An Inquiry into the Epistemic Properties of Entrepreneurs' Theories of Action

Boundedly rational managerial actors struggling to process information often use a limited set of "theories of action," or simple rules. However, simple rules may have a hierarchical structure, with some simple rules guiding others. Assuming the existence of such "keystone rules," we establish the complexity of determining an efficient set, and therefore the necessity of using meta-heuristic approaches. We explore the development of "keystone rules" among entrepreneurs as a genetic algorithm, where the computationally hard problem of picking rules is solved by social calculation. We find that the emergent keystone rules among the observed entrepreneurs do not match existing "scientific" theories but have particular epistemic properties. The identification of keystone rules could fill a theoretical gap between the rational decision and the social construction perspectives.