Good Law in the Age of AI: How Citation Systems Established in the 1800s Remain Fundamental to AI-Powered Legal Research

Every day, lawyers check whether their cases remain "good law" using tools whose basic design emerged over a century ago. This post, the second installment of our four-part series on the impact of AI on legal research, recounts the history of citators and their role in shaping AI-powered legal tools.

In the initial post of the series, we explored legal digests and their role in organizing case law by topic. But finding relevant cases is only the first step. Lawyers also need a way to ensure the cited cases remain "good law"— that they have not been overruled, questioned, or otherwise weakened. This is where the legal citator enters the story.

A citator is an index that allows attorneys to track the treatment of cases by later decisions and determine whether a case has been followed, distinguished, criticized, or overruled.

This post will look into the history of legal citators, highlighting the early work of Simon Greenleaf, the entrepreneurial vision of Frank Shepard, and the innovations of King & Leonard. We'll uncover how their combined efforts shaped the modern citator, a tool whose fundamental design has endured for over a century, and how that design fits with artificial intelligence.

The historical content in this post is based on two indepth journal articles: Patti J. Ogden, Mastering the Lawless Science of Our Law: A Story of Legal Citation Indexes, 85 Law Libr. J. 1 (1993) and Laura C Dabney, Citators: Past, Present, and Future, Articles by Maurer Faculty. 2113 (2008).

Simon Greenleaf: The Seed of an Idea

The story of legal citators begins with a young lawyer's embarrassing courtroom moment. In 1807, Simon Greenleaf was one of only about fifty attorneys practicing in the Maine territory. 

During a court appearance, Greenleaf cited an English decision that seemed decisive for his client's case. Opposing counsel informed the court that the case had been overruled. The court promptly declared the case "of no authority whatsoever."

This embarrassment sparked an idea. Greenleaf decided to create a resource that would help lawyers identify which cases had lost their authority through subsequent decisions. He began compiling a list of overruled cases.

Joseph Story, a U.S. Supreme Court Justice, took interest in Greenleaf's project. (Story also played an early role in the development of case digests. See discussion) Story not only provided his personal list of overruled cases but promised to continue sending supplements.

After several years of work, Greenleaf published "A Collection of Cases Overruled, Doubted, or Limited in Their Application" in 1821. The volume's straightforward format listed cases alphabetically with brief explanations of how and by what later decisions they had been affected.

However, Greenleaf's compilation had limitations that he readily acknowledged. The collection wasn't comprehensive—an impossible task for any individual given the growing body of case law. More significantly, he didn't consistently indicate which specific point of law had been overruled or limited in each case. Still, Greenleaf had planted the seed for what would become an essential legal research tool.

Frank Shepard: The Law Book Salesman that Taught the Profession to Shepardize

The rapid expansion of American case law during the 1870s made it increasingly difficult to track a case's authority. This was due to the growing number of published court decisions and the shift towards relying on American precedents over English opinions.

Frank Shepard, a businessman with experience in legal publishing, would become synonymous with legal citation checking. Shepard's tables of numerical citations were an innovation from previous citation indexes that used alphabetical lists with explanatory text. These tables were printed on gummed paper so they could be cut and pasted directly into case reporters. This allowed lawyers to quickly reference any later cases that cited the one they were reading.

The other citation indexes of the time seemed to be superior to Shepard's system because Shepard's lacked treatment analysis, which would indicate how a citing case treated the original case, and it didn't provide any information about the point of law being cited.

Shepard's system was successful despite its shortcomings. His business background allowed him to see the commercial potential of citation indexes. Unlike other compilers who only published a single volume, Shepard issued citation books for all states and provided regular supplements to keep his annotations current.

As one legal historian noted, "Uniform, reliable, and current citation information seem to have been the hallmarks of Shepard's annotation system, and these distinctions rescued what was otherwise a mediocre product."

