A Model-theoretic Account for Structural Realism

In this paper I will give a model-theoretic account for structural realism. I will introduce the so-called ‘hierarchy of models’ which can be used to formally analyse the structure shared by a ‘high-level theory’ and the ‘phenomena’. I will show that structural empiricism (as characterized by Bueno (1997, 2000) and Van Fraassen (1980)) cannot be accommodated in the hierarchy of models and show that this is ultimately due to the absence of a justified account for the structure of the ‘phenomena’. Next I will introduce Bertrand Russell’s causal theory of perception and show that this theory can account for the structure of the ‘phenomena’. Then I will show how, based on this account, structural empiricism can be accommodated in the hierarchy of models. After that, I will show that, by the addition of a layer called ‘external world’ to the hierarchy of models, we can accommodate structural realism in the thus obtained ‘extended hierarchy of models’. This will provide us with the notion of a ‘high-level theory’ being ‘structurally correct’, which means that the ‘external world’ can be embedded into a ‘high-level theory’. The extended hierarchy of models allows us to formally analyse certain relations between structural empiricism and structural realism. Based on these analyses we will see in what way structural realism can be considered to be an extension of structural empiricism. Finally, I will show how the extended hierarchy of models can be extended even further to account for the ‘problem of theory-change’ (Bueno, 2000).