One of the many things the discipline of law and language owes to Peter Tiersma is the description of a process that seems central for our legal culture: the transfer of legal authority to written language, a process he calls textualization, elaborated in his book, Parchment, Paper, Pixels (2010). This concept is, however, embedded in a broader and pervasive discussion of the extent to which “ordinary” language and legal language are the same and where they differ. It is an issue that Tiersma touched on in “Some Myths About Legal Language” (Tiersma 2006), and that I want to take up and develop a bit. Later I will return to the textualization phenomenon and related issues.
The issue for Tiersma is whether and to what extent the “language of the law” is different from “normal” language. There are difficulties in defining both ends of this comparison. Both are actually misleading myths in themselves. On the one hand, we cannot specify exactly what a “normal” language variety is. On the other hand, the “language of the law” is, as Tiersma has repeatedly pointed out, structured minimally into “operative,” “persuasive,” and “expository” genres (Tiersma 1999: 139–143). Restricting our discussion to the language of written statutes (an “operative,” norm-establishing written type of genre), traditional lore has it that the main differences and distinguishing features are on what linguists now call the “microlinguistic” level, which is the grammar of the language, including syntax, morphology, the lexicon, and phonology. The lexicon includes legal vocabulary—concepts special to law that appear to lawyers to be the defining characteristic of legal language. While the existence of domain-specific concepts is beyond doubt, specialized vocabulary does not exhaust a linguistic characterization of this genre. Tiersma discusses what may be termed a “positive” list of grammatical features that are observed in operative legal texts, concluding that such a list does not offer conclusive evidence that the legal genre is radically different from other written formal language. He notes: “It turns out that just about all the features attributed to legal language are also ‘characteristic of formal written prose’ and that this conclusion offers small comfort to those who claim that the language of the law is special and distinct from ordinary communication” (Tiersma 2006: 46).
There are two main points that can be added to this argument. The first takes off from the work of Biber (1995), who characterized text types (clusters of linguistic co-occurring features, which are then, in a second step, interpreted as notional genres) in terms of what occurs and what does not occur. Obviously, what is absent reflects choices among grammatical options for selecting, presenting, and focusing content. Certain structures highlighting affected entities tend not to be chosen in legal texts, such as prepositional passives or middle voice structures. It appears that such a list of “negative” features has not been given enough weight in a linguistically oriented characterization of legal language. It is an interesting theoretical issue for the theory of varieties whether the absence of such features would be used to define either linguistic markers (exclusively characterizing a genre) or features (restricted to a class of genres) as set out by Biber (1995: 28). Generally, these terms are applied to actually “occurring forms.” It is understandable that, in the age of computerized corpus linguistics, the focus is on positive features, whereas the absence of features is well-nigh impossible to handle computationally.
Tiersma himself pointed the way to an even more important repository of “negative” features that can serve as “markers” or “features” of legal discourse. He discusses a number of local interpretive conventions of a type represented by contrasting the following two examples:
John kissed John’s girlfriend
Buyer promises that Buyer will pay