Discussed ongoing Hathi Trust litigation concerning library copies produced as part of Google’s book scanning project. Hathi Trust (with repositories at Michigan/Indiana) has about 10 million volumes, Google scans as well as library digitization projects. What can the Trust do with these volumes now that there’s no settlement? Access for the visually impaired, possibly; preservation. Potential orphan works project: let one user check out a copy at a time if their library had contributed the book to the database. Complaint targets not the orphan works project but the entire project.
Corynne McSherry, EFF
Sovereign immunity is a big barrier. Can sue state officers, ordinarily with the same effect.
Standing: default principle in copyright is that you can only enforce the rights that you have, not other people’s rights. Authors Guild sued on their own behalf and there were some individual authors. Even suing on behalf of the membership, not that many books are by AG members. Did manage to find one author to join the lawsuit whose works were previously considered orphan works. Does this lawsuit create a lot of leverage? Procedural difficulties. Potential laches defense.
What is the lawsuit over? The scanning (because the delivery of information about what words are on what pages from the corpus is just facts). The Hathi Trust’s fair use argument is good. Copying the entire work is ok if you need it for your purpose, which here is to create a corpus. Limited access to researchers and students aids in the positive purpose. As for the orphan works project, the market harm argument is quite interesting. There’s a policy argument that there can be no market harm and allowing people to find the work can enable identification of authors, as with the one guy who joined the Authors Guild case. An opt-out aids the fairness argument as well. Authors will say that it’s all about licensing.
Allan Adler, Association of American Publishers: because of love/hate relationship between authors and publishers, will be discreet in commenting on legal theories of AG case and why publishers aren’t participating in the case; note that he is under nondisclosure agreement with respect to Google settlement discussions which are ongoing. Applauds attention to old books renewed by digital environment. Google was unwilling to talk with the publishers about copyright; Google had billions of dollars and wasn’t willing to change its interpretation of fair use. Lawsuit = way to get Google’s attention in a serious way, as is often the case.
AG case against Hathi Trust—not the same. Hathi was perfectly willing to talk to the publishers about what was being done with the digital collections, including the Google scans. Because we saw the beginnings of a dialogue that’s continued, there was no rush to sue in order to get their attention.
Books are not dying, they’re transitioning to other formats and means of creation.
Users, even in consortia, are not the ones who should be making decisions about when to use an orphan work without the author’s permission. This is a job for Congress. Van Gogh never sold a work in his lifetime; it’s possible that works that went out of print because it wasn’t worth investing further in them may now find a market. Authors & publishers should be given the first shot at trying to see whether those works have value in the new world.
[He’s very good.]
Kevin Smith, Scholarly Communications Officer at Duke University: Licensing schemes to solve orphan works—what might they look like? His guess is that the AG suit is a way to make sure we don’t have a practical illustration of a way to exploit orphan works in a noncommercial way without licensing.
What would it mean for Congress to solve the problem? Virtually every country has used a licensing scheme; the alternative proposed and failed in the US in 2008 would have limited damages. If we’re talking about a solution that might make money for powerful groups, then don’t laugh at the possibility of Congressional action.
A licensing scheme inevitably gives someone who is not the rightsholder a right to sell a license. Canada has a licensing scheme. Good article about it. Only for published works; no formal standards for a diligent search, but informal standards evaluate the type of work and the use proposed—look like fair use factors. Licenses always require payment. Usually the CRO is to hold the money for 5 years, after which that money reverts to the general funds of the CRO. 60% of applications wanted a single work, but a handful of licenses were granted for large numbers. Extended collective licensing. (We’d have to have a cultural policy to do this, wouldn’t we? In the other countries, the funds are generally supposed to go to cultural development; hard to imagine we’d do that here unless we equated cultural development with “giving large sums to a private owners’ organization,” which I certainly won’t put off the table.)
No matter how low the cost is set, it inevitably discourages beneficial uses of obscure works. These are exactly the works that somebody needs, but often only one or two people. Who should administer this license? Unlikely to be a gov’t body in the US. Could a CRO like the CCC serve as an honest broker? (Hisses from the audience of librarians.)
