Dissertation

Time’s Deep Rhythms: Models, Mechanisms, and Narratives in Historical Explanations

My dissertation is focused on the question of how it is that explanation works in the “historical sciences” (i.e. scientific disciplines dealing with questions of the deep past). I survey the extant literature in order to develop a loose taxonomy of approaches as developed up to this point. While there are some deep differences in these approaches, at least one common theme emerges: the overwhelming majority of philosophical (and scientific) work on historical narrative explanation understands such explanations to be focused on highly particularized explanatory targets. That is, narrative explanations are understood to involve the reconstruction of highly particular causal histories that are supposed to explain the generation of a highly particular explanandum of interest. The majority of the literature, then, has (wittingly or unwittingly) upheld the distinction between the so-called nomothetic sciences (e.g. physics) and the idiographic sciences (e.g. paleontology). The nomothetic sciences, it is supposed, offer explanations in terms of laws and regularities, whereas the idiographic sciences explain by “depicting” narrative structures. Further, the idiographic sciences, insofar as they explain, supposedly do so in a way that is impoverished relative to the nomothetic sciences. I think this distinction rests on several mistakes, which I attempt to remedy in my dissertation.

As a solution to this problem, I argue for a conception of narrative explanations as structural-mechanical models, and at the core of this account is a concept I term mechanistically generated contingency (MGC). A key motivation for maintaining that historical sciences are idiographic is that they so often deal with contingent events. What MGC shows is that this motivation is misplaced. With this model-based conception of the nature of narrative explanations combined with MGC, we can appeal to the explanatory power of regular mechanisms underlying contingent evidential traces, which gives us a much deeper understanding of just how central causal regularities are to explanation in the historical sciences. This brings together several literatures that have been mostly isolated from one another up to now. More specifically, it brings together and synthesizes work on narratives in science, model-based explanation strategies, and the explanatory uses of mechanistic thinking. By putting these pieces together in the right way, we see that there is no good motivation for upholding the bifurcation between historical/narrative explanations and other kinds of scientific explanations. Properly analyzed, explanations in the historical sciences turn out to be model-based, mechanistic explanations that make free (and essential) use of causal regularities, and are epistemically on par with scientific explanations across the spectrum of scientific research.

An early version of the work done in my dissertation has been published in Philosophy of Science. My views on many of the central issues tackled by my dissertation have evolved since that publication, though it still captures many of the basic commitments of my account. Further, I’ve been invited to submit an article for a special issue of the Journal of Philosophy of History, wherein I extend my analysis of historical explanation into the world of archaeology and anthropology (more specifically, the scientific debate on how the Americas first came to be populated (i.e. the “Clovis First Debate”)). In that project, I develop an account of what I term speculative narratives, and explore their epistemic and non-epistemic benefits in assessing the credentials of purportedly explanatory narratives.