Reforming Copyright Law
Last time, I questioned the premises and theses of the common notion that the “Blurred Lines,” “Stairway to Heaven” and “Dark Horse” cases are chilling creativity. My main point was that the only way to determine whether this was happening is to assess whether quality of popular music has been declining as a result. No one has tried to make that case, and it would be difficult to disentangle the effects of these cases from other economic and cultural factors.
But the critics of these cases aren’t wrong, either. Copyright law is something of a contradiction. It promotes creativity by preventing creativity. More precisely, it encourages creativity by giving creators the right to prevent others from reproducing, adapting, distributing (copies of), publicly performing or publicly displaying their work. When the infringement takes the form of rote (or “literal”) copying, there’s no problem.
But copyright protects more than just the literal expression—if it didn’t, I could just re-write the story from your novel using my own words, or lift a catchy tune from your song. And these non-literal uses are fertile ground for creativity. Creativity depends on the works of others, but copyright law requires a certain creative distance—the actual border is fuzzy, sorry—from the preceding works. Thus, a great deal depends on how one balances the rights of preceding work’s creator, who was initially incentivized to create in the first instance, against those who would like to build upon, reference, criticize or celebrate the preceding work.
For the purposes of this blog post, we’re going to put aside fair use, even though that’s already going through your mind. Fair use is a defense (though a special defense) and thus like a backstop. Also, fair use does enough heavy lifting. Instead, I want to focus on the affirmative parts of copyright law, i.e., the elements that the copyright holder must prove (though, as we’ll see, the burden of proof can get flipped).
“SUMMONS FOR JURY SERVICE” by elycefeliz is licensed under CC BY-NC-ND 2.0
Rule No. 1 of Copyright Reform
Any reform of copyright law must be equally applicable to all the subject matter covered by copyright law. This means that any reform of copyright inspired what what we’re seeing in popular music cases must also work with novels, essays, certain kinds of databases, sculpture, product designs, fashion, paintings, “design,” architecture, dance1And pantomime![/ref], computer programming, photography, compilations, movies, brochures and all kinds of other fields of creativity that I just can’t think of right now.
There are specific parts of the Copyright Act that are limited to very circumscribed fields—basically §§ 108–120—but they are usually very specific, often so specific as to be useless, or are codifications of pre-existing rights (like the first-sale doctrine). The industries affected by your proposed reforms are much better at lobbying than you are (though they aren’t necessarily better at drafting statutory language!).
Rule No. 2
You can’t reform copyright just so cases like “Blurred Lines” and “Dark Horse” always come out the way you think they should. Any reforms must work with the system of dispute resolution we’ve been using for centuries. This means working with the rules of evidence, juries (shudder), and the possibility that things sometimes just go completely wrong.
The Jury Is ALWAYS In
The single most common suggestion for reform is to get rid of juries in copyright cases (or at least music cases?). Caramanica explicitly makes this point, suggesting replacing a jury trial with some sort of arbitration. Wang quotes several musicologists and industry folks who clearly blame juries, including some manager who believes that the English just ask a musicologist (which can’t possibly be correct).2Strangely, this quote comes right after Wang notes that a jury ruled in FAVOR of Led Zeppelin. Even the musicologists’ brief seeks to overturn a jury verdict. The theme in all these complaints is, to put it bluntly, that jurors are bad at music (because they reach different determinations than what seems obvious to those in the industry who have a stake in the outcome).
This is a criticism leveled at the civil jury system in a number of fields: jurors just can’t possibly understand your field and/or the law is too complicated for jurors to understand. My own feeling is that it’s premature to criticize the civil jury system because the jury instructions (and the way they’re given) are terrible, so we don’t know how a well-instructed jury would do.
Regardless of how one feels about the civil jury system, getting rid of juries is the reform least likely to succeed. That’s because civil juries are enshrined in the Seventh Amendment of the U.S. Constitution. That means either the Supreme Court would need to re-interpret the Seventh Amendment (as it applies to civil trials), or the Constitution would need to be amended.
