In an earlier paper by Glänzel and Schubert [Glänzel, W., & Schubert, A. (1988a). Characteristic scores and scales in assessing citation impact. Journal of Information Science, 14(2), 123–127; Glänzel, W., & Schubert, A. (1988b). Theoretical and empirical studies of the tail of scientometric distributions. In L. Egghe, & R. Rousseau (Eds.), Informetrics: Vols. 87/88, (pp. 75–83). Elsevier Science Publisher B.V.], a method for classifying ranked observations into self-adjusting categories was developed. This parameter-free method, which was called method of characteristic scores and scales, is independent of any particular bibliometric law. The objective of the present study is twofold. In the theoretical part, the analysis of its properties for the general form of the Pareto distribution will be extended and deepened; in the empirical part the citation history of individual scientific disciplines will be studied. The chosen citation window of 21 years makes it possible to analyse dynamic aspects of the method, and proves sufficiently large to also obtain stable patterns for each of the disciplines. The theoretical findings are supplemented by regularities derived from the long-term observations.
