The Main Event
Last night I attended a great event at Cardozo Law School titled Our Bizarre System For Proving Copyright Infringement, also the name of Professor Mark Lemley’s paper which the four distinguished panelists, including Lemley, discussed. Thanks to the school and panelists for hosting a very interesting and informative event. Each panelist provided commentary on the paper before a few audience questions, followed by a reception where I once again had the opportunity to engage in a spirited discussion of the merits of Creative Commons licensing (round one here). In other words, arguing with old ladies.
Professor Lemley began by providing an excellent summary of the existing standards and the slight variations by circuit in the test for proving copyright infringement. Two formulations of essentially the same procedure are followed to determine whether there is substantial similarity between two works, resulting in infringement. Arnstein v. Porter, a Second Circuit case, articulates the classic rule that a plaintiff typically must show ownership of a valid copyright and copying by the defendant. Because there is rarely direct evidence of copying (e.g. the defendant admitting it), circumstantial evidence is used to show copying by demonstrating that the defendant had access to plaintiff's work and that defendant's work is "substantially similar" to plaintiff's. The first step of the Arnstein analysis involves the court analyzing and dissecting the work to filter out the unprotectable elements. This is usually done with the aid of expert testimony. Next, improper appropriation of more than a de minimus amount of protectable expression must be shown. In this second step, the “ordinary, lay observer” standard is applied by the jury to determine whether the copying in question rose to the level of improper appropriation.
The First and Third Circuits follow Arnstein, while the Fourth, Sixth and Eleventh Circuits generally follow the test set forth in the Ninth Circuit’s Sid & Marty Krofft Television v. McDonald’s. Krofft is about the whether McDonald’s copied the plaintiff’s characters. Trip down memory lane and see for yourself (pun intended): Krofft’s H.R. Pufnstuff, McDonald’s Land commercial. (If you haven't seen the NSFW Mr. Show parody of H.R. Pufnstuff, it is hilarious.) In Krofft, the court called the factors “extrinsic” and “intrinsic”, which gets us to the same place. The extrinsic inquiry considers whether plaintiff’s ideas were copied, while the intrinsic test looks at whether the expression was copied. In decisions under the intrinsic test, "analytic dissection and expert testimony are not appropriate." At least for “books, scripts, plays, and motion pictures”, extrinsic factors eventually came to include filters for both ideas and expression. “The two tests [extrinsic and intrinsic] are more sensibly described as objective and subjective analyses of expression, having strayed from Krofft's division between expression and ideas.” Shaw v. Lindheim (emphasis added). As Lemley summarizes, regardless of which test is followed, the general approach “permits expert testimony in the first step – inferring copying – but excludes it when it comes to the second step of determining whether copying is unlawful.” (emphasis added)
The paper goes on to discuss “outliers on expert testimony” in Part I.C. For example, the Sixth Circuit, while similar to Krofft, applies “a more stringent standard” for allowing expert testimony in the first prong, allowing the trial court more discretion according to Federal Rule of Evidence 702. Things get more interesting when we start to look at outliers on the second prong. Lemley notes that “courts that follow Krofft are more willing to treat as the ‘ordinary observer’ the likely customer of the copyrighted products, which may or may not be a member of the jury.” He begins with the Fourth Circuit in Dawson v. Hinshaw Music, which allowed expert testimony on both prongs of the inquiry in a case involving classical music, instructing district courts “to consider the nature of the intended audience of the plaintiff’s work.” While this usually consists of the general public and thus the ordinary observer standard applies, the standard shifts when “the intended audience is more narrow in that it possesses specialized expertise, relevant to the purchasing decision, that lay people would lack…” The Sixth Circuit similarly allows the “intended audience” inquiry to influence the use of expert testimony. Lemley calls these cases the “extraordinary observer” category.
Although the Ninth Circuit, like the Second in Arnstein, requires the use of an ordinary observer standard for musical works, it embraced the standard of a “a discerning 17.5-year-old boy” in a case about infringement of a video game. Similarly, in software cases, how is an “ordinary observer” able to make distinctions between the abstract functions of computer programs and their actual implementations without the aid of an expert? The danger lies in the rules of evidence, which prohibit experts from testifying to the actual legal conclusion. In other words, an expert can’t say “Defendant's work clearly infringed,” or vice versa. Lemley’s paper argues that the current system produces bizarre results, and will continue to do so as we move further into the digital age where copying is often established, but jurors are not able to make the required distinctions about whether it was too much copying of protectable elements. Lemley argues that “the analysis of proof [is] exactly backwards – permitting analytic dissection of the works and expert testimony where the question is one that should be handed to the members of the jury, and falling back on the “ordinary observer” test on the very questions that require careful dissection by the court.” (emphasis added)
He uses two familiar cases as examples, Feist and Steinberg. In Feist, copying was established—the defendant copied fictitious phone book entries. Shouldn’t this have counted for something? In Steinberg, Columbia Pictures’ movie poster for Moscow on the Hudson clearly copied plainftiff’s New Yorker cover. Lemley notes that the court wastes too much time talking about the unprotectable elements while avoiding the difficult question of whether protectable expression was copied. Even in the “virtual identity” cases, Lemley notes, courts go through the same two-step procedure of first dissecting out the unprotectable elements, then having the jury determine whether there is “virtual identity” between the two works. The “virtual identity” test applies to works that are protected by a “thin” copyright because they consist primarily of unprotectable elements. In other words, there are only so many ways to make a glass jellyfish or children’s doll. Thus, the scope of protection ought to be limited. The Second Circuit takes a “more discerning” approach when comparing products that contain both protectible and unprotectible elements, and the First Circuit would likely do the same.
