Stairway to Heaven Is Not Blurred Lines

I Know They Look the Same, But the “Stairway to Heaven” Court Gets Something Very Important Right

We interrupt the regularly-scheduled blog post about copyright in public murals in Nashville with this counter-intuitive defense of the Ninth Circuit’s recent surprising reversal of the jury verdict in the “Taurus”-“Stairway to Heaven” case. I know, I know. I looks like “Blurred Lines” all over again, only in reverse. Whereas in “Blurred Lines,” the jury reached the “wrong” conclusion, and the Ninth Circuit refused to fix the jury’s mistake, here it looks like the jury reached the “right” conclusion,” and the Ninth Circuit is screwing up the jury’s work. Techdirt all but says so, in an article 9th Cir Never Misses a Chance to Mess Up Copyright Law: Reopens Led Zeppelin ‘Stairway to Heaven’ Case.¹The Hollywood Reporter has a more balanced approach to the decision.

This might be the day the Ninth Circuit stops repeating the same nonsensical test for substantial similarity. Have they truly learned their lesson?

I’m pleased to report that, far from taking this opportunity to further screw up copyright law, as Techdirt fears, the panel of judges is attempting to improve copyright law by replacing the Ninth Circuit’s (very bad) framework for copyright infringement with a much better one. Indeed, the “Stairway to Heaven” opinion may be seen as a rebuke to the “Blurred Lines” opinion. The pity is that Led Zeppelin must bear this burden by having to do the trial all over again.

The reason the “Stairway to Heaven” has to do with our old, misunderstood frenemy, the “Inverse-Ratio rule,” which is only tangentially related to why the case is being sent back for a second trial. I blogged extensively about the “inverse-ratio” rule in connection with the “Blurred Lines” case, and how the Ninth Circuit (correctly) amended its opinion to excise (its terrible) discussion of the rule. But here’s the deal: the inverse-ratio rule provides that the more similarity there is between the two works, the less proof access you need, and vice-versa. It is highly beneficial when applied to the right legal framework for copyright frame work. It is a perverse disaster when applied to the wrong legal framework.

Two Legal Frameworks for Copyright Infringement

Generally, there are two leading legal frameworks for determining when someone has infringed copyright:

1. Plaintiff must prove (a) defendant’s access to the copyrighted work, plus (b) “substantial similarity” between the two works. This framework is used by the majority of courts, so we’ll call it the “Majority Framework.” It is also a bad framework.
2. Plaintiff must prove (a) “copying” and (b) “unlawful appropriation.” To prove “copying,” the plaintiff must prove (1) access and (2) “probative similarity.” “Unlawful appropriation” is pretty much the same thing as “substantial similarity,” but the different term is used to avoid confusion with “probative similarity” (and is also more accurate). This framework is preferred by scholars and is used by one half of the Second Circuit, so we’ll call this the “Preferred Framework.”

Both of these frameworks grapple with the same thing: to infringe copyright in a work, the defendant must both copy the work and take enough of the work that, well, um, it’s too much. But the key is that copying is something that must be proved. In theory, if a magical monkey banged out Harry Potter and the Chamber of Secrets without ever have looked at a copy of Harry Potter and the Chamber of Secrets, that would not be copyright infringement. It would just be unbelievably unlikely.²More realistically, this is principle behind “clean rooms,” when computer code is (legally) copied by having one team examines the software and breaks the copyrighted software down into functional components, then another team takes those components and writes its own code, without ever having access to the original software. But even then, any coincidence between the new and old code can be explained by having similarly trained coders seeking to achieve the same functional ends, which probably means that those components were never protectable in the first place. The need to prove copying is actually a fairly important limitation on copyright law. We shouldn’t lose sight of it.

Imagine that you saw me with a book and that I was writing something down on a pad of paper while obviously reading the book. Have you proved that I have “copied” the book? You certainly have proved access: there I am, with the book in hand! But, you haven’t. You need one piece of evidence. You need to see what I am writing. It’s possible that I’m multitasking, perhaps writing a shopping list as I read. It would take just a glance to figure that out, though, right?

Now imagine that you didn’t see me with the book. No one did. And yet my notebook is word for word the same as the book. Do you need any evidence of access? Maybe just the slightest bit, like I live in the same city as a bookstore or library that has a copy of the book. (This is sometimes known as “striking similarity.”)

Does this make intuitive sense? If so, congratulations! You understand the “inverse-ratio” rule, at least when correctly applied to “copying.” The more access, less probative similarity is needed to prove copying. More probative similarity, less access needed.

Proving copying is only step one of a two-step process, so let me hit you with another hypothetical. Let’s say that I was seen with the book at some point, and my notebook contains an extensive data table that is exactly the same as a data table in the book (but in my handwriting). There is nothing else in my notebook related to the book. The data table is otherwise unremarkable. Have you proven infringement? No! You have proven copying. I had access to the book, and the data table is too much of a coincidence to explain away. Clearly, I copied the data table. But not all copying is infringement. What I took isn’t protectable because it’s just facts. You haven’t proven the “unlawful appropriation” (AKA “substantial similarity”) element.

This hypothetical shows the difference between the two kinds of similarity: probative and substantial. (See, this is why we should re-name “substantial similarity” as “unlawful appropriation.”) You can use unprotectable elements to establish probative similarity. You can’t use unprotectable elements to establish substantial similarity/unlawful appropriation. Hopefully, this makes intuitive sense to you. With probative similarity, we’re just trying to prove copying, so we’re just looking for coincidences that, in light of the amount of access, we can’t explain away. Substantial similarity/unlawful appropriation is limited to protectable expression.