How would orphan works schemes relate to other user rights, particularly fair use? Would the board make a fair use determination before determining whether or not a license was appropriate? What about the other exceptions in the copyright law?
Adler: doesn’t believe that fair use can justify an entire system/process/set of procedures for determining what works are orphans and when one can go ahead and what the consequences should be if a copyright owner emerges. In an individual case, of course, fair use would still be a valid defense if in fact a copyright owner emerged. His argument: what the University of Michigan seemed to be doing was saying that the entire orphan works project is fair use based and that goes too far.
Smith: yet Michigan hasn’t made one orphan work available to one student. The lawsuit is based on asserted risk.
Adler: but we do know their intent—designation of a work as an orphan work was declaration of intent such that if a copyright owner didn’t protest within 90 days they could use the work. McSherry says the use would be limited to students and researchers, but that’s insufficient. If it’s an orphan work, you should be able to make it available to others, as long as that’s done pursuant to law with procedures in place.
Smith: but in the absence of orphan works law, we think that a more limited solution comports with fair use.
Band: Adler says you can’t use fair use as a systemic solution to a problem. That’s wrong; courts have applied fair use to business models. Search engines. iParadigms and its plagiarism detection software. The Betamax.
Adler: the search engine cases shouldn’t have relied on fair use but on implied consent. Google respects opt-outs. And you can’t cite a case that precedes the digital era as relevant. That was 5-4 anyway. (As the late, great Judge Becker said about panels, though, 2 is more than 1.)
McSherry: two underlying problems—the copyright term and statutory damages. Congress has done more to create this problem than it’s ever done to solve it. Fair use looks better in San Francisco.
Adler: takes a (well-deserved) whack at the SCt’s sovereign immunity jurisprudence, which is misguided and distorts the possible responses to use of materials by a public entity that knows it’s not subject to statutory damages.
McSherry: but copyright liability is divorced from any actual harm, and that’s part of the problem. Fair use is a backwards way to think about actual harm, but that’s only because we started from this premise that copies generate liability. Downstream uses—filmmakers, documentarians, etc. who want to get this stuff out there still have to worry. Sad and ironic that as tech progresses we’re shutting those mechanisms down.
Smith: one of the reasons that there’s a conflict between publishers and universities is that copyright works so badly for academic authors. Editors and academic authors don’t make money; academics don’t benefit from the copyright system as it stands. They may benefit from control, but not money.
Q: what about a threshold trigger—if you get a certain number of downloads then there should be payment. If authors are worried about potential markets, why not wait and see if that potential materialize?
Smith: thinks a licensing board could do something like that. Zero license for a certain period, then reconsideration if changes occur.
Band: Windfall profit scheme might work. Instead of fearing the possibility, let’s keep costs down up front and if unfairness develops find a way to compensate that person. Even in the Hathi Trust framework, that unfairness can’t develop because it can only circulate within the community. If a book is suddenly in demand even though it hadn’t been checked out in 10 years, then you could treat that differently. It is so unlikely/rare, so we shouldn’t prevent uses on the grounds that the situation might development.
McSherry: 99% v. 1%.
Adler: Orphan works is a misnomer. We’re really talking about orphan uses. The one thing that’s clear is that no one believes we should designate a work as an orphan for all uses. Rights are inevitably divided. We’re talking about a particular use that doesn’t fall within an exemption; that may be orphan because you can’t find the person who can give permission for that use.
Q: Authors Guild defense. Won’t trust Google to treat their works fairly. If one of these things goes forward, benevolent/nonbenevolent, commercial/noncommercial, authors should be at the table, like a union, participating in the structure. AG doesn’t speak for all authors, but does demand a place in the conversation. Maybe they didn’t do it right, but they did it—somebody powerful is coming to the table even though I don’t have the power, money or time to do it myself.
Band: Authors were invited to the table; publishers took up Hathi’s offer and AG didn’t.