Usage-based linguistics (UBL) incorporated probability as a theoretical notion right from the start, making frequency not just an arbitrary performance phenomenon but a crucial component of the inner workings of mental grammar. This chapter discusses why and how statistical methods have become increasingly more prominent in UBL studies. More specifically, it discusses the “why” and outlines how the very nature of UBL—its commitment to certain kinds of data and its central notions and mechanisms—virtually requires a certain amount of statistical methodology. After that, the chapter discusses examples of statistical applications scholars find in current UBL work in approximately ascending order of complexity: frequencies of (co-)occurrence and association measures; predictive modeling (e.g., regressions and other approaches); and exploratory methods. Finally, each of these three areas is revisited with an eye to discussing necessary next steps that UBL could benefit from.
© 2001-2025 Fundación Dialnet · Todos los derechos reservados