On theory development: Applications to the study of family formation

The further opening up of demography to related social science disciplines is a felicitous outcome, provided that this does not lead to a segregated jet of mono-paradigmatic approaches or explanations. Hence, we need to go beyond the anchored narrative approach. In the philosophy of science, several paths for such further theory integration are available, and the present essay makes use of two such approaches: Imre Lakatos's program of "progressive problem shifts" and L. Jonathan Cohen's "inductive knowledge of comparative reliability." An application is made to three theories of the so-called second demographic transition, by showing that (1) these theories are by no means mutually exclusive, (2) their mechanisms are often interrelated and synergistic, and (3) the plausibility of a theory may depend on a chosen subgroup or context in time or space. Given these properties, such partial theories or separate narratives are prime candidates for inclusion into a more overarching multiparadigmatic and multi-causal theory.