Abstract
As microeconomic calculus and macroeconomic estimation superseded earlier approaches to political economy, broad questions about how things are (ontology), how things might be known (epistemology), and how science should proceed (methodology) were neglected. As a corrective, Critical Realism (CR) has been proposed as an alternative to the orthodox deductive‐nomological (ODN) tradition; i.e., to mathematical deduction and statistical induction. In their place, retroduction—the use of analogy, metaphor, intuition and ordinary language—is supposed to illuminate root causes by identifying the deep mechanisms that govern events. CR offers guidelines for social science that are of a most general kind: from initial “premises,” retroduction proceeds to hypotheses about deep structures and mechanisms. The initial premises are determined by a desire to understand events that surprise us. However, nothing is thereby excluded, including ODN. And since historical processes are revealed neither by assumption nor by the net effects of whatever initial conditions hold, it might be apposite to drop the search for (deep) socio‐economic laws and to use whatever evidence is at hand to see whether, and the extent to which, ideal types apply to any given historical sequence.