Let’s return to the “inverse-ratio” rule. Can you see why it’s appropriate in the context of copying, when we’re weighing access against probative similarity? And why it’s inappropriate if you are instead weighing access against substantial similarity/unlawful appropriation? In fact, it’s perverse. Imagine if you and I both took the same material from an unpublished manuscript, and that the case for substantial similarity is kind of borderline. I am found with a copy of the manuscript on my computer. You, however, have no connection with the manuscript except that you are the manuscript editor’s niece’s best friend (and the niece visited the editor at least once while in possession of the manuscript). We both took the same thing, but under the misapplication of the inverse-ratio rule, I’m an infringer, and you are not. That’s stupid. Considering that most courts do not correctly handle the distinction between copying and substantial similarity, you can see how the inverse-ratio rule has gotten a bad rap.

The Ninth Circuit Uses a Weird Hybrid Legal Framework for Copyright Infringement, but “Stairway to Heaven” Opinion Might Fix That.

Which of these two frameworks does the Ninth Circuit use? Neither, of course. This is the Ninth Circuit we’re talking about. Its framework looks like a combination of two. On the one hand, it follows the Majority Framework by requiring proof of access (instead of copying) and “substantial similarity.” But then is splits the inquiry into “substantial similarity” into two tests, both of which must be proven: the “extrinsic test” and the “intrinsic test.” The extrinsic test is meant to be “objective.” The jury is permitted to consider both protectable and non-protectable elements, to dissect the elements and compare those, and to consider expert opinion. You know what that is? Probative similarity, but here’s it’s pressed into the service of substantial similarity and has been separated from the inquiry into access.

The intrinsic test is just the normal test for “substantial similarity”/“unlawful appropriation” that every other court uses. Calling it a “test” is overstating it: it’s more like the jury’s gut instinct, based on the “total concept and feel” of the works. Sorry, but no one has come up with a better way of formulating the test, or how “total concept and feel” works when both protectable and non-protectable elements have been taken.³To be fair to the Ninth Circuit, the distinction between intrinsic and extrinsic evidence was innovative at the time and served as an important stepping stone to the preferred legal framework. Alas, the way precedent works, the Ninth Circuit has been stuck at this half-way point for decades. It should go without saying that the application of the “inverse-ratio” rule to the Ninth Circuit’s framework can be ugly.

But that’s what makes the court’s opinion in the “Stairway to Heaven” case so exciting. The court straight up endorses the Preferred Framework, even using the preferred terminology:

Whether Defendants copied protected expression contains two separate and distinct components: “copying” and “unlawful appropriation.” A plaintiff must be able to demonstrate that a defendant copied his work, as independent creation is a complete defense to copyright infringement. In cases such as this one where there is no direct evidence of copying, the plaintiff can attempt to prove it circumstantially by showing that the defendant had access to the plaintiff’s work and that the two works share similarities probative of copying. When a high degree of access is shown, a lower amount of similarity is needed to prove copying. To prove copying, the similarities between the two works need not be extensive, and they need not involve protected elements of the plaintiff’s work. They just need to be similarities one would not expect to arise if the two works had been created independently.⁴I’ve omitted citations and most of the internal quotation marks.

It then tackles the plaintiffs’ argument about the missing “inverse-ratio” rule. I’ll bet the plaintiffs were wanting an “inverse-ratio” instruction that told the jury to weigh access against substantial similarity. If so, the Ninth Circuit had some bad news for it: the “inverse-ratio” applies only to copying, not to substantial similarity.

This [inverse ratio] rule assists only in proving copying, not in proving unlawful appropriation. Even if a plaintiff proves that a defendant copied his work, the plaintiff must still show that the copying amounts to unlawful appropriation. The showing of substantial similarity necessary to prove unlawful appropriation does not vary with the degree of access the plaintiff has shown.⁵Again, I’ve removed citations and internal quotation marks.

The court then went on to suggest strongly to the trial court that an “inverse ratio” rule might be a really good idea, in a case where there’s proof of access.

But what about that intrinsic-extrinsic distinction? Surely, it’s no longer relevant, right? Well, I guess the Ninth Circuit wasn’t quite prepared to get that radical. With a little hocus-pocus, it sort of re-purposed the distinction in the service of substantial similarity/unlawful appropriation. Now, instead of being objective and subjective, it’s about making sure you have both evidence of appropriation of protected elements and that gut-feeling. OK, whatever—that might actually be an improvement, since its one way to reconcile “total concept and feel” with a need to filter out non-protectable elements.

Emphasis on “Might”

The fight’s not over. When a Circuit court takes up an appeal, it assigns a panel of three of its judges. In theory, a panel can’t overrule an earlier panel—for that to happen, you need the entire court (called “en banc”) to take up the case, which is rare (but not unheard of).⁶The Ninth Circuit is so large that “the entire court” doesn’t mean all of the judges, just most of them. Right now, later panels have the option of picking which framework they want to use: follow “Blurred Lines’” traditional Ninth Circuit framework, or adopt “Stairway to Heaven’s” newfangled Preferred Framework. Obviously, parties are going to pick the one that works best for them. We might end up with an internally-split circuit (which is essentially the current situation with the Second Circuit). Of course, the Supreme Court to step in and fix it, but let’s be serious.⁷It is almost scandalous that, for decades, there have been multiple frameworks for a concept CENTRAL to copyright law, and the Supreme Court has never expressed interest in the subject.

Now, as it happens, the court’s discussion of the “inverse-ratio” rule, and its related description of the Preferred Framework, is not what caused the court to vacate the jury verdict and send the case back down for a second trial. I’ll explain the twin brain farts that caused that disaster in a subsequent post.
Thanks for reading!

Rick Sanders

Rick is currently General Counsel for Software Freedom Conservancy. Previously, he has been practicing law as an intellectual-property litigator since 2000.