The Seventh Amendment does not give litigants the right to try all issues by a jury. It’s limited to “suits at common law,” and whether a modern cause of action is governed by the Seventh Amendment depends on how the courts back in the day3I.e., when the Constitution and Bill of Rights were drafted and ratified. treated analogous causes of action. Correctly or incorrectly, the Supreme Court is very confident that copyright actions were tried by juries back in the day (even though, apparently the English have since stopped the practice).
So, hate on juries all you want, but you’re probably just wasting your time. A more fruitful avenue of reform like go like this:
- Write jury instructions so that jurors can understand them, which means coming up with a new standard to determine their “correctness.”
- Let jurors familiarize themselves with the relevant instructions before evidence. (Radical, I know.)
- If nothing else, don’t read the instructions to the jury at the end of trial, when they’re dead tired. Wait until they’re fresh. And take steps to make sure they understand.
- And at least let the jury take the instructions with them into deliberations. Even law students get to study the law before they take the exam!
- Stop putting off work on jury instructions until the very last minute. They’re not an afterthought.
Scènes à Faires: The Creator’s Toolkit
When you read the articles and the musicologists’ brief, what they’re all getting at is a legal concept called (unfortunately) scènes à faires, a term borrowed from theater that means roughly “scenes that must be made” or perhaps better “scenes that naturally flow.” But a much better way to look at it is this: it’s a toolkit.
If an element of a work is scènes à faire, it’s by definition unoriginal and thus not protectable by copyright. The practical reason for this is that it’d be impossible to create anything without a shared understanding of your field. All painters, novelists, photographers, sculptors, industrial designers, songwriters and computer programmers know certain techniques. They learn them at school, at conferences, from each other.
The theoretical reason for this is, if you learned the technique or whatever, it didn’t really come out of your own head. You’re getting it from somewhere else.
So, when the musicologists (especially) talk about how a certain rhythm isn’t original, this is what they mean.
The way scènes à faires should work is like this. You find an expert in the field and ask her about the practice in the field: is this technique, element, figure, etc. something that ordinary practitioners in the field use? This is actually a fairly common inquiry in law, though it’s usually called “industry practice.”
But it’s not this simple because scènes à faires is bound up with the broader issue of originality. Originality is the sine non qua of copyright. Without it, there is no copyright. It does NOT mean novelty. You don’t have to be the first to ever have come up with it. It just has to have come out of your own (human) mind.
But courts look at originality kind of backward. First of all, they look at the copyrighted work as a whole, asking whether the work as a whole is original. Furthermore, the obstacle posed by originality is very low. The way a parts catalogue is organized can be original, provided it’s neither arbitrary (i.e., random, thus displaying no creativity) nor regular (i.e., the product of a system and thus an unprotectable idea).
You can immediately see the problem. If you say “Joyful Noise”4The song “Dark Horse” supposedly infringes on. is unoriginal because it uses scènes à faires, the court will look at you funny and say that it obviously is original. So what you need to say is: the parts of “Joyful Noise” alleged to have been infringed by “Dark Horse” aren’t original (because they are scènes à faires). The law doesn’t really provide an easy way to say that. And the pattern jury instructions don’t even come close to describing it like that to the jury.
This problem is exacerbated by copyright law’s problem with “filtering.” Copyright law instructs us that we are never to filter out elements when comparing the copyrighted work with the accused work because we are always to compare works as a whole. But it also teaches us that copyright protects only those elements that are original, which implies some kind of filtering. These two seeming contradictions can be reconciled (using something akin to the abstract-filter-compare method usually reserved for software), but doing so is a difficult mental exercise. More commonly, judges and juries end up comparing the two works as a whole, but focusing on the elements asserted by the plaintiff, which are frequently exactly the sort of elements that should be filtered out.
