Google’s legal winning streak continues with last week’s decision by the Second Circuit Court of Appeals, which rejected the Authors Guild’s appeal of the dismissal of its lawsuit against Google over the latter’s massive book-scanning project.
I won’t take up space here recapping the timeline of this now-ten-year-old legal saga – there’s a pretty good summary at Wikipedia. But I’d like to share some thoughts about the rejection of the Authors Guild’s appeal.
First of all, the appeals court’s central finding is that “the defendant’s copying is transformative within the meaning of Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 579-585 (1994), does not offer the public a meaningful substitute for matter protected by the plaintiffs’ copyrights, and satisfies § 107’s test for fair use.”
Why? Because the copying that Google did involved something very different from the photocopying, file-“sharing”, or home taping activities that have given rise to so many previous important copyright lawsuits. Google was not simply creating a bunch of new copies of books, intended to be used in the way the originals are. But this issue goes beyond Google’s intentions. More important, I think (and apparently more influential on the appellate court’s decision in this case) is the fact that the resulting database of digitized books simply cannot be used the way the original copies of those books can be used. No one can read the copies contained in the Google Books corpus; what you can do is search their content and locate terms, and in most cases see very brief snippets of the text.
In other words, Google’s use of the copyrighted works was, in the court’s view, clearly transformative for precisely this reason: what were once books have been transformed into databases — or, more accurately, a single massive database — which offer tremendous public benefit without posing any significant threat to the rights or legitimate interests of copyright holders. Access to the Google Books corpus is not, by any reasonable standard, the same thing as access to the books themselves, and therefore doesn’t constitute a competitive substitute for them. That makes all the difference in the world when it comes to determining fair use.
But wait. Google is a highly profitable commercial entity. How can its wholesale copying of copyrighted texts, undertaken for an explicitly commercial purpose, possibly be considered fair use? Here the court points out something that many of us in libraries have tried to help students, faculty members, and sometimes even publishers understand for many years: the fourfold test for fair use doesn’t contain any trump tests—fair use is determined by considering the proposed use holistically. Professors and students who take the “Purpose and Character of Use” test for a trump card and therefore believe they can do anything they want with copyrighted works as long as the use is educational are falling victim to this misunderstanding, but the same is true of copyright holders who believe that fair use is always fatally undermined by commercial reuse. As the court puts it in this decision, “Our court has… repeatedly rejected the contention that commercial motivation should outweigh” an otherwise convincing fair use argument. In this case specifically, “we see no reason… why Google’s overall profit motivation should prevail as a reason for denying fair use over its highly convincing transformative purpose, together with the absence of significant substitutive competition, as reasons for granting fair use.”
But wait. The plaintiffs in this suit aren’t only concerned about what Google is doing intentionally with the digitized books, but also about what might happen if someone were to hack into Google’s server, copy the books, and then do things with those copies that genuinely infringe on the copyright holders’ prerogatives. In response to this concern, the court noted that Google has stored its digital scans “on computers walled off from public Internet access and protected by the same impressive security measures used by Google to guard its own confidential information.” While the court allowed that the plaintiffs’ concern in this regard is “theoretically sound” (since it describes a scenario that could actually take place) it felt that the likelihood of that fear being realized was too low to warrant a finding in the plaintiffs’ favor.
But wait. What about the digital copies that Google left with their library partners? For one thing, the libraries themselves might make infringing use of the copies; for another, the copies held by libraries aren’t protected by anything like the industrial-strength security measures Google can afford. The court recognizes the possibility that libraries might fail to live up to their copyright obligations either by active infringement or by negligent handling of the copyright materials, and observed curtly that if they should do so, “such libraries may be liable to Plaintiffs for their infringement.” In the meantime, “the possibility that libraries may misuse their digital copies is sheer speculation” and no basis for a finding against Google.
Here’s something I think will be interesting to watch: many in my profession are every bit as suspicious of Google, its motivations, and its plans for our information future as they are of the motivations and plans of publishers. What will be the library profession’s response to this finding?
Some responses are already coming in: laudatory quotes from the American Library Association, the Association of College & Research Libraries, and the Association of Research Libraries — all members of the Library Copyright Alliance — are compiled (for attribution) here. The California Digital Library likes it too. So far, it looks like my colleagues mostly agree with me: this finding is a win for Google, for libraries, and for researchers.
The Authors Guild may not believe it, but on balance it’s almost certainly a win for authors and publishers as well.
Discussion
24 Thoughts on "Another Big Win for Google Books (and for Researchers)"
It does seem to me though that this judgement significantly increases the liability risks for libraries, if they don’t have adequate security against hackers (cf Aaron Swartz) or do not take reasonable steps to ensure that copyright is protected. Librarians may need to display more respect for authors’ intellectual property rights than their uncritical passion for OA to everything would suggest to date. As the judgement notes, there is a big difference between the creation of a giant database and the expropriation of scholarly, and other, property rights. I can understand librarians’ frustration that the finitude of their resources means that they cannot offer access to everything. However, the answer is not necessarily to compound the abuse of content creators that has emerged as the internet has developed.
Robert, can you explain how OA advocacy equates to a disregard for intellectual property rights? Or perhaps provide a concrete example of how librarians seek to “compound the abuse of content creators”?
I think Harnad & Co advocate immediate deposit of all subscription articles in an open repository, without publisher’s permission or embargo. That might be considered disregard for intellectual property rights. In fact he openly advocates the destruction of the subscription journal industry by this means, as do others.
Harnad certainly advocates for policies that would allow authors legally to self-archive immediately and without embargo. I’m not aware that he has advocated such archiving when it’s not allowed by copyright or license. If he has, I’d be genuinely interested in a citation.