Finally, the paper mentions the software infringement cases, which usually allow expert testimony on both prongs. The leading case, Computer Associates v. Altai, established the “abstraction-filtration-comparison” test. Lemley suggests taking a cue from the software cases in all circumstances. While expert testimony is often useful in considering the first prong, probative similarity, Lemley claims that “the second stage of the infringement inquiry – whether the defendant copied a more than de minimis amount of material that was in fact legally protected – does require analytic dissection of the work and expert testimony in the vast majority of cases.” (emphasis in original) In short, Lemley’s thesis is that we should sometimes consider unprotectable elements to establish copying, and allow more filtration and dissection, possibly through the use of experts or applying “specialized observer” standards, when determining whether the copying is improper.
The Other Speakers
The other professors generally agreed with Professor Lemley's persuasive paper. Professor Sonia Katyal commented on the idea of changing the ordinary observer standard depending on the type of work involved. The gist of her argument was that we should have a sort of sliding scale from ordinary to very specialized observer depending on the type of case. Works aimed at popular audiences, for example, could be compared using an ordinary observer standard but software would require a much more technical standard to the point where infringement sometimes becomes a question of law taken from jury. On this point, many of the commentators also spoke of developing an analogous procedure to Markman hearings in patent law. However, this would greatly raise the expense of bringing suit (perhaps not such a bad result), and the necessary questions cannot be taken from the jury under existing law.
I largely agree that it seems silly to have a jury consider technical questions of infringement, especially when there are specialized elements of a particular work or style of work at issue. On this point, Professor Alfred Yen brought up the obvious point that if the jury isn’t allowed to dissect a work at all, what else would they have to talk about? However you slice it, there will be some element of subjectivity involved. Lemley's paper implicitly criticizes Justice Holmes’ maxim that judges are not in the business of determining the value of works by noting that “if the problem is bad when judges are making the gestalt judgment, it is even worse with juries, who don’t have the experience judges do with disregarding facts they are told are legally irrelevant.”
(I’m not a literary critic, so the following is my best attempt at distilling Professor Yen’s discussion.) Yen went on to draw comparisons of copyright infringement cases to literary criticism. A formalist view assumes one encounters a work and determines its meaning without the influence of anything other than one's own prior experience. Intentionalism looks at how the artist’s intended meaning affects the experience. Finally, there is the audience-based view, where meaning is determined by a member of the intended audience for the work, such as the discerning 17-year-old. Professor Yen sketched an interesting timeline of cases, beginning with Nichols, where Judge Learned Hand famously dissected stock elements of the play Abie’s Irish Rose as unprotected generalities present in many dramatic works. Interestingly, Judge Hand castigates the overuse of experts, opining that they have no place at all in infringement cases. Yen's spectrum ends with more recent cases such as Dawson and Epix, where the audience-based view is at the forefront of the court’s consideration.
Finally, Professor James Grimmelmann provided another twist on the above, broadly agreeing with the previous panelists’ points. He stated that the tests for copyright infringement attempt to balance three categories, each consisting of things that either the plaintiff has the burden to prove or the defendant will raise as a defense. First, there is the idea of coincidence, or drawing on common sources. Second, the issue of unprotectable elements, or things the plaintiff does not own or control. Third, the issue of de minimis copying, which is not actionable, or alternatively, how much copying is improper. This third issue takes the first two into account. For example, two parties drawing on the same sources or stock elements of a genre very well might produce a similar work.
According to Grimmelmann, Krofft doesn’t provide much guidance because of its focus on the third issue while confusing the relevance of one and two. He also notes that although the Arnstein two-step might be sound, its wacky result (a serial litigator accusing the mighty Cole Porter of hiring “stooges” to steal his songs overcame Porter’s motion for summary judgment) might be the result of Judge Frank having struggled with the summary judgment standard, or just getting it wrong. Indeed, Frank applies the summary judgment standard of whether "there is the slightest doubt as to the facts." (citations omitted). The Federal Rules of Civil Procedure were enacted a few years before the case, and the Celotex trilogy, which greatly changed summary judgment practice, was still forty years away. Indeed, even the standard that the Celotex cases overruled provides a slightly higher standard than that applied by Judge Frank. Regardless, the copyright infringement test from Arnstein remains good law.
