Systemic Functional Linguistics and Computation

Systemic-functional linguistics (SFL) has a long history of interaction with computational linguistics, but no area stands still. In the 1980s and early 1990s, various computational systems for natural language generation informed by SFL achieved state-of-the-art performance. Subsequently, however, advances in statistically driven, rather than rule-based, computational linguistics eclipsed many of these earlier systems. In this contribution, we focus on several areas of contemporary computational applications of SFL to show how the field is now developing. We begin with a brief historical introduction to the context of this interaction, moving on with characterizations of the current state of the art and a consideration of the relationship between SFL and the statistical paradigm now central to computation and natural language