Analytical ideas

[lmullen]

Modern Law Mockup 01 copy.pdf

[kfunk074]

Ideas for more advanced stages (from Columbia junior faculty workshop 8/17/22):

• IP historians argue that patent holders created the field of "intellectual property" in the 1860s to associate patents more closely with copyright, which was regarded as a more sacrosanct property interest shielded from Congressional divestment. Opportunity to study pattens of field cohesion (opposite of differentiation).
• Does citation of a case in a treatise lead to greater citation of the case in the case law?
• Johnson v. M'Intosh is today a canonical case in Property but began as a Constitutional/Sovereignty case in the Marshall Court. Can we show its canonical migration? (Same with Marbury v. Madison, a case about prerogative writs that becomes a case about judicial review at the turn of the twentieth century.)
• Wheeler, Practical Law of Slavery, is an early attempt at a treatise aiming to create a field that never took off. What canons did the slavery cases ultimately end up in, if any? What other aborted attempts at genre/field formation can we detect from similar patterns? See also Justin Symard's Citing Slavery project.
• The Zurich economist Elliott Ash has a lot of work on text analysis that may provide helpful models/packages. Especially on educational network effects?
• Supreme Court practitioners today are launching a cottage industry mining old treatises for authority. There will be an interested audience in this work and especially in any kind of interface that might help them with their research.

[lmullen]

In general we have two categories of ideas here. (1) Start with something we know to be true, figure out how to find that in the data, then apply that process to other cases to learn interesting things. (2) Look for general patterns. And while usually I prefer the first option, the idea of looking whether a case cited in a treatise increases citations in cases seems like a great starting place. There is an easy way to visualize that, but probably also a more rigorous way to define likelihood. But it's also fundamental to the idea of the project, about treatises' role in lawmaking.

[kfunk074]

Another idea from a colleague: What kind of a citational afterlife do overruled cases have?

This is a more complex question than it might at first appear, because 1) cases may be overruled in part or by degrees--it's rarely so stark as say Dobbs overruling Roe, and 2) a whole industry arose in the late nineteenth century that attempted to index cases and signal whether they were still "good" law, so the very concept of an overruled authority shifted during our period.

But as I think of instances of overruled cases, I can add them to a list and we can see if any interesting patterns develop. Off the top of my head: Respublica v. Oswald, 1 U.S. 319, and Respublica v. Passmore, 3 Yeates 438, both held that defendants could be held in contempt for defaming the courts in publications. The Pennsylvania legislature overruled the cases in 1819 (statutory cite: 5 Sm. I. 55). Thenceforth, contempts could only punish in-court activity or the direct disobedience to a writ. Published libels against the courts had to be prosecuted as an ordinary crime.

[kfunk074]

If we want to mess around with a small but highly interesting (to me) corpus, here's a list of the books in MOML used to teach civil procedure in the law schools from 1870 to 1920. Procedure Education.xlsx

[kfunk074]

First steps in analysis useful for my book.

Analysis Ideas for Laws Machinery.docx

[kfunk074]

Just sketching an idea to visualize dueling canons. I have a table of procedure textbooks used by half a dozen law schools over fifty years. All the books are in MOML, and all cite cases as authority. So each school could be represented as a cloud of nodes, each node representing a case in the procedure curriculum. Common nodes shared by multiple curricula could be clustered together/overlap, while uncommon nodes stay to a periphery. My skimming leads me to suspect four or five of the schools will cluster together and possibly grow denser over time, while Harvard and maybe one other will break off and go their own way.

[kfunk074]

Figure out if self-aggrandizing judges attracted citations. Refer to Wright's preface: https://cite.case.law/wright/1/

And code:

SELECT j.name_long, c.decision_year, c.reporter, count(*) as n FROM cap.cases c LEFT JOIN cap.jurisdictions j ON c.jurisdiction = j.id GROUP BY j.name_long, c.decision_year, c.reporter HAVING j.name_long = 'Ohio' AND c.decision_year < 1900 ORDER BY c.reporter, c.decision_year

Notebook: https://observablehq.com/d/cfee63563c3ebd75