Harry Potter and the Lexicon of Fair Use

No, it’s not the eighth installment of the Rowling series – rather, it’s the latest installment of the ongoing legal fistfight over RDR Books and Steven Vander Ark’s attempt to publish a book version of the on-line guide to the Harry Potter wizarding world. (I posted briefly on this earlier, when I was annoyed by clueless coverage of the case by the NY Times and Joe Nocera.) The trial in the case starts 9:30AM on 24 March [update: it's been moved to 14 April, though I don't know why - thanks to Brandy Karl for the logistics tip!] before Judge Robert Patterson at courtroom 24 of the Moynihan Courthouse, 500 Pearl Street, in New York City. (Hat tip to Lauren Gelman of Stanford’s Center for Internet and Society for the information, and to Fred von Lohmann for distributing it.)

In short:

See below the jump for my analysis on the copyright issues. (I’ll defer the TM analysis for a day or so since I’m technically still on spring break and already feel like a nerd for writing this much.)

1. Copyright Infringement?

First, Rowling and Warner Bros. have to show a prima facie case for infringement – in other words, is it plausible to think RDR and Vander Ark have violated copyright law? To prove a violation of the reproduction right, Rowling / WB need to show 1) access by RDR / VA to the Potter works (easy) and 2) similarities between the Lexicon book and the Potter works that demonstrate copying of protected material. Since the Lexicon is nearly an encyclopedia of the Potter world, point 2 seems a lock. Thus, absent a defense, liability for infringement seems easy. (To be precise, there seem to be at least 3 theories for copyright infringement: 1) violation of the 106(1) reproduction right, for copying characters, plot, dialogue, etc. from the books and movies; 2) violation of the 106(2) derivative works right, for recasting or modifying the Potter works; and 3) violation of the 106(3) distribution right, for distributing the infringing works. Since there haven’t been public sales yet of the Lexicon book, claim 3 of these seems the weakest link.)

2. Fair Use?

Assume Rowling / WB make out a prima facie case for infringement. RDR and VA might still be off the hook if their use of the Potter works is fair under 17 U.S.C. 107. Congress set up 4 non-exclusive factors to evaluate a fair use claim: what is the purpose and character of that use (is it transformative)? What is the nature of the copyrighted work? How much – amount and substantiality – of the copyrighted work is used? And what effect will that use have on actual and potential markets for the work? (Importantly, these are non-exclusive factors – courts are free to find and use others – and there’s no explicit guide to weighing them. Fair use is like gumbo: you dump everything in together, mix it, simmer, and see how it tastes in the end. Lawyers call this a “standard” rather than a “rule” – “don’t drive too fast” versus “don’t drive faster than 55 mph.”)

I want to take these 4 factors out of order. Start with amount and substantiality (factor 3). The Lexicon incorporates a lot – if not most – of the Potter works. It’s an encyclopedia – if it wasn’t comprehensive, it wouldn’t be worth much. So, there’s significant copying from a quantitative perspective. Qualitative analysis seems somewhat less damning, since the book is more judicious about quotes from the sources, but it still clearly cuts against RDR / VA: there’s substantial copying of important expression. So, factor 3 cuts against the defense.

Now, nature of the copyrighted work: the Potter books / films are published, which helps a fair use claim, but they’re fictional, which hurts, since fictional works get greater (“thicker”) copyright protection. Factor 3 goes against the defense.

Market analysis: generally, market analysis (factor 4) vies with transformation (factor 1) for pride of place. The two also interrelate: the more transformative a new work, the more market harm to the copyrighted work courts are likely to tolerate. Here, market analysis has to be speculative: the book hasn’t been published yet. So, Rowling / WB have to project likelihood of harm, mostly by focusing on Rowling’s plans to write her own encyclopedia. (Interestingly, for you copyright geeks, note that this approach – by now the standard for courts in copyright cases – tracks Justice Blackmun’s dissent in Sony for market analysis, not the majority opinion by Justice Stevens. It’s a nice example of how a clear (though snippy) dissent can influence, significantly, future legal analysis.) Rowling’s lawyers clearly overstep by arguing that “[not] every person who purchases the [RDR/VA] Book [would] purchase a second encyclopedia, even if it is written by Ms. Rowling.” (Mem. in Support of Prelim. Inj. at 24.) The test isn’t any market harm, but rather the level of market harm relative to other considerations. My own take is that the market factor also cuts against RDR/VA, but that’s mostly because I don’t think the new Lexicon book adds much in the way of transformative elements, and hence we should be less willing to tolerate displacement of sales of the Potter books / movies, or future sales of a Rowling encyclopedia.

