BOILERPLATE SEMANTICS: JUDGING NATURAL LANGUAGE IN STANDARD DEAL CONTRACTS

Many corporate finance lawsuits involve the interpretation of commonly-used boilerplate contracts, the meaning of which is thought to be widely understood. In some cases, however, judges interpret these contracts in ways that upend market actors’ expectations about the meaning of terms and frustrate the presumed intent of the parties. Given this experience, and the legal profession’s long history with boilerplate, it is a source of frequent surprise that certain standard provisions continue to be used, sometimes almost verbatim, even after becoming notorious sources of conflict. A number of persuasive explanations have been advanced for this phenomenon, but this Essay argues that they are incomplete, and an overlooked additional factor helps explain the persistence of trouble-making language. Many boilerplate clauses that become the subject of controversy share a type of ambiguous semantic structure that linguists know well, but that we as lawyers are rarely trained to identify. This type of structure lends itself well to boilerplate, but contributes to confusion and opportunistic reading of contract language. Semantically ambiguous constructions can seem straightforward enough to a drafting lawyer on first read, but they contain multiple layers of often-hidden meaning that provide fertile ground for later disputes. Despite the confusion that these structures create, their ambiguities are difficult to spot and correct, especially using the interpretive processes that we lawyers are accustomed to using. Thus, even earnest attempts to correct problematic language can end up falling short. This Essay identifies these structures using three well-known boilerplate provisions whose interpretations have proved controversial. The Essay also discusses ways in which lawyers can learn to recognize these structures and suggests that algorithms designed to process natural language may be able to “see” them even when humans struggle to do so.