Continuous Time Modeling in the Social Sciences: History and Philosophical Background

Against the backdrop of the methodological situation just after World War II with its strongly cross-sectional orientation, the introduction of continuous time modeling into social science by Simon and Coleman is depicted. Further development in the two directions of structural equation modeling (SEM) and state space modeling (SSM) is followed by the detour in SEM via the indirect parameter estimation procedure. It is observed in detail how the two merged into presently available flexible software, allowing totally arbitrary time intervals within and between subjects and thus missing values at arbitrary places. Recent developments discussed include hierarchical Bayesian analysis, time-varying parameters, and mediation analysis. Philosophical issues related to continuous time modeling are discussed, culminating in the question, which was raised already in antiquity, whether change in continuous time is possible at all.