Finally, transformation. The RDR/VA Memorandum opposing a preliminary injunction sets out several arguments that the new book is transformative: it creates a reference tool, like a search engine (citing Perfect 10 v. Amazon and, by implication, Kelly v. Arriba Soft); it organizes a voluminous, scattered Potter universe into an easy-to-use supplement; it adds commentary and analysis of the Potter works; it unpacks obscure references and allusions; and brings in substantial outside research. The memorandum implicitly takes up economic concerns under the market (4th) factor here also, calling the Lexicon book an economic complement rather than a substitute. (See p. 12.)

For transformation, let’s dispose quickly of a red herring: RDR/VA’s work – and it’s extensive – in cataloging and assembling information about the Potter world gets zero weight in fair use analysis. Feist makes this clear: “sweat of the brow” copyright is dead – you get no protection for your work because of the effort involved in pulling it together. Rather, the key is the new expression you add – or, here, the new transformative expression. And I don’t think there’s enough of it.

The search engine analogy is intriguing but flawed. The Web operates largely on an implied license model: we let Google index sites (unless they opt out via ROBOTS.TXT) because society needs search engines and the transaction costs of negotiating a license with each site owner are prohibitive. Moreover, as Kelly points out, there’s no deliberate targeting of a specific work – search engines scoop up everything. And finally, indexing via search engines helps rather than hurts copyright owners: the thumbnail images in Kelly and Perfect 10 didn’t compete with the full-size originals, but instead pointed consumers to them. (And if I link to Perfect 10′s Web site, thereby pointing you to it, this blog goes from PG to NC-17 really fast.) The Lexicon book can advance this argument somewhat – it will probably drive some sales of the original Potter works, as most derivatives do – but it’s clearly less powerful than for search engines.

The best argument for RDR/VA is the outside research and the analysis / commentary. This isn’t a trivial argument, and I do want to protect non-academic research. But my subjective analysis is that, in the Second Circuit, there isn’t enough original analysis / criticism / commentary to save the Lexicon book. The primary purpose of the work isn’t to critique the Potter world; it’s to provide a guide to it. We might ask, as Justice Blackmun did, whether this work is “productive”: does it add sufficient value to society to tolerate its harms to Rowling / WB’s copyrights, and to the implied harm the precedent would do to other copyright owners? In short, does the Lexicon generate positive externalities? The answer is “Yes,” but I think Castle Rock, Twin Peaks, and Paramount v. Carol Publishing (11 F. Supp. 2d 329 (SDNY 1998)) show that the Second Circuit employs rather stringent standards for transformation that cut against a reference or supplementary work. In the end, I think the first factor – is the work transformative? – also cuts against RDR/VA.

4-0 looks like a clean sweep, as the Red Sox have demonstrated. If Judge Patterson disagrees with my analysis (not that he’ll ever learn about it!) on factors 1 or 4, it’s a closer call, as the inverse relationship between these factors would help the defendants. But I don’t think that’s likely.

Is this a good outcome? I think so. Remember that the 4 factors are non-exclusive. I’d argue Judge Patterson should consider an additional factor here: behavior by the copyright owner. Rowling has been supportive – very much so – of the Lexicon as long as it remained on-line and relatively non-commercial. To the degree that free speech concerns arise in this case (as the memo in opposition of the injunction argues, at p. 6), Rowling’s conduct mitigates those worries. (Eldred v. Ashcroft, which establishes fair use as constitutionally necessary, also notes there’s a difference between making one’s own speech and making someone else’s speech; 537 U.S. 186, 221 (2003).) She’s allowing this information to be presented to her fans and the public in general, while trying to minimize financial harm to her works. Copyright is often presented as a balance between incentives to produce and access to that production; here, Rowling’s approach seems to find that balance. When a use is commercial, it is slightly disfavored for fair use (copyright geeks: compare Sony‘s presumption here with Campbell‘s tacit repudiation of that position), though courts recognize that even fair use has to pay the bills. As a derivative work becomes more transformative, I think we should tolerate greater commerciality (because there’s less risk of substitution for the original), but where a copyright owner wields her legal entitlements not to block a derivative (as Acuff-Rose tried to block 2 Live Crew’s parody of Roy Orbison) but to cabin its economic impact, I think we should give that approach greater deference. In short, we want to encourage good behavior by copyright owners. (The plaintiffs and defendants take up this question: RDR/VA notes that Rowling/WB haven’t sued other compendiums, and Rowling/WB point out that we don’t want to press copyright owners to police fanfic in the same way that TM owners police their marks.)

