Measurement, decomposition and level-switching in historical science: Geochronology and the ontology of scientific methods.

Philosophers of the historical sciences have focused to a significant extent on the problem of epistemic access facing these sciences: how do historical scientists overcome the relative scarcity of data about the past, compared to the present? Solving this problem usually requires solving another one, which I call the 'problem of ontic access:' how do historical scientists get access to entities and processes with properties that are potentially informative about the past? The case of geochronology illustrates one solution to this problem: historical scientists can get access to entities and processes with properties that are potentially informative about the past by exploiting the metaphysical structure of their domain. Geochronology experienced a spectacular explosion of its research boundaries in the 20th century. I explain this productivity by analyzing the ontology implicit in geochronological techniques. The productivity of isotope geochronology was based on (a) mereological decomposition in order to (b) exploit differences of properties obtaining between the parts and the whole, and (c) an exceptional complementarity between mass spectrometry and the lower-level properties, allowing application to a wide range of geological contexts. The technologically mediated ability of the scientists to exploit the metaphysical structure of their domain was crucial to their success.