King & Leonard: Innovation from a Small Texas Town

King and Leonard, two attorneys from Dublin, Texas, made significant contributions to citator design while Frank Shepard was building his citation product. In 1892, they published "Annotations of the Texas Court of Appeals Criminal Cases and Civil Cases." Their format was similar to Shepard's, with numerical columns, but they also added key innovations.

The addition of superscript numbers allowed researchers to quickly identify which cases discussed the specific legal issue they were researching. These numbers indicated which numbered point or headnote in the cited case had been referenced by the citing case.

They also used letter abbreviations to show how the citing case had treated the original case—whether it had followed, distinguished, or overruled the case. This treatment analysis was not included in early Shepard's publications.

In 1894, West Publishing became the "exclusive special agent" selling "Citation Manuals" for the National Reporter System, compiled by King and Leonard. These citators were designed to complement the West reporter series. However, the project vanished after this date, with no further references to new volumes or the National Citation Company. The venture likely failed due to the amount of expensive editorial work required.

Despite the lack of commercial success, King and Leonard's work, especially headnote references and treatment analysis, were significant advancements in citator design. 

These features would later appear in Shepards.The improvements borrowed from King and Leonard made Shepard's Citations significantly more useful. A reviewer in 1903 called it "quite a unique and colossal undertaking." The system became so established that "Shepardize" entered the legal lexicon as a verb meaning to check a case's subsequent history—a term Shepards would later register as a trademark.

The Digital Age and Beyond

The introduction of computer-assisted legal research in the 1970s marked the end of the print era's long reign in citation checking. With the launch of Lexis and Westlaw's competing computer research services, Shepard's citation data was licensed for online use. In 1979, Lexis introduced Auto-Cite as a supplementary service, and Westlaw followed suit in 1984 with InstaCite.

In 1997, West introduced KeyCite. This development was spurred by Reed Elsevier, Lexis's parent company, acquiring partial ownership of Shepards, which threatened West's access to citation data.

Initially, West estimated that developing a rival citation system would cost $300 million and take seven years because of the extent of the required editorial work. However, West later acquired Auto-Cite data when Thomson Corporation bought West and had to divest some overlapping products.

The time-consuming printing and distribution process made print citators inherently less current than online alternatives. Coupled with the tedious nature of using print citators, the shift to online citation checking has steadily decreased their use. Due to cost considerations and the availability of affordable online alternatives, many law libraries have canceled their print Shepards subscriptions.

Legal Citators and AI

The editorial content that has been produced over the last 100 years resulted in citators that provided a few legal publishers information resources that could not be matched. The nuanced analysis contained in citators was a “monumental” task taken on by an army of legal editors throughout the years.

AI’s ability to independently generate editorial content at scale alters the commercial availability of citation resources. More interesting than AI's ability to build upon the structure of citators that were set in the late 1800s, will be the new insights and relationships that AI will uncover as the AI constructs new citators.

Just like citators grounding research in “good law” remains an essential tool for lawyers to conduct legal research, AI requires access to citators to ground the AI in “good law” as AI progresses to automate legal research.  

Conclusion

The story of legal citators reveals how a simple idea—tracking which cases have been overruled—evolved into sophisticated systems essential to modern legal practice. Simon Greenleaf, responding to a courtroom embarrassment, created the first specialized index of overruled cases. Frank Shepard, with his business acumen rather than legal expertise, developed a system of regular updates that gave his product an edge over competitors. King and Leonard, from their small-town Texas practice, introduced innovations that enhanced the usability of citation indexes. Together, these pioneers established the fundamental structure of the modern citator: numerical citations, treatment codes, and headnote references.

What's remarkable is how enduring this structure has proven. Despite technological advancements, the core function and organization of legal citators remain recognizable to anyone familiar with Shepard's bound volumes or King and Leonard's pocket parts in the late 1800s.

While online citators have made citation checking faster and more convenient, they still operate within the framework established over a century ago. The next evolution in legal research—the application of artificial intelligence—promises to move beyond these established structures, changing how we understand and use legal authority.