There is one final problem. Originality is naturally something the copyright owner must prove. BUT one of the benefits of timely registration is that the registered work is presumed to be originality. This means the burden shifts to the defendant to prove lack of originality. Of course, this presumption goes to the work as a whole, so when the U.S. Copyright Office grants a registration, it’s only saying that the work as a whole is original, not that every element is original.
If you want to reform the way scènes à faire operates, therefore, here are my suggestions:
- Whenever a copyright holder asserts non-literal infringement, the burden remains on the copyright holder to prove originality as to the asserted elements.
- Re-define scènes à faire as figures, techniques and other elements that are common to a field of creative endeavor.
The first suggested reform will have two major effects. First, it will force plaintiffs who bring cases for non-literal infringement to identify those elements that, in isolation or in combination, are infringed by the defendants; AND to allege how the accused work infringes those elements as so identified; AND that those elements (as so identified) are original. This should dissuade lawsuits whose attitude is: everything is presumed original (thanks to the timely registration) and we’ll work out the details later with the jury.
Second, once the defendants can bring forth evidence that at least some of those elements are scènes à faire, the plaintiff will be vulnerable to a motion for summary judgment if it can’t bring forth substantial contrary evidence (enough to create a genuine question of fact) that they aren’t.
The second suggested reform re-casts scènes à faire in intuitive terms that judges and juries can grasp. Further, it should make proof easier. Instead of overwhelming a jury with “prior art” (a term unfortunately borrowed from patent law), experienced songwriters can testify that the elements are in their fellow songwriters’ toolbox.
Substantial Similarity Is in Crisis
Substantial similarity sets the boundaries between what otherwise protectable content you can and can’t appropriate from a copyrighted work. You are permitted to deliberately copy elements of another song, provided you stop short of taking “too much.”
Thus, substantial similarity is important in determining how and how much you may be influenced by others’ previous works. By design it’s fuzzy. Even the great Learned Hand had to throw up his hands when trying to describe it: “The test for infringement of a copyright is of necessity vague. … Decisions must therefore inevitably be ad hoc.” Importantly, we leave the question up to the jury (something that deeply offends the amici musicologists) because the issue comes down to whether an ordinary observer would regard the aesthetic appeal of the two works the same.
Although substantial similarity will never be clear, the way it is applied by most courts is terrible. The way it is applied in the Ninth Circuit, where these three cases were brought, is even worse. I explained in detail in Techdirt what is wrong with substantial similarity as commonly understood and applies, and how it should be understood and applied. I’ll refer you to that article, but here’s the shortest summary I can come up with: There are two different kinds of similarity relevant to the inquiry into copyright infringement that play very different roles, and it’s very, very important to keep them separate.
I’ll add this much: if the jury had been properly and clearly instructed about substantial similarity, “Blurred Lines” would probably have come out differently, and “Dark Horse” would definitely have come out differently.
The main reform here: adopt the Second Circuit’s Laureyssens v. Idea Group (including, you, Second Circuit courts, which often forget their own controlling authority).
One final point: Remember how I said any reform must work with all areas of copyright’s subject matter, not just popular music? Well, that can work in reverse. It’s possible that popular music is unusually restricted creatively. This isn’t to say that writing popular songs isn’t hard—in some ways, these restrictions would make it harder!—but that it’s very difficult to avoid using lots and lots of scènes à faires and other unprotectable elements. But juries assume a normal range of creative choice and thus tend to overlook just how much of popular music borrows from other popular music, and that such borrowing is OK. This would also mean that popular songs’ copyright is “thin,” meaning that the degree of similarity to constitute infringement must be much higher—perhaps even near-identical—than for other types of works.
I’m not sure I buy that, though I suspect there’s something to that. I do resist the idea that melody is the only aspect that can be protected.
It doesn’t appear I’ll get to all the potential reforms in this post. I’ll finish up in the next post. Thanks for reading!
