Kevin Clarke and David Primo begin their book on the role of models in political science with a statement about models' importance, they write: ``[m]odels have come to be the dominant feature of modern political science and can be found in every corner of the field" (Clarke and Primo 2012, 1). They go on to clarify just how far they believe models have spread: ``[w]e do not mean to imply that all of political science concerns models. Normative theory is one example of an area where models are less central" (Ibid., 1n). Unfortunately, Clarke and Primo are quite mistaken on both claims. The first claim, that models have come to dominate modern political science, is certainly true in the sense that models are everywhere in modern political science. They're wrong, however, because it was ever thus. Only a too-restrictive understanding of just what a model is would yield such an historically minimizing view of the disciplinary trajectory. So too would one need a restrictive and unreasonably mathematical understanding of what counts as a model to identify normative political theory as a clear standout in the several fields of political science.
Primo and Clarke are generally correct, however, in their claim that models feature prominently in the social sciences and political science in particular. Given the prominence of this near-ubiquitous tool, it is surprising that the model itself receives relatively little attention from thinkers concerned to better understand the foundations of political enquiry. This paper is an effort better understand the foundation, structure, and role of models for analytically parsing the political world. To that end we require a broader conceptualization of the model than political scientists concerned to discuss their import generally employ, one that takes seriously their extension to normative political philosophy.
I proceed by arguing for the following claims: first, models are ubiquitous precisely because they are the fundamental tool with which social and natural scientists seek to understand and explain the world. If our goal as social scientists is to understand and explain human interaction, we cannot but simplify the social world through abstraction and idealization. Second, this ubiquity extends to normative theory. Third, the ways in which normative theorists employ models sheds important light on our understanding of just what a model can be, and how social scientists ought to think about them.
The implications of this unified approach to thinking about theory as a modeling activity are several. Important among is the reorientation toward thinking about things like thought experiments in normative arguments as purpose-specific attempts at modeling particular aspects of some theoretical investigation. The veil of ignorance is then an attempt to model intuitions about the legitimacy of threat advantage, demographic attributes, and even motivation in distributing the benefits of cooperative society, for instance. Another important way in which this scientific mode of thinking might benefit normative work is to push us to attend more systematically to how errors in theorizing and application ought to feed back into our theoretical efforts themselves.
