<p>Defining the field of forensic linguistics is no easy task. Taken in its broadest sense, forensic linguistic work can examine topics as disparate as the language of statutes, authorship analysis, and police interview techniques. Research methods may be quantitative or qualitative, and while in some cases the data examined is itself “legal”, in others the connection of a given study to the legal system comes purely from its intended application. Scholars have put forward a number of conflicting definitions of “forensic linguistics” over the years which further complicate efforts to understand the field. In this article, I suggest that rather than imposing a top-down definition of what the field should be, the best way to understand the scope of forensic linguistics is by considering what studies that are broadly held to belong to field are actually doing. Based on a review of existing definitions, relevant studies, and consultations with colleagues who consider themselves to belong to the field, I propose that “forensic linguistics” is best defined as “that set of linguistic studies which either examine legal data or examine data for explicitly legal purposes”. Such a definition is useful not only because it captures the breadth of the work which is commonly understood to be “forensic linguistic” in nature, but also because it establishes a baseline which can be used for the further subclassification of forensic linguistic studies. To demonstrate that latter point, I show how all data examined by forensic linguistic studies can be classified according to whether it has a legal form and/or a legal function and how categorizing data in this and related ways provides novel insights into the nature of the field.</p>