Textualism, i.e. the search for an interpretation of texts faithful to their original age, is a new emerging trend in legal interpretation, but its practical implementation is often incomplete. What are the opportunities and risks involved in the spread of this approach? In this post, we examine the advantages and risks of the relationship between textualism, corpus linguistics, and legal corpus linguistics, and the exchange of methodologies between disciplines.

How did linguistics enter the courtroom?

Textualism is a legal approach that focuses on the original text and its “ordinary meaning” when interpreting laws and regulations. It plays a particularly important role in the US legal system, where precedential decisions can date back to a hundred years. Textualism has been used in several famous cases. For example, in US Supreme Court decisions where judges have sought to interpret the words of the Constitution or statutes in their original meaning. Such was the case of King v. Burwell in 2015, a case about the interpretation of certain provisions of the Affordable Care Act (Obamacare), where textualism played a significant role in the decision. In addition, the 2020 case “Bostock v. Clayton County” received a lot of press coverage, in which the Supreme Court used a textualist approach to decide whether the prohibition of employment discrimination extends to discrimination based on sexual orientation and gender identity. These cases highlight that textualism can be an important tool for legal interpretation, especially when the historical context of the text is decisive.

However, it should also be remembered that this methodology can only be effective if it is applied correctly. To make the pitfalls clear, let’s examine how data-driven linguistic research has evolved and what exactly it has to do with legal decision-making today.

Since the 1950s, linguistics has been dominated by the so-called generative approach. One of its pillars is the assertion that all native speakers can judge the well-formedness of a sentence by virtue of their innate linguistic competence. On this basis, linguists used self-generated example sentences for analyses of language use, a process known as introspection. However, such artificial examples were only able to describe a small slice of real language use. Already in the 1970s, a functional approach, now known as corpus linguistics, emerged and became widespread around the 1990s. Its aim was to make real language use the basis for research. It brought a data-driven turn in linguistics: linguistic phenomena were studied by analyzing large data corpora of representative and digitally stored texts. This has made it possible to explore linguistic patterns objectively using statistical methods, thus making linguistic research more systematic and reliable.

The results of corpus linguistics have not escaped the attention of the legal world, particularly regarding studies of the changes in the meaning of single expressions. The methodology of the field could describe accurately the nuances of the meaning of a word or phrase based on contemporary textual data. This has proved extremely useful in the USA for judges working with legal texts that are often more than a hundred years old. In essence, textualism, which emerged independently of linguistics, introduced the analysis of historical corpus as a methodological tool in jurisprudence.

As mentioned, the textualist approach, which is particularly prevalent in the Anglo-Saxon legal system, puts the original meaning first. This means that it seeks to interpret the meaning of legislation and precedents solely based on the language used at the time. Databases such as the Corpus of Historical American English (COHA) or the Corpus of Contemporary American English (COCA) are, according to textualists, fully capable of identifying the “ordinary meaning” of key terms. This, according to the proponents of the methodology, can help to ensure that legal decisions are based on an interpretation that is perfectly faithful to the original. This is particularly important in cases where the meaning of the original text of the law has changed over time and there is a need to understand the context in which the original text operated.

The textualist approach and corpus linguistics are thus closely related, as both approaches aim to analyze authentic language use. While textualism focuses on the interpretation of the original text of the legislation, corpus linguistics examines linguistic phenomena by analyzing real linguistic data, i.e. corpora. However, it is no coincidence that legal corpus linguistics has also gained currency as a separate concept.  In essence, legal corpus linguistics combines the above principles in the analysis of legal contexts, thus creating a new interdisciplinary field combining linguistic and legal analysis. However, legal corpus linguistics, and more specifically its practitioners, have been criticized for their superficial or even incomplete application of linguistic methodology, their ad hoc selection of texts, or their failure to provide a basis for conclusions.

Lawyer-linguists or linguists-lawyers?

The advantage of the textualist approach and legal corpus linguistics is that, when properly applied, they provide an objective and empirical basis for the interpretation of legislation. They help to avoid misinterpretations of historical texts, for example by identifying words with their present-day meanings. Such a methodology can be particularly useful when the original meaning of the text of a law or statute needs to be determined in a distant historical context. The analysis is based on the corpus, which allows the real characteristics and proportions of language use to be considered, thus ensuring that the interpretation of legal texts does not deviate from the original meaning. Textualism can thus also serve as a kind of safeguard to ensure that legal interpretation does not become arbitrary, but that the basic meaning of the text remains the basis for interpretation.

However, there are also limitations to the textualist approach and especially to legal corpus linguistics. One major criticism is that textualism ignores the pragmatic context, which (in linguistics) refers to the social, cultural, and communicative context of utterances. The pragmatic context helps to interpret the meaning of texts beyond the dictionary meaning of words. It considers the relationship between speaker and listener, the situation and time of communication, and cultural and social factors. By deliberately ignoring this context, a textualist approach can simplify the meaning of legislation, thereby giving a distorted interpretation in cases where it is important to understand the intention of the legislation. Ignoring pragmatic elements can be particularly problematic where the interpretation of legal texts is in the interests of a particular social group and ignores the social context in which the text was produced.

Moreover, legal corpus linguistics often fails to consider the highly complex nature of language use. For example, the meaning of words is determined not only by syntactic relationships (co-occurrence) but also by paradigmatic relationships (what words are used interchangeably for each term in a given context). The nuances of meaning and the relationships between co-occurring words are extremely complex and ignoring them can reduce the accuracy of legal analysis. The representativeness of corpora can also be questionable. Linguistic research often builds specific corpora to investigate a particular problem. In contrast, legal corpus linguistics often relies on only a few publicly available corpora, which may not be representative of all aspects of legal texts. This lack of representativeness may risk misleading legal analysis if the corpus does not adequately reflect the linguistic diversity and the different contexts of legal texts.

Another limiting factor is that a corpus, by its nature, contains only positive examples. Simply speaking, the absence of a term in a given textual context does not necessarily mean that it never existed. The lack of negative data means that legal corpus linguistics is not able to provide a complete representation of language use, and analyses often provide only a one-sided picture of the phenomena under study. In legal corpus linguistics, the analyses are mostly carried out by the judges themselves, without the involvement of a specific expert. Although judges are excellent in their field, they cannot necessarily master the analytical methods of a completely foreign discipline in the short time available for decision-making.

This also raises the question of the value of involving linguists in legal analysis, since this would effectively delegate the task of interpreting the law to another discipline. This kind of cooperation, while offering the possibility of more thorough and precise analyses, raises questions about the blurring of the boundaries between law and linguistics. The expertise of linguists can be useful in understanding linguistic patterns, but ultimately the responsibility for interpreting the law must remain with the lawyers.

The textualist approach and legal corpus linguistics have great potential for the interpretation of legislation and precedents, as they intend to analyze language use objectively and on an empirical basis. However, it is important to keep in mind all the methodological limitations and to pay attention to the importance of pragmatic context and representativeness, without which legal analysis and interpretation can be significantly distorted. The application of textualism and corpus linguistics to law offers promising possibilities, but only if used with due care and with an awareness of methodological limitations. It will be crucial for future legal interpretation practice to consider, in addition to the current principles of the textualist approach, the social and pragmatic factors that are necessary to understand the true meaning of legislation.

István ÜVEGES is a Computational Linguist researcher and developer at MONTANA Knowledge Management Ltd. and a researcher at the HUN-REN Centre for Social Sciences. His main interests include the social impacts of Artificial Intelligence (Machine Learning), the nature of Legal Language (legalese), the Plain Language Movement, and sentiment- and emotion analysis.