In theories that idealize the object of study, falsity is inserted somehow. However, the actual propositions by which the idealization takes place need not be false at all. An example from physics illustrates that the Ideal Gas Law and Boyle's Law are respective idealizations of the van der Waals Law. The idealizational procedures involved in reasoning from the latter to the former can be repeated at a higher level of abstraction than that of the laws as we know these from physics textbooks. Thus, idealization and abstraction can be seen as relatively independent methods of reasoning, the one to be carried out with or without the other at the same time. This underlines Uskali Mäki's taxonomy, which shows that horizontal isolations in economic reasoning form a procedure completely different from vertical isolations. In contrast, however, to his taxonomy, I propose to define idealization as horizontal isolation. The lessons for the policy relevance of science – and of economics in particular – are that the use of ideal models does not necessarily imply a total lack of their external validity. Further, I show that abstraction in theorizing, under certain conditions, may increase the policy relevance of theories, rather than that it is decreased. theories tend to count the actual social world – where policymakers try to intervene – among their models more easily than concrete theories. Finally, this paper also deals with one very problematic aspect of the common use of clauses in order to hedge economics hypotheses. Many idealizational clauses have a propensity for imprecise reference, due to which it is impossible to judge the external validity of economic models. I shall indicate how this problem relates to issues of interdisciplinarity in social science.
Read full abstract