphasize in this paper. It is the coincidence of predictions that appeal to the same theory, which we propose to call “internal convergence.” We will characterize internal convergence metatheoretically, and provide some examples showing how such an epistemic virtue is relevant in scientific practice. We will pay particular attention to the case of cladistics, where internal convergence plays a special role, as the theory is difficult to test empirically.