The Web of Law:
A post by Dave Hoffman at PrawfsBlawg reminds me of Tom Smith's very cool draft article on SSRN that I have been meaning to mention: The Web of Law.

  Smith's article looks at legal citations from the standpoint of network theory, and presents the results of a citation study performed on Tom's behalf by the people at Lexis Nexis that looked at the citation structure of nearly 4 million American legal precedents. According to Smith, the study reveals that the citation of legal precedents creates a scale-free network, and that allows for all sorts of modeling from network theory that can shed light on legal dynamics. From the introduction:
  What determines whether a case makes it into the elite of cases that are cited hundreds or thousands of times, instead of just a few, or never? By studying the statistical dynamics of citation over time, scholars using network theory could shed significant light on what accounts for the success of a legal authority. How much, for example, does a case's merely being decided earlier account for its citation frequency? How much of a difference does a case having been decided by a higher court make? What about the "fitness" of a case, in terms of its persuasiveness or analytical acuity? Do these attributes explain in part a case's flourishing, survival, or extinction as a precedent? Do cases have a natural life span? Does their authority tend to wax and wane, and does this depend on the type of case? It is possible that if we study the evolution of the legal network, the "Web of Law," we will discover unsuspected historical dynamics in legal authority. Perhaps the characteristics of legal evolution themselves have changed over time. If there have been changes in the dynamics of legal evolution, or other noticeable changes, perhaps they correspond to recognized watersheds in legal history. As I discuss briefly below, using network theory to analyze law may enable us to understand, in a much more rigorous way than previously possible, the dynamics by which interpretations of important laws, such as the Constitution and landmark statutes, change over time.
  Very interesting stuff, and also both well-written and relatively short (with lots of cool graphs, too). I'll enable comments in case readers want to discuss the argument.
Texas Brian (mail):
I have not read the paper. Though the empirical work was probably nontrivial, this approach strikes me as both obvious and, to a certain extent, trivial. It is trivial because the network model is similar to former English major law profs using cross-textual analyses. Since I am acquainted with little more than the terms, I cannot say much. Cross-textualism appears to exploit and examine the linkages that the network approach models. Combine the two and you get what is probably an interesting inter-disciplinary problem.
6.7.2005 2:46pm
Anthony D'Amato (www):
"The Web of Law" is marvelous. It's just the kind of bridge-building paper that we need a lot more of in today's fragmented legal scholarship. I've been looking at network theory in connection with a project on the substantive links among rules of international law, and I have learned much from Professor Smith's paper.

His approach is heuristic (which is all it can be at the very first stage of interdisciplinary work). He looks at network theory and spins off a number of hypotheses about the law, all of which are very interesting. As a next stage, he might consider starting with a legal problem and seeing what light if any network theory can throw upon the problem.

One such problem occurred to me as I was reading his paper. Back in the Middle Ages, when I was in law school, we had a class in Legal Process given by one of the co-authors of that now-famous set of legal materials, Professor Albert Sacks. I recall, not too sharply, that one of the modules in the legal process materials (which unfortunately have never been published in book form but which are probably available in most law libraries) had two cases, A and B, involving roughly the same set of facts but with different outcomes. When we looked at the precedents relied upon by the judges in the two cases, we found that none of the precedents cited by A were cited by B. Thus there were two lines of cases, in Professor Sacks' words, that never communicated with each other.

It would be interesting to network these two lines of cases. At some degree of separation, the precedents should link up. What is the first instance when they link up? What happens after that? How far back in time do we have to go to find connections? Are there some hub cases in the deep past that spawned the two different lines?

Here's a far-out speculation: Suppose the lines of cases resemble a double helix? Does that mean that when they connect up at the end (as they must after the "secret" of the Hart-Sacks materials is out of the bag and a new case comes up where both lines of cases for the first time are cited), a new gene will be born?

And here's a hypothesis which I've often wondered about but never knew how to test. Perhaps network theory can be harnessed here. My hypothesis is that the non-connectiong lines of cases are becoming increasingly frequent! My intuition is that back in the medieval period, judges and lawyers were familiar with many of the precedents--there were, after all, fewer of them back then. The main reason for their greater familiarity was the absence of word-based computerized retrieval. There was, of course, Westlaw's "Key Word," but it was regarded as full of errors because it all had to be entered by hand. But with the Googling and Lexing of today, nearly all precedents are "read" when you retrieve them by a word-based search. This means, I think, that we are all getting to know a lot less about case law.

Citation-superficiality may indeed make the power law curves straighter, but is this a jurisprudential improvement?

