Generality is a defining modelling virtue of the economic tradition. It is often assumed that generality is not only a normative standard but also a costless commitment. This paper takes up a criticism of this assumption: degree of generality of a model is a choice which comes with costs under certain conditions. It is argued that if economic systems are heterogeneous, then generality trades-off with empirical fidelity. Trade-offs have the net effect of supporting virtue pluralism. The degree to which trade-offs bite, and thus the degree of virtue pluralism, depends on the degree of ontological heterogeneity of economic systems. Generality as a normative standard is undermined because it is a gamble on ontology