Against the New Fictionalism: A Hybrid View of Scientific Models

ABSTRACT

This article develops an approach to modelling and models in science—the hybrid view—that is against model fictionalism of a recent stripe (the ‘new fictionalism’ that takes models to be imaginary entities that are analogous to works of fiction). It further argues that there is a version of fictionalism about models to which my approach is neutral and which makes sense only if one adopts a special sort of antirealism (e.g. constructive empiricism). Otherwise, my approach strongly suggests that one stay away from fictionalism and embrace realism directly.

Acknowledgements

A shortened version of this article was presented at the Pacific APA meeting, San Francisco, 30 March to 3 April 2016. I thank my commentator, Susan Vineberg, for many constructive criticisms; and I thank the audience of the session for a critical discussion. I thank an anonymous referee of this journal for detailed criticisms.

Notes

1. Throughout this article, I assume that geometrical figures are abstract entities, not any physical things that we routinely use to show or describe them. In Frigg’s tripartite structure for modelling, the markings on a particular map belong to the ‘description’, while the geometrical figures the markings show or describe are the abstract model elements, and the actual streets and buildings in the city the real systems. However, if one rejects all abstract entities as an ontological commitment, then one can easily dispense the description talk such as line segments and dots describing geometric figures and just talk directly about line segments and dots on paper and their geometric relations.

2. Note that the names (or referring device) are not attached to the geometric figures directly. They do that by being printed or drawn by the side of, or close to, the markings that show or describe the figures on the map.

3. I say ‘most’, for there are exceptions to this claim, as we shall discuss separately later. There are ‘toy’ models or ‘fictional’ models in science that do not fit this claim here.

4. NB: this point remains valid if one rejects the abstract-entity conception of modelling and takes all models to be concrete physical objects (such as a scale model of the Rutherford atoms or diagrams on paper as the one shown above). Stripping the labels from the parts and removing the name of the model would make the ‘thing’ just another physical object or set of ink marks.

5. Much more liberty may be taken in the arts in how and what things and pictures can represent: we are here concerned only with scientific representation.

6. I thank Susan Vineberg for raising this point.

7. One must, however, be clear again that the problem for the antirealists is not whether it makes sense to regard the model sans RDs as existing. If it makes sense to be a Platonist on abstract entities while remaining an antirealist about hypothesized unobservable systems, then a model of quarks exists for such an empiricist while quarks are not regarded as real on a par with desks and chairs.

8. Some violence is done to van Fraassen’s view of models (the semantic view of theories), but hopefully no confusion arises from this.