All copying is not created equal: borrowed language in Supreme Court opinions.

• Author: Feldman, Adam
• Position: A. Introduction into III. Measuring Language Overlap F. Lifted Relationships (High Percentage Overlap and High Word Overlap

1. Introduction

   Imitation may be the most sincere form of flattery, (1) but in the language of the law, the demarcation between imitation and wholesale borrowing remains opaque. This is especially true in the Supreme Court's opinion-construction process. This article examines the implications of the Court's use of substantive language directly from merits briefs. To do this, the article critically analyzes the relationship between merits briefs and opinions, focusing on instances in which there is strong evidence that briefs played a substantial role in the Court's choice of majority-opinion language. The analysis compares phrasing from briefs and opinions and locates a case type in which briefs have played an especially influential role.

   While there is an expectation among those experienced in Supreme Court practice that language from briefs will seep into the Court's opinions, (2) there is considerable variation in the extent and types of overlapping language. Maybe due to this expectation, there is minimal scholarship dedicated to examining instances of high levels of language sharing between Supreme Court opinions and the sources of opinion language such as merits briefs, as well the normative consequences of this language-sharing practice. (3) To help explain the variation in shared language from case to case, this article contains a typology of relationships between briefs and opinions and presents examples of the different types before focusing on the brief-opinion relationship in which briefs play the greatest role: what I characterize as cases having "Lifted" opinions.

   The manner in which Justices and their clerks utilize language from merits briefs in Supreme Court opinions shows distinct practices across Justices' chambers and across time. By inquiring into the inputs for the information and wording in Supreme Court opinions, we may begin to better understand the opinion-construction process in a fashion that interviews with clerks and Justices alone cannot convey. (4)

   Recent evidence shows that from 1946 through 2013, Supreme Court opinions shared 9.55 percent of their language, on average, with individual merits briefs. (5) That analysis of almost 10,000 briefs elucidates several patterns in the ways that merits-brief language makes its way into Supreme Court opinions. The main foci of this article are cases in which the briefs have the greatest impact on the Court's opinions or, put another way, cases in which the language from a merits brief is most fundamental to the language of the opinion. (6)

   In many of these instances, the Court uses language from the briefs as its own without attribution. With these examples, this article confronts the normative question about when the Court should note that the opinion language does not originate with the Justices and their clerks. (7)

   In addition to examining specific brief-opinion relationships, this article also looks at the factors that tie the set of cases together. In this sample of cases there are common threads in terms of the attributes of the cases, Justices, and litigants. These shared features show that cases with high levels of language overlap are not entirely random.

   The underlying premise of this article is that the language in Supreme Court opinions matters. The article begins by looking at why opinion language matters and by developing a hypothesis for why we might expect the language in briefs to filter into Supreme Court opinions, sometimes in especially high doses. After this introduction and a short discussion of how the article tracks the linguistic similarities between briefs and opinions, the article compares the cases of interest with their respective briefs. To conclude, this article examines the normative implications of this language-sharing practice by focusing on the ways in which bringing it to light may affect the perception of the way the Court conducts its business.

   2. WHY FOCUS ON LANGUAGE

      The strength of the Supreme Court's precedents is only as iron-clad as opinion language permits. Time and again, issues left unsettled by the Supreme Court lead to new test cases that percolate back up to the Supreme Court. This can be illustrated by the Court's sinuous precedent in Fourth Amendment cases that define the exclusionary rule. (8) As language from past cases is repeated in subsequent cases, the specific wording of the opinions affects future decisions of the Supreme Court and lower courts. Indeed, when interviewed on the subject of Supreme Court opinions, Justice Scalia underscored the importance of opinion language:

      [T]he only important part about an appellate case is not who wins or loses; it's not ... affirmed or reversed. The important part is the opinion. And if you affirm or reverse for the wrong reason, you've done everything wrong.... [I]f you haven't made clear what your holding is, instead of reducing litigation, instead of making life simpler for courts and lawyers below you, you've complicated it. (9) Similarly, Justice Kennedy noted that "the law lives through language and we must be very careful about the language that we use." (10) With this understanding, many scholars take seriously the downstream effects of the Court's opinion language in guiding and constraining future decisions in the Supreme Court and other courts. (11)

      The source of Supreme Court opinion language is relevant to its institutional standing. (12) There is an expectation for the Court to use language that comes from the machinations of the Justices based on their understandings of the Constitution as well as from citations to existing law and secondary sources. (13) One of the main sources that provides potential language for Supreme Court opinions is parties' merits briefs. (14)

      Briefs organize and synthesize cases for the Justices and clerks, minimizing their need to turn to other sources. (15) They also distill the information so the Justices and clerks can focus on other case attributes. (16) Chief Justice Rehnquist underscored this point when he wrote, "The brief writer must ... bring order to [the case materials] by organizing--and I cannot stress that term enough--by organizing, organizing, and organizing, so that the brief is a coherent presentation of the arguments in favor of the writer's client." (17)

      Given that the Justices' own statements describe the briefs importance in organizing all of the case material for their digestion, it may not be surprising that there is directly overlapping language in some cases between the opinion and a brief. The extent of this relationship, however, varies immensely and studies have up to this point been generally devoid of such analysis because qualitative assessments--like the one that I discuss here--of the language shared between Supreme Court briefs and opinions are sparse. (18)

      This article relies on mixed methods, cutting between quantitative and qualitative ends of the spectrum. As I explain in the following section, I use quantitative methods to locate the specific cases and briefs of interest. I then analyze the content of the shared language to generate inferences surrounding these relationships in cases with the greatest percentage of shared language. (19)

   3. MEASURING LANGUAGE OVERLAP

      Text-analysis software allows for precise measurements of shared language. A drawback, however, is that the language from the two sources must be a near match, and as a consequence, the software does not pick up on shared meaning when the words differ. (20) I use WCopyfind software for the analyses on which this article is based. (21) This program compares two documents and reports the similarities in their words, both in terms of percentage of overlapping language and number of shared words. The user chooses and inputs certain settings to calibrate the requirements of language similarity necessary for the program to note an instance of overlap.

1. Program Settings and Examples

   Understanding the WCopyfind settings is essential to contextualizing what a high percentage of overlap in a Supreme Court case means. I use the default settings as others have done in studies looking at similar relationships of language in legal and political texts. (22) The first setting deals with the number of words in a phrase that must be similar for the program to mark it as an instance of overlap. I set the minimum phrase length at six words. This means that phrases under five words will not be indicated, while phrases of six words or more will be processed. The second significant setting deals with the percent of language commonality that phrases must share for the program to recognize a phrase as relevant. I set this to eighty percent, which allows for slight differences between phrases in which the majority of the content is the same. Finally, the program is set to allow at most two imperfections in the shared language, so it will not pick up phrases that overlap at eighty percent or more if there are more than two differences between them.

   These settings may be easier to visualize through a straightforward example. In Lawson v. FMR, (23) the program recorded this sentence in the opinion as overlapping with the petitioner's brief: "[The Report concludes]: 'Congress must reconsider the incentive system that has been set up that encourages accountants and lawyers who come across fraud in their work to remain silent.'" The bracketed words were not shared with the brief (the brief and the opinion referring to the same document using different phrasing for its title), but due to the more than eighty-percent similarity, the program marked the entire passage as overlapping, and highlighted the words that did not overlap. Lawson also contains this language: "[the provisions require] accountants and lawyers for public companies to investigate and report misconduct, or risk being banned from further practice before the SEC." (24) The respondent's brief has similar--but not exactly the same--wording: "[u]nder...