Various interpretive theories recommend using dictionaries or corpus linguistics to provide evidence about the “original public meaning” of legal texts. Such an interpretive inquiry is typically understood as an empirical one, aiming to discover a fact about public meaning: How did people actually understand the text at the time it became law? When dictionaries or corpora are used for this project, they are empirical tools, which might be reliable or unreliable instruments. However, the central question about these tools’ reliability remains unanswered: Do dictionaries and corpus linguistics reliably reflect original public meaning?
This paper develops a novel method to assess this question. It begins by examining the public meaning of modern terms. It compares people’s judgments about meaning to the verdicts that modern dictionaries and corpus linguistics deliver about (modern) public meaning. Eight experimental studies (total N = 1,327) reveal systematic divergences among the verdicts delivered by ordinary concept use, dictionary use, and corpus linguistics use. For example, the way in which people today apply the concept of a vehicle is systematically different from the way in which people apply the modern dictionary definition of a “vehicle” or the modern corpus linguistics data concerning vehicles. Strikingly similar results arise across levels of legal expertise; participants included 999 ordinary people, 230 “elite-university” law students (e.g. at Harvard and Yale), and 98 United States judges. These findings provide evidence about the reliability of dictionaries and corpus linguistics in estimating modern public meaning. I argue that these studies also provide evidence about these tools’ reliability in estimating original public meaning, in historical times.
The paper develops both the positive and critical implications of these experimental findings. Positively, the results reveal systematic patterns of the use of dictionaries and corpora. Corpus linguistics tends to generate prototypical uses, while dictionaries tend to generate more extensive uses. This discovery grounds normative principles for improving the use of both tools in legal interpretation. Critically, the results support five argumentative fallacies that arise in legal-interpretive arguments that rely on corpus linguistics or dictionaries. More broadly, the results suggest that two central methods of determining original public meaning are surprisingly unreliable. This shifts the argumentative burden to public meaning originalism and other theories that rely upon these tools; those theories must provide a non-arbitrary account of these tools’ use and a demonstration that such methods are, in fact, reliable.
Donwload the article from SSRN at the link.