Explaining with Models: The Role of Idealizations

Abstract

Because they contain idealizations, scientific models are often considered to be misrepresentations of their target systems. An important question is therefore how models can explain the behaviours of these systems. Most of the answers to this question are representationalist in nature. Proponents of this view are generally committed to the claim that models are explanatory if they represent their target systems to some degree of accuracy; in other words, they try to determine the conditions under which idealizations can be made without jeopardizing the representational function of models. In this article, we first outline several forms of this representationalist view. We then argue that this view, in each of these forms, omits an important role of idealizations: that of facilitating the identification of the explanatory components within a model. Via examination of a case study from contemporary astrophysics, we show that one way in which idealizations can do this is by creating a comparison case that serves to highlight the relevant features of the target system.

Acknowledgements

Previous versions of this work have been presented at the Society for Philosophy of Science in Practice conference, Toronto, June 2013, and at the Biennial Meeting of the Philosophy of Science Association, Chicago, November 2014. We thank the audiences for their comments. We would also like to thank Anouk Barberousse, Aki Lehtinen, Mauricio Suárez, and two anonymous reviewers for this journal for their insightful comments.

Notes

[1] Non-representationalist accounts of model explanation have also been developed (Elgin 2007, 2009; Knuuttila and Merz 2009; Knuuttila 2011; Kennedy 2012).

[2] Note that McMullin (1978, 1985) describes a kind of explanation that is structural. He writes, ‘When the properties or behavior of a complex entity are explained by alluding to the structure of that entity, the resultant explanation may be called a structural one’ (McMullin 1978, 139). McMullin takes this kind of explanation to be causal, since the structure that is identified is generally the cause of the feature that is being explained.

[3] That said, Morrison (2005, 2009) showed that there is a theoretical inconsistency in the de-idealization thesis. According to the de-idealization thesis, if a model can be de-idealized, then it is a good approximate representation: it is approximately true. And the more we de-idealize the model, the better the model explanation will be. However, this is only possible if there is a stable ‘structure’, or set of approximately true assumptions, that remains constant throughout the de-idealization process. As Morrison points out, the problem is that such a structure does not always exist. In some cases, a model that results from a de-idealization process will contradict the model from which it was derived, which seems to show that there is no constant underlying structure that survives the de-idealization process. While we agree with Morrison that there does seem to be a theoretical inconsistency in the de-idealization thesis, we think that the following question is still worth addressing: in practice, even if successive de-idealized models conflict with each other, is de-idealization always explanatorily beneficial? We argue, in the remaining of the article, that a de-idealized version of a model is not always in itself explanatorily beneficial.

[4] While many philosophers of science focus on the epistemic problem raised by simplifying idealizations in models, some have focused on their role as cognitive aids (Dilworth 1992; Hartmann 1998; Forster 2001; Teller 2001; Elgin 2007, 2009; Morrison 2009). In particular, Suárez (2009, 2010) contends that fictional assumptions play an inferential role in scientific modelling. On his view, the primary function of a scientific model is not (only) to faithfully represent the target system but also to provide ‘inferential shortcuts’ from which we can access the properties of the target.

[5] Photoionization refers to the state of complete (or near complete) ionization of hydrogen.

[6] The Lyman alpha forest is an absorption phenomenon seen in the spectra of high red-shift galaxies.

[7] Contrastive explanations were first systematically developed by van Fraassen (1980) and Garfinkel (1981).

[8] The comparative work that we have described in our examples is not limited to de-idealization. Such comparative work might be also done by scientists when they replace the laws in their initial model by ones that are considered to be more fundamental, in an effort to improve their models.

[9] We use an example from physics, however, explanation by comparison occurs in other disciplines as well, such as economics (see Lehtinen and Kuorikoski 2007; Lehtinen 2013 for examples).