Casetext CARA subscribers have recently seen two new  search enhancements appear on the CARA Platform. These new features ” Black Letter Law” and “Holdings”can help a researcher quickly locate the fundamental principles governing an area of law  in order to bolster a brief.

The Black Letter Law feature was created by harvesting all of cases which identify accepted principles of law which are  introduced with the phrases ” it is well settled,”  ” it is well established,”  “It is axiomatic.”  The “black letter law’ feature has one limitation and that is identifying black letter law related to completely new concepts e.g. blockchain.

The Holdings database leverages judicial summaries of cases and compares similar holding across an area of law. Search results provide the name of each case, the citation and a concise case summary. Results can be filtered by jurisdiction.

In The Beginning. The Pablo Arredondo, one of the co-founders of Casetext provided me with some background on the development of these features. The initial idea behind the Black Letter Law and Holdings database goes all the way back to 2011. While studying law at Stanford Arredondo had recognized the  value of “explanatory parentheticals” in judicial opinions.  Arredondo taught himself some basic coding so he could create a proof of concept: “Through a lot of trial and error, I was able to mine tens of thousands of principles and hundreds of thousands of case summaries. The prototype was posted on the web using a domain we called “well settled.” When Arredondo showed the prototype  to Paul Lomio,  who was Director of Stanford’s Law Library at the time, who suggested that Arredondo reach out to another Stanford alumni.  Jake Heller had just embarked on a legal research start-up. Heller and Arredondo met a few weeks later and they went on to build Casetext and the suite of CARA enhancements.


Black Letter Law

The Holdings Database Results


The limits of AI.  According to Arredondo,  AI still can’t summarize a case’s holding…  Or in Arredondo’s words, “judicial intelligence” still beats “artificial intelligence” in being able to summarize case holdings. CARA uses legal informatics patterns which are unique to the common law.  Phrases like “it is well settled” have appeared in common law case,s in all common law jurisdictions for hundreds of years. The folks at Casetext have calculated that it would take a judge over 350 years to read all the cases and write the case summaries which are included in CARA’s holdings database.

Holdings vs Synopsis I asked Arredondo how CARA’s  holding summaries are different West editorial “synopsis” which are added to cases in their reported cases. Arredondo  pointed out that the West summaries are written by an editor and Cara’s  holding statements are written by judges. I also asked  if CARA could flag cases where a judge had misinterpreted the holding of a case.  He responded that since CARA provides  a researcher with multiple summaries from different judges – presumably any misinterpreted holding would  be materially different from the others. It might be worth considering whether CARA could use their linguistic analysis to flag and highlight the “outlier holdings” which might be misinterpretations of a case holding.
One fundamental fact remains  — lawyers must still read the underlying cases whether they are utilizing  CARA’s  new features or a more traditional citator tool. It is exciting to see legal research innovators enrich lower cost research systems with sophisticated editorial-type  enhancements built on linguistic analysis, machine learning and algorithms.  CARA’s Black Letter Law and Holdings features are designed to help lawyers quickly strengthen their briefs with the best precedents and the strongest principles of law…. and then have time to read the cases themselves.