Data sharpening and linguistic theorizing: a case study of the causative derivation of Urdu change-of-state verbs

Abstract

AbstractThis study examines the evidential value of corpus and introspective data from the perspective of morphological operations involved in the causative derivation of Urdu change-of-state verbs. By analyzing multi-source data on the transitivity status of causative variants of 112 Urdu verbs, it discovers a crucial relation between the type of evidence and the aspect of linguistic competence it addresses. More specifically, it compares data from lexical translation, Urdu WordNet and Urdu Lughat on the one hand, and data from a judgment task coupled with dialogical introspection on the other hand, and finds that the former can reveal the gradient nature of morphological productivity, but not its dynamic nature; the latter, however, can help explore the dynamic nature of productivity in the causative alternation. Such an observation confirms that both corpus and introspection complement each other, and that a particular source-specific piece of evidence may limit the coverage and generality of analysis. Thus, the study shows the importance of sharpening data by multi-source evidence for examining how various components of a phenomenon interrelate in the context of the overall grammatical organization of a language.