Data-driven approaches to syntactic research

Abstract

In recent decades, the nature of data used in linguistics has evolved considerably. Relative ease of access to large corpora and various databases, in addition to an increasing number of experimental methods, has enabled new approaches to traditional issues in linguistics. In this presentation I will describe some of my research and how it has benefited from various types of data. The first research area is differential argument marking, and I will outline the work carried out in two Himalayan languages: one study focuses on the optionality of ergative case on subjects of transitive clauses in Bumthang (Bhutan), the other on the optionality of dative case on objects in Nubri (Nepal). Both languages are little-known, and the data were collected in the field. It makes use of traditional elicitation methods, but also incorporates data from a variety of elicitation genres, in which the key data were found. Follow-up studies are being carried out using more targeted ‘experimental’ approaches, which I briefly outline. The second study I will discuss focuses on a question of linguistic history and particularly in determining linguistic areas. This is interdisciplinary joint work, comprising specialists in morphosyntax, phonology from the linguistics side and an archaeologist. The linguistic hypotheses were investigated using large databases to identify linguistic areas, setting up some claims about possible areas. These were then compared with what is known from archaeology and corroborated by data from outside the field. The ability to triangulate results using multiple data types, and also drawing on data from outside the field, enables us to be more confident about our claims.