Skip to content

My failed “zombie” cognitive research, Part 2: character does not equal causation

December 23, 2014

So this is part two of presenting some old research ideas. The previous post was about some really silly cognitive studies I tried. They didn’t work out so well.

This post is about another failed study. I really love this one. I still think it’s super fascinating and deserves more attention. Like the work described yesterday, this work did not particularly stir interest amongst my so-called advisor. Particularly, it was pushed aside because it was not experimental.

Well of course it’s not experimental, how could I experiment on the Federal Courts? Even the jurisdictional experiments of our 50 state democracy are not really experiments, they are iterations. They can’t test their legal theories in controlled setting, every repetition is a new case in a new court. Judicial opinions are always unique decisions in time.

Lexis gives incredible access to the public record. I don’t understand why the public can’t provide equal access to it’s own record. Pay-walls are a plague on the intellectual community. Still, as a student with academic access, Lexis was great. I would frequently spend time digging around and randomly search for funny words. This led me to create my book series “Law of the Horse” compiling some of these funny words from the courts’ opinion record.

I started noticing some unusual associations. I’ve discussed these association regarding zombies before. Early ZombieLaw posts frequently noted the abundance of Social Security references, and particularly regarding medications. Also multiple Statute of Limitations puns, corporate Ponzi-scheme banking, and other crimes.

So for another research attempt I tried to show that there was some kind of conceptual pattern in the association of this character word to those correlated legal topics. I figured that if it was happening for zombie maybe other character words also had conceptual character patterns. And I think I found some.

My theory is that characters adhere to contexts which associate to themes and meanings in the world. I hoped this was evidence for the transmission of cultural ideas and conceptual blends. My hypothesis was that character words associate to legal subject matter words in federal court opinion in ways that differ by character and suggest a portrait of attributes relevant to the character.

The character words chosen for this research were inspired by Mark Rayner’s game variation on rock-paper-scissor: “Monkey Ninja Pirate Robot Zombie”

And also by Tim Buckley’s artwork of a robot-pirate fighting a zombie-ninja-on-a-dinosaur, entitled: “This is the way the world ends”

My methodology was very simple. I searched the text of federal court opinions using Lexis Academic search. I counted how many court opinions have the character word and then how many of those court opinions with that word also had another word. This method is admittedly unreliable for some reasons discussed below but it demonstrates an exploratory method to use the corpus of federal court opinions as a tool for cultural study. A fuller semantic analysis is warranted. Perhaps one day all of these cases can be put into a Coh-metrix by someone more organized than me. That would help account for word correlations that were not counted here.

The sample size (number of Federal Court opinions) varies for each character word. “Zombie” had about 360 cases at the time of this research; “Ninja” about 120, “Robot” almost 1000, “Dinosaur” about 500ish, “Monkeys” about 2000ish, and “Pirates” about 2500. There were very few cases that included more than one of these character words in the same opinion.

In the table below I report some of the most interesting associations that show differences between the characters. These are straight correlations of the word counts from Lexis. Either a case has the two words or it doesn’t. The graph reports the proportions of those cases within the character word set that also have the other target word. It reveals something like a unique fingerprint for each of these characters.

In the first graph notice that each of the searched characters correlated with some legal topic words. From these graph we can see that at the time of this research the characters had somewhat different patterns of association.

“Zombie” shows strong proportion of social security, criminal, contracts and constitutional. The word “ninja” is predominately in cases of criminal and constitutional law. Recall my blog of NinjaLaw and consider that “constitutional” cases often correlate with the criminal claims because many of the “ninja” cases are habeas petitions filed by convicted criminals alleging violations of their constitutional rights at trial, or rights violations in prisons.

Notice also that “pirate” and “robots” and “dinosaurs” have sizeable proportions of copyright related cases but probably for different reasons. The dinosaurs are more the intellectual property being infringed, whereas the pirates are that too but also the persons doing the infringing. Those kinds of narrative details are not available in this kind of raw correlation research.

Interestingly, “monkey” has a noticeable proportion of Equal Employment Opportunity Commission cases (EEOC), (“robot” and “dinosaur” too) because these words have applications as slang about worker performance. They are quotes from testimony used in much the same way that “zombie” gets quoted in the social security opinions. Still, the judges (or their clerks) had to decide which quotes to include in the opinion, and so it is still noteworthy to try to track these associations.

In the second graph, the character words are analyzed for correlations with some non-legal words. Otherwise it is the same method of raw counting word association data. Here we see “zombie” related to “depression”, “death” and also some “technology”. Compare “ninja” and “pirate” which both correlate to “technology” and “death”, but “ninja” is more “death” than “technology” and “pirates” are more “technology” than “death”. “Dinosaur”, “monkey” and “robot” are all somewhat correlated to “death” and “technology” in the federal courts, so it’s not really clear what any of this means. Still I found it fascinating to learn that only the “robot” has noticeable connection to “welding”.

“Correlation does not equal causation.”

It’s possible that some of these patterns would wash away as the sample size increases. Alternatively, once a case is decided, the rules of stare decisis and of citing precedent make it more likely that the words will be quoted again for the same topic. This research does not account for those repetitions of case precedent quotations. Nor does it account for any other connections to the ideas that don’t use the exact word. This is a pure computer word correlation. Additionally the word correlations are not mutually exclusive categories, though, as mentioned above, the character words themselves have very little overlap.

Since most of the readers of this blog are primarily interested in the “zombie”, here are some more of the associations for “zombie” in the federal courts opinions. This graph shows words that appear in a sizeable proportion of the “zombie” cases but not so much in the other character word sets. By way of contrast, consider some words that appear proportionately more frequently in the “ninja” cases.

The purpose of this research was to show that these character words correlate to other terms and that the record of federal court opinions can demonstrate some intriguing pattens of the cultural character. This is not experimental research and the variables are not controlled in any way. It only shows by raw correlation of words in federal court opinions that there may be some pattern amongst these character words that adhere through language and into the law. This suggests the possibility that these character memes exist in an idea space of metaphor and language in which they are attached by other semantic features.

Unfortunately, none of my advisors really had any idea what to do with this data, and again I didn’t find any cognitive effects or implication that the words were impacting the outcomes of these judicial decisions. So, I moved on. In my next post, I will describe the zombie research that ultimately earned my Masters, but honestly I still think the research reported in this post was a more important idea. I might like to return to this idea in the future and if anyone has any suggestions please let me know.


From → Uncategorized

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: