Pinpointing prescriptive impact

Abstract This paper presents a single-author case study which demonstrates that the statistical modelling technique change point analysis (CPA) can provide compelling evidence of prescriptive impact at an idiolectal level. It has been hypothesized that Late Modern English review periodicals consistently pushed a prescriptive agenda, and that this impacted language use ( McIntosh, 1998 ; Percy, 2009 ). A lack of empirical research has, however, left these claims unsubstantiated, partly because evaluating prescriptivist endeavours has proven challenging. Using a purpose-built 3-million-token idiolectal corpus spanning 7 decades, this paper reports that it is possible to discern a striking change in usage. Use of CPA enables this change to be located precisely, and correlated to the author’s exposure to a prescriptive review of her work. In demonstrating how effectively CPA can provide a sophisticated correlation indicative of causality, this paper showcases the suitability of this technique to the study of prescriptivism.