Models - Predictions - Data: An (Un)problematic Relationship?

Abstract

Reflecting the ongoing debate on whether linguistics is an empirical discipline, Erich Steiner's chapter addresses the question of whether Systemic Functional Linguistics in general, and the Cardiff version of it associated with the work of Robin Fawcett in particular, are empirical disciplines. The author argues that many functional linguistic models are not empirical disciplines as prototypical sciences like physics, chemistry, or biology and that compared with Halliday's Sydney Grammar the Cardiff Grammar may have more transparent operationalizations, but it also has more limited scope in terms of domains of modelling and of application. The chapter concludes that becoming exclusively empirical may not be desirable for Systemic Functional Linguistics and for any socially responsible linguistics.