As I’ve written, I would favor a change in the law that would immunize derivative works, and other, smarter folks advocate changes that would do away with injunctions in favor of a compulsory license / revenue-sharing scheme. But with the current state of the law, I think the Lexicon loses. In Potter terms, though, I think this is a triumph for Dumbledore’s Army, and not for the Death Eaters.

Stay tuned!

7 Responses to “Harry Potter and the Lexicon of Fair Use”

  1. What about RDR’s claim of estoppel (or is it laches? IANAL, I just read Groklaw) because of JKR praising the non-profit Lexicon and giving it an award, that giving the non-profit online Lexicon an award implicitly gives them permission to publish a for-profit non-online version?

  2. Thanks for posting this clear analysis!

    Matthew – RDR is claiming both laches and estoppel (and a number of other affirmative defenses). They haven’t specified what facts/arguments particularly they are alluding to with those claims, but we can probably guess.

    Estoppel: The standard as applied to this case could go something like this:
    If RDR/SVA had an expectation based on JKR’s statements or actions, and reasonably relied on that expectation, and acted on that reliance, and would suffer a detriment if the expectation were false, JKR could be estopped from enforcing legal rights.

    Here, I assume RDR would argue that JKR commended the site and awarded it a fan award, and SVA thus expected that JKR would approve of his project to produce the Lexicon in another format, and acted on it, and now RDR may have put time and money into preparing to publish (it’s unclear what stage of production they were at, so who knows) and now they’ll argue if they can’t print, they’ll be harmed.

    One key point here is “reasonable” reliance. Was it reasonable for SVA to think that, because JKR commended him for creating a really good free fansite (something she did for at least 2-3 other fansites), she would approve of him publishing that work in a different form for profit? Even after he approached her about collaborating on such a project and she turned him down? And then her lawyers contacted RDR to try to stop publication, well before the publication date? Yeah, I don’t think that’s “reasonable” reliance. I think there were plenty of clues in advance that this would not go down well with JKR.

    Laches is a form of estoppel that is based on a delay in the other party asserting its rights. In this case, RDR might argue that by JKR/WB waiting until right before the book publication date to file suit, instead of earlier, RDR had acted on the belief that it could publish to a point where it is now in a worse position than before – for example, if RDR had prepared for and/or begun printing of the books, and then JKR won the case and there was an injunction against distribution, RDR would be harmed by losing the time and money it put into preparation and printing, and by having to deal with deals already made that now can’t be carried through (i.e. rights sold to publishers in other countries) for which it may even have already received some payment.

    On the other hand, if JKR’s lawyers can pull out the 3-5 letters they claim they sent trying to meet with RDR and work something out, and can prove the delaying tactics that RDR supposedly used in order to avoid talking to them, then RDR really has no argument. It’s likely that WB/JKR can show their delay in asserting their rights was based on a good-faith effort to work something out, and that they filed suit late in the publishing game because they had been trying to avoid a suit at first. The judge isn’t going to fault them for that (I would hope).

    One interesting thing I remember reading is that RDR had consulted with its lawyers and been told it was not infringing, or something to that effect, during the time right before suit was filed. If that’s the case, RDR must have talked to the worst lawyers that ever passed the bar.

    (Sidenote: I love how RDR is claiming the doctrine of unclean hands as an affirmative defense. As the internet cats would say: “WHUT?” That’s just ridiculous.)

  3. Matthew,

    Praise is not implied consent, and the Christopher Little Agency and WB have monitored the Lexicon site over the years. On several occasions, they asked Steve to remove certain information for various reasons, and he complied.

    From West’s Legal Thesaurus/Dictionary (c. 1985): “Laches arises when there has been an unexplained delay in asserting a right of such duration and character as to render the enforcement of the right inequitable.”

    I will quote from an online user at The Leaky Cauldron’s Leaky Lounge by the name of dresdenfilesfan, who puts this very succinctly:

    “The book contract between Vander Ark and RDR was signed August 23, 2007; there was no monetary advance. According to Roger Rapoport’s private emails dated September 6th and October 5th {2007}:

    1. “We do not want to sell this to any of the publishers currently publishing the Potter books.”
    2. “I am assuming you will NOT be discussing this book with Editorial Presence, the publisher of Harry Potter and the Deathly Hallows. Please confirm this. They are not a candidate for this book.

    In regards to a flyer to market the book, Rapoport states in an email date September 30, 2007 “I do want to have it ready to go on short notice and it sounds like you are close.”

