Abstract
This paper analyses two methods commonly used to understand legal language: deontic logic and the analysis of concepts taken as fundamental for any one or more areas of the law (sometimes called the philosophical foundations of law project). In doing so I introduce what I call the phenomenon of linguistic regress, and I do so in order to show why and how these methods necessarily fail as theories of legal language. I argue, in short, that any form of content-determination of concepts or norms fails as a method for understanding legal language because of the phenomenon of linguistic regress. Given the failure of these two methods, I argue that legal theorists should instead adopt an approach that focuses on explaining the community-based use of commonly-recurring terms and phrases found in legal language - a use, furthermore, that is most commonly exercised by legal experts (of varying degrees of skills), operating in different and interacting contexts of expertise (judicial, legislative and scholarly), and more broadly, as members of a legal community not impervious to influence from other kinds of communities.