Professor Grimmelmann’s point about the correct standard for summary judgment made me think of the recently filed petition for certiorari in Aftermath Records v. F.B.T. Productions, a case about the proper construction of terms in a music contract as applied to digital downloads. The petition claims that the Ninth Circuit went out its way to ignore the plaintiff’s repeated forfeitures of various arguments in order to overturn a jury verdict they did not agree with. For a summary of the Aftermath case, see here. The petition seeks further clarification on the question of whether a party may appeal an order denying summary judgment after a full trial on the merits. Recently, the Supreme Court decided in Ortiz v. Jordan that where summary judgment is sought on a qualified immunity defense, the answer is no. After a first read, I don’t see Ortiz as limited to the qualified immunity context, and the three-judge concurrence suggests that the answer is and should typically remain a resounding no. (“Most such orders are not appealable at all, because they neither qualify as "final decisions" capable of appeal under 28 U. S. C. §1291 nor come within the narrow class of appealable interlocutory orders under §1292(a)(1).”) This result seems practical, predictable and efficient, especially in the copyright context. At issue across the board is the proper role of the jury, and what questions can and should be decided as a matter of law. As Professor Yen suggested, changing the role of the jury, judges or expert testimony might lead to better informed decisions but not necessarily more predictable results.
Activist by Default
After the presentation, the school hosted a reception where I had the pleasure of once again defending sanity in response to antiquated views on copyright that ignore the case law, statutory and policy developments of the past fifty or so years. It seems that whenever the words “Creative Commons” (CC) are mentioned, some people simply can’t maintain their temper or produce coherent statements. (For the first installment, see here.) I mentioned to an attendee that I once interned at CC and all hell broke loose. Sweet. Anytime you can’t get past the fact that CC licensing, copyleft and rampant piracy are not the same thing, it is usually best to politely excuse yourself and not engage your mis- or uninformed adversary (unless, of course, they're willing to listen and/or have a civil discussion about the merits of CC as one approach to the copyright difficulties presented by technology etc.). Piracy is a legitimate issue that unfortunately wreaks havoc on certain business models. Last and every other time I checked, however, CC and the Pirate Bay are not the same thing. Sure, there’s overlap amongst their supporters. But just because I’m a fan of the New York Yankees, and there are criminals who also like the Yankees, that doesn’t make me a criminal. (This is the first example I thought of and isn’t meant to suggest that piracy is automatically criminal!)
We started by discussing incentives, or at least trying to. I began by noting that the incentives to create provided by copyright do not necessarily have the same effect on all creators. I offered a few examples of where incentives are not purely financial. However, I was then reminded that the Constitution says copyright is to incentivize creation. I replied that it does not say “financially incentivize,” and that one must have been sleeping the past twenty years (e.g. open source software, remixing, hip hop etc.) and selectively ignoring oodles of court cases to not realize that our copyright system does not discriminate amongst those who rely on it for a living and those who don’t. From a constitutional perspective, even an originalist would recognize that “money” does not appear in the Constitution, and further, the early copyright acts that didn't protect all forms of works certainly didn't inhibit their creation. For a progressive, the scope of incentives surely would change over time. But I didn’t go down the Con law route. Instead, I tried to establish common ground by telling this copyright attorney that CC simply provides a few licensing options to creators to choose the downstream permissions they wish to grant. Well, she replied, why can’t a creator simply license their work? Umm.. they can. They can also avoid the hassle, time and friction of finding someone to write one if it’s a scenario that a public license, such as CC, is appropriate for.
Then the circularity of our discussion started making me dizzy. First, I was mockingly asked whether I’d heard of fair use. After replying that if one reads any of the fair use cases it's a relatively unpredictable judge-made beast as any first year copyright student will tell you, I was told that I was wrong, because fair use is written into the Copyright Act! I don’t even know where to begin commenting on that, but I pity the poor clients who rely on such misunderstandings of how the law works, and the poor attorney who thinks that CC licenses rely on fair use. Huh?
One must remember, all of this came after the presentation about copyright infringement discussed above. I guess not everyone paid attention. I tried to explain the paradox in her position that one cannot support the broad application of fair use and also think that authors should be remunerated for every possible use of or reference to their work, no matter how tenuous the connection. In other words, not all copying is illegal and four of the leading copyright scholars in the country just spent nearly two hours explaining the difficulties of drawing those lines! Then we circled back to piracy. “The [insert large copyright trade group] said that electronic sales are lost to piracy.” Even the U.S. government has admitted that many of the figures regularly tossed around are bogus and use questionable methodology that simply labels every possible use a lost sale, multiplies, and voila! billions of dollars down the tubes. Continuously trying to bring the discussion back to something positive, I once again acknowledged that creators should be compensated (not sure where/when I or CC ever said otherwise), that there is plenty of room for different levels/ecosystems of copyright, and that these new ideas are responsive to technology, in the same way that radio changed the licensing landscape. At this point I’d had enough. I encouraged my adversary to do her homework, offered a few resources and left to catch the train, pretty certain that encouraging sharing is a far cry from promoting stealing and glad that not all copying is illegal.