Brooklyn Law Review


Judges and lawyers often appeal to the “ordinary meaning” of the words in legal texts. Until very recently, claims about the ordinary meaning of words in legal texts have not been informed by evidence of the way that words are used or understood by ordinary people. This is because no such evidence—and no method to gather such evidence—was available. Instead, judges, parties, and scholars have been left to rely on their own linguistic intuitions and dictionaries, both of which are problematic guides to the usage or understanding of ordinary people. This symposium on Data Driven Interpretation focuses on recent developments that have challenged these traditional limitations: The rise of the use of linguistic corpora in legal interpretation, and recent scholarship evaluating interpretive claims with survey methods. Corpus evidence and survey evidence may also disagree, providing inconsistent answers to what appear on the surface to be similar questions. This has led some to conclude that the corpus-based answers to these interpretive questions must be incorrect. But where survey responses and corpus-based evidence differ on similar questions, we may have reason to prefer the corpus-based answer because it is premised on evidence of natural language that was produced in a natural communicative setting. At the very least, where survey-based answers and corpus-based answers to interpretive questions differ, interpreters should very carefully interrogate the reason for that difference. At bottom, both corpora and surveys can provide evidence of linguistic use and understanding that was not previously available to interpreters. This essay will explore some of the comparative advantages and disadvantages of using corpus methods or survey methods in legal interpretation.