Abstract

Nuclear models are incompatible in their thing-terms: terms that refer to static entities like objects, structures, and substances. Specifically, the two most prevalent models—the liquid drop and shell models—treat the nucleus, its internal structure, and the component nucleons as entities that contradict each other’s properties. These differences allow these two models to describe and explain different nuclear experiments: fission and scattering in the liquid drop model, and single-nucleon excitation and nuclear decay in the shell model. However, prima facie, these differences also suggest that these models are incompatible in their ontology. However, by taking seriously the experimental methods by which these models are constructed and the calculational tools these models provide for interpreting experimental outcomes, I construct a new form of realism about these models that renders them ontologically compatible. Namely, I argue that nuclear models are consistent in the dynamic entities to which they refer. I therefore advocate a pure process realism with respect to nuclear models. Critical to this process realism is the recognition that the processes referred to within nuclear models are essential parts of the observation acts that form nuclear experiments. In particular, because the dynamics within the nucleus must always be a continuous intermediary of our experimental interventions and the receptions of signals from the system, these dynamics are essential dynamic parts of nuclear experiments. We are therefore licensed in inferring these dynamic parts on the basis of experimental practice alone. In contrast, the thing terms reified by the thing realist in these models require additional inferences, the premises of which cannot be supported on the basis of experiment alone. Thus, inferences to processes are experimentally supportable, whereas inferences to things are dubious at best.