Lexical learning and language change: raising in English from the perspective of acquisition and diachrony

Abstract Raising and control structures in Present-day English are surface-identical, yet differ in crucial ways both syntactically and semantically. We study the difference between the two from a diachronic and acquisitional perspective. Historically, raising has been argued to have emerged out of control (Higgins 1990). Similarly, raising has been found to be developmentally delayed compared to control (e.g. Wexler & Culicover 1987). We use quantitative methods to study the spread of TO -infinitives in child-directed and child naturalistic speech. In parallel, we study the same process in Early Modern English and what distributional properties may have facilitated the emergence of the distinction. Furthermore, we put to the test a learning model, the Tolerance Principle (Yang 2016), against the diachronic data. We argue that the emergence of raising historically was most likely a consequence of the broader rise of the requirement for an overt subject – an argument that has a clear parallel in recent research on the acquisition of raising using the same methods (Irani 2019).