Quantum chemistry and the theory-driven conception of science

JUAN CAMILO MARTÍNEZ GONZÁLEZ; HERNÁN ACCORINTI

Quantum Mechanics: Theory, Analysis, and Applications

NOVA science

Año: 2018; p. 197 - 217

Over the last decades the notion of model has gained interest and philosophers ofscience have recognized that models constitute a fundamental methodological resource ofmodern science. In this context, the following question acquires a special relevance: whatis the relationship between theories and models in science?According to the traditional conceptions of scientific theories, theories apply toparticular situations through the use of specific models. This theory-driven conceptionsupposes that models depend on theories: a model is always a model of a certain theory.Thus, any approximation introduced to build the model should be legitimated by thetheory, since only in this case the supporting evidence can be transferred from the modelto the theory.This latter position was challenged by the so-called ?toolbox? conception ofscientific theories, which calls into question the dependence of models on a specifictheory. The arena of the debate between the two views was mainly London brothers?model. But the discussion around it model has reached a kind of dead end.The present work tries to provide a fresh perspective to the debate by considering thecase of the molecular models used in quantum chemistry. Those models integrate twoincompatible theoretical domains: quantum and classical. Indeed, the ?clamped-nucleusapproximation? (which assume the nuclei as classical particle), introduced by BornOppenheimer, is incompatible with the Heisenberg uncertainty principle, which establishes the impossibility to assign simultaneously a definite position and a definitemomentum to a quantum particle.The analysis of the molecular models of quantum chemistry shows the independenceof models from theories. We believe that the tolerance of theory incompatibility reveals aconceptual independence that is constitutive of the modeling process. The existence ofmodels that integrate incompatible theories constructively and in an empiricallysuccessful manner provides an argument to call into question the traditional view,according to which a model is always a model of a certain theory, on which it depends.