Harnad’s mantra has always been immediate OA deposit. He says it over and over. In fact I have debated him several time on this. I have never once heard him mention publisher approval. If you have a case of him saying it I would love to see it.
First of all, “publisher approval” is a non-issue; what publishers want or approve of doesn’t matter. What matters is the terms of copyright law and/or of the licenses or transfers that govern reuse.
Second, Harnad’s mantra is indeed immediate self-archiving, but I don’t believe he has ever advocated that it be done in breach of copyright law or transfer agreements. You’re the one accusing him of “disregard(ing) intellectual property rights,” so the onus is on you to provide examples of him doing so — it’s not on me to provide examples of him not doing so. (Nevertheless, you will find instances of him calling for self-archiving policies and licenses here, here, and here, just to cite three very easily-found examples.)
What Harnad has always advocated is immediate self-archiving. What he has NOT advocated is ignoring publisher-imposed embargoes. In the time that an embargo exists, Harnad suggests a work-around: using a mechanism to send an email directly to the author and request a copy that way. He blieves this to be legal, in line with the traditional practice of researchers sending copies of their artkicles to other researchers who request them, and I tend to agree with him, though I would not call this fair use, as he once used to do. (Apparently, i convinced him to drop that term for the mechanism; he used to call it the “fair use button.”)
In response to the ability of Google to protect its information: SNOWDEN!
It’s certainly true that anyone can be hacked, given a sufficiently skillful and motivated hacker. But that’s not really an argument against Google Books — it’s an argument against entrusting any valuable, sensitive, or important information of any kind to a networked digital environment. Feel free to make that argument if you’d like, but you won’t find many takers for it.
Firstly, OA requirements on scholarly publication expropriate the author’s copyright in their work, especially when certain Creative Commons licences are mandated. Authors lose control of the integrity of their work and the right to derive benefit from the sale of secondary rights, in the case of articles, or primary rights in the case of monographs. This applies whether the requirements are the result of contractual terms imposed by funders or employers or by attempts to use ‘soft law’ methods that pressure authors to release rights without compensation. In the UK, we are seeing this with HEFCE policies that then transmute into university policies that only OA publications will count for certain kinds of recognition and reward. There is no statutory basis for these policies, nor are they incorporated into contracts of employment as a result of collective bargaining and compensation for the potential economic losses involved. Recently, I have been refused publication of a paper in conference proceedings because I chose to insist on my copyright, although OA was not a condition of the invitation to present. As someone who largely works as a freelance, I can live with this exclusion but other presenters did not have the same choice and felt bullied into compliance.
As a member of the UK Authors Licensing and Collection Society, I do see quite a lot of the other side of the impact of these policies on content creators. Traditionally, in my experience, librarians were prominent defenders of content creators’ rights in the xerox era, as I recall from the punctilious behaviour of staff at the Bodleian and other major university libraries in response to copying requests. This is much more difficult to achieve online but this is not necessarily a reason to abandon the attempt. Again, referring to the UK, I will confess to having crossed swords with one of the leading librarian advocates of OA as long as 20 years ago in an faculty senate meeting and being told that my concern with copyright and the integrity of my work in the contexts of reuse was obsolete and irrelevant. My concern to be compensated for the loss of income from those rights was ridiculed, although, at various stages in my career, it made a modest but useful contribution to the low salaries of a junior scholar in social science/humanities. As far as this person was concerned, academic content creators simply did not have any rights in their work. Similar views have been expressed to me more recently in committees that I have attended as a representative of the UK Academy of Social Sciences, particularly by representatives of a major research foundation supporting work in life sciences and associated social sciences/humanities. I am also noting the content of the courses being run by the librarians of the institutions that I collaborate with, which are intended to promote compliance with the various UK policies in the area. I do not criticize them for this – it is what they are employed to do – but it is notable that the courses do not encompass those areas where copyright is still relevant. I do not, for example, see courses on the evaluation of publisher’s contracts for textbooks or monographs with a potential crossover market. Given the increasing opportunities for work in other media, I would also have thought it might be an emerging role for librarians to alert faculty at all levels to issues with TV or internet video contracts: it is not a matter of substituting for an agent’s role so much as alerting people to the issues, the need to protect their, financial and other, interests, and the potential value of an agent as a professional adviser and intermediary.
I don’t know whether these examples are specific enough to meet your requests, Rick, but some of them are attended by concerns about confidentiality and a desire not to compromise individuals, many of whom have less freedom than I do to question something that they are being asked to deliver. One of the reasons I read Scholarly Kitchen regularly is that it is an environment where honest debate can take place in a way that is shut down elsewhere. Clearly, you could respond by saying that I have just had a bad set of individual experiences. However, I would hold that they are representative of the reluctance to see scholars as ‘content creators’ alongside authors, songwriters, musical performers and other cultural professionals with a legitimate interest in protecting their work and securing whatever revenue stream it may be capable of generating.
I actually agree with you about the impact of compulsory CC BY on authors’ copyright prerogatives–the thing is, there’s still a lot of controversy in the OA community over the degree to which OA necessarily has to mean CC BY. The more people learn about what CC BY means, the more controversy there is (yes, even among librarians) about whether or not it should be forced on authors by funders, publishers, or institutions.
The problem with your examples of librarians being abusive of content creators isn’t that the examples aren’t specific, but that they aren’t (I don’t believe) representative of the profession you are characterizing as dismissive of authors’ rights. Speaking as a librarian who has been in this profession for over 25 years, I can say that my experience suggests that the great majority of my colleagues do, in fact, see scholars as content