And that leads to a final speculation: is the straightness of the Supreme Court's power law as revealed by Professor Smith a sign that the Supreme Court pays hardly any attention at all to its own precedents? The less attention they pay, the more their clerks will be instructed to "throw in" a lot of citations to precious cases to dress up the justices' opinions.
6.8.2005 12:49am
SupremacyClaus (mail):
Yes, that article was worth reading.

It failed to address how many justices were paid off to account for the wackiness of most influential decisions. How does one get out of a contract to buy a barren cow, Rose of Aberlone, after finding out it is fecund, without paying off the judges? That is the purpose of a contract, to enforce a promise one does not want to keep. Only a payoff can explain.

It failed to account for the Grand Unifying Theory of Appellate Decisions, rent seeking on behalf of a lawyer criminal cult enterprise. That is many orders of magnitude bigger and more influential than bribery, or as the man calls it, campaign contributions. That is similar in pervasiveness and importance as the background radiation from the Big Bang. It the universal, ubiquitous, cultural training and indoctrination of the lawyer. It is interior in any who wants to be admitted to the Bar. Any legal theory must address it.

It failed to return back far enough, to address the paper trail from Scholasticism, and its thuggish, crime against humanity enforcement, the Inquisition, to modern legal doctrines. The 1600 A.D. methods are smoother, but unchanged ideologically or physically. The robe, the bench, the gavel, the court architecture, the buffoonish pomposity, and self-seriousness of its dumbass occupiers, all from Church, and illegal in this secular nation.
6.8.2005 9:58am
Stephen M (Ethesis) (mail) (www):
It failed to address how many justices were paid off to account for the wackiness of most influential decisions. How does one get out of a contract to buy a barren cow, Rose of Aberlone, after finding out it is fecund, without paying off the judges? That is the purpose of a contract, to enforce a promise one does not want to keep. Only a payoff can explain.

Not hardly. Who could afford such a pay-off over a cow.

A much better approach is to look at the philosophy of language changes that were processing at the time -- in one case (the gemstone case) identity was denotation. In the other case, identity included connotation.

If you read the dissent, it is obvious that the real disagreement was over identity including connotation and what the proper connotations were to apply to the cow (and what connotations were applied by the party).

It is a reflection of the weakness of cross-disciplinary work that while this seems immediately obvious to anyone with background in that area, the analysis is still not in print. Too bad there aren't any trained philosophers doing work in law ;)

What is interesting is the way other theories intertwine and are cited -- often against a parallel effort by their authors to control their meaning -- the doctrine of "good faith and fair dealing" vs. the tort of "bad faith" is a good example.

I know, I've been sidetracked into discussing the examples rather than the principle ...
6.8.2005 11:38am
SupremacyClaus (mail):
Stephen: You could be correct. A philosophy of language, Wittgensteinian, ontological divergence could have taken place in the rural courts of Michigan of the 1880's.

The gemstone case is straight theft of a little girl's property, enforced by cold law. Wacky, but especially inscrutable, in following the cow case. I still do not understand the justification for the difference between mutual mistake, and unilateral mistake. Who cares? A contract is a contract, but somehow, the poorer party gets screwed every time, and a contract is not a contract in the face of defenses. This is suspicious to me, since only a certain type of person knows the defenses.

What was the wage of a Justice in Michigan in the 80's? How do we know a dollar, or a 2 bit bottle of whiskey would not be sufficient to permanently imprint contract law for 120 years? The legal fees spent should be posted on the LexisNexis appellate cases, or in every appellate case from here forward, includint the campaign contributions by the law firm, and straight bribery budget. The rich are likely to have ruled, quite easily, with the richest litigant being government. Decisions made in favor of a campaign donor should be reversible per se, better yet, subject to the Exclusionary Rule II, applying to judges. See how they like it for a change. If there are 3 such exclusions, the impeachment should also be per se.

A money trail explanation is elegant and has physical, measurable reality. A network explanation, or ontologic speculation is too hard to understand, harder than the cases themselves. These should follow the disproving of human corruption theory, not precede it.
6.8.2005 1:04pm
Stephen M (Ethesis) (mail) (www):
A philosophy of language, Wittgensteinian, ontological divergence could have taken place in the rural courts of Michigan of the 1880's ... it did take place all over the American English speaking world about then -- a shift from a thing being the item denoted (so there is no mistake that the rock was the rock regardless of what it was made out of) to the thing including its connotations (so that the cow was not the cow if it wasn't barren).

I found the dissent moving in that case, and the legal doctrines interesting.

Unfortunately, it appears that these posts have diverted the thread to the point no one else has posted -- an amazing thing on this blog.
6.8.2005 6:57pm