    RDR deliberately kept this book below the radar to keep Potter publishers out of the loop. One is dated prior to the first C & D, and one is dated after RDR asked for extra time to respond because of a death in his family. The marketing flyer was authorized by him all the while knowing that WB/JKR were issueing him C & Ds.

    Then there is the Publisher’s Market Place site, which lists the page generation date of September 6, 2007 and a page update of September 10, 2007. Neil Blair stumbled upon the listing and sent off an email to Steve and RDR on September 12. After waiting six days with no response, he sent another email on September 18th. From that point there were several more emails, formal letters, and informal phone calls all attempting to get RDR to stop publication and/or allow them to review the book.

    For RDR to claim that WB/JKR didn’t act soon enough to prevent financial harm to them is completely absurd. They asserted their rights as soon as they became aware that their rights were being violated.”

    For a number of years, Steve Van Der Ark had personally eschewed enquiries with regard to a print version of the lexicon and emphatically denied that he had any plans to do so, claiming he felt it would be a copyright violation. In several of the Plaintiff’s exhibits are email exchanges of this nature, and one or two in particular where he discourages others from publishing their own Encyclopedias. CLLA, JKR, Bloomsbury/Scholastic and WB were all aware of Mr. Van Der Ark’s stance.

    I would also direct you to a couple of interesting entries in the Lexicon’s FAQ, as well as in their What’s New section:

    From the FAQ: http://www.hp-lexicon.org/help/hp-faq.html#copy

    From the What’s New section:
    http://www.hp-lexicon.org/whats_new.php?year=2007 (you have to scroll down to the May 11th entry: Something I had to do.

    This is no simple, cut and dried case for either side, however, I would submit that both the Defendant RDR and Mr Van Der Ark have been less than forthright, and their inclusion of both “Estoppel/Laches” and “Unclean Hands” Affirmative Defenses in lieu of the above is fairly laughable.

  4. Hi Emily, Hi Matthew,

    The short version is that I think Emily’s analysis is right. Laches is a standard affirmative defense to raise, and courts treat it with a great deal of skepticism. (For legal geeks, it feels in this sense somewhat like adverse possession in property law – open and notorious violations of one’s rights, if left unchallenged over time, can diminish or extinguish those rights.) The brief period of time that Rowling / WB waited isn’t anywhere near laches. The argument is a waste of time, in my opinion.

    I think the estoppel claim founders based on the course of dealings (if they can be called that) between the parties: Vander Ark was surely on notice that the book wouldn’t meet with Rowling’s approval, and it would be a stretch indeed to construe positive comments on a fan Web site as a license to produce a printed encyclopedia version of it.

    And I’d be deeply surprised if RDR’s lawyers had, in fact, told them that the book either didn’t infringe or was protected by fair use. Fair use is simply too nebulous a doctrine to make any such assertion with legal precision, and the encyclopedia clearly infringes on its face. I’d guess that RDR’s lawyers told their client that the publisher and author had a good, but not clear-cut, fair use claim, and that RDR / VA went ahead based on that. If the lawyers said, “No problem, the encyclopedia doesn’t infringe at all!” or “It’s definitely fair use!”… well, that would be malpractice in my opinion (which is why I doubt it happened).

    Having seen the work that a patent firm puts in before issuing an opinion letter on non-infringement, I’d be shocked if copyright lawyers (who work in a more doctrinally murky area) were any less careful.

  5. Hi Derek,

    Thanks for your further comments on this.

    I agree that the laches defense is very weak – I’m also inordinately amused by the image of SVA as a squatter.

    For clarification, I didn’t mean to make it sound like the lawyers told RDR it was definitely fair use or completely safe. I’m sure they didn’t. I wish I could remember where exactly I read something to the effect of “RDR had consulted with its legal advisors and was told they should be able to go ahead with publication.” I may be remembering wrong, but I really thought I read that in one of the docs available on Justia.com. Anyway, even if I did read that somewhere, it’s likely that the attorneys did say something much more cautious, and RDR just interpreted that in layman’s terms as, “it’s ok to publish!” I wouldn’t be surprised.

  6. [...] considerations — as shown by the judge wanting to move forward from this issue. In the Info/Law blog, Derek Baumbaeur, discusses how an additional fair use factor should be considered regarding the [...]

  7. [...] any help in my view from Derek Bambauer of the Harvard Law School. Even before the decision he had pretty much decided that the Lexicon didn’t deserve to be covered by fair use, because it failed to meet any of [...]