The Supreme Court's Language Goes Its Own Way
Often criticized for its perceived political leanings, the U.S. Supreme Court has indeed developed a language in its opinions that is markedly different than that of the next-lower federal courts, new research co-authored by Professor Michael A. Livermore of the University of Virginia School of Law has found.
The finding will be published late 2017 in the Arizona Law Review. The paper, "The Supreme Court and the Judicial Genre," is part of the ongoing computational research Livermore is conducting in association with collaborators at Dartmouth College.
The researchers used an algorithm to compare trends in language between the Supreme Court and the U.S. Courts of Appeals between 1951 and 2007. Their "topic modeling" method can look at a body of text, detect co-occurring words, and then identify the subject-matter categories in specific opinions based on the words they contain.
The "marquee finding" of the analysis was that Supreme Court opinions "have become increasingly distinctive over time," Livermore said. The project defined "distinctiveness" with respect to the topic categories, rather than stylistic measures such as the number of footnotes or the length of the document.
The research builds on earlier social scientific work that suggests that part of the way the Supreme Court maintains high levels of popularity with the public is by distinguishing itself from political institutions, such as Congress. Part of the way the court does this is by issuing its decisions in texts that are recognizable as judicial opinions. The authors suggest that the Supreme Court may lose some of this legitimacy if its link with lower courts becomes too tenuous.
Part of the motivation for the research was to apply quantitative tools to questions that, until now, have been difficult to study in a comprehensive fashion.
"Some people may have had an intuition that the court's opinions are becoming more political, or the justices are writing in a less-traditional style; that doesn't seem like a crazy thing to hypothesize," Livermore said. "At the same time, somebody can always come back and say, 'Where's the evidence of that?' It's really hard using traditional methods to identify these trends, and there is always a risk of cherry-picking anecdotes. Our analysis bring quantitative tools to the table."
While the language is more distinctive, the researchers found, the types of cases chosen by the Supreme Court for review are not.
"It's about as easy to distinguish a case selected for review now and a case selected for review in 1950," Livermore said. "It would be plausible to think that maybe the court is grooming its docket more carefully, or selecting stranger cases, but we don't find evidence of that happening."
Livermore's co-authors are Daniel Rockmore of Dartmouth College and Allen Riddell at the University of Indiana (who complete work on the project while a post-doctoral researcher at Dartmouth). Livermore and Rockmore have published previous research findings in the paper "A Quantitative Analysis of Writing Style on the U.S. Supreme Court," published by the Washington University Law Review in 2016.
- "Is the Supreme Court Acting Less Like a Court?" (The Conversation)