Ahistoricism in Time-Series Analyses of Historical Process: Critique, Redirection, and Illustrations from U.S. Labor History

Most time-ordered analyses of historical processes are rendered because the premises that direct quantitative inquiry obscure changing historical relations through time. These conventional practices stem from three problematic presuppositions: (1) the separation of theory and history, (2) ahistorical conceptions of time, and (3) methodological autonomy and the primacy of the theory of statistical technique. We illustrate how these premises shape sociological-historical investigation by examining several quantitative historical analyses of labor organization and conflict. Reanalysis of the same historical series with temporally moving covariance analysis demonstrates theoretically important historical contingency as well as the error of presupposing temporal homogeneity in historical relationships. Our critique of conventional practice and our reanalysis enable us to pose alternative premises to ground quantitative historical research in the interdependence of history and theory, in historical conceptions of time, and in the historicization of quantitative methodology. Our redirection aims at moving us closer to historical actuality, that is, to historical relations and processes, when we use time-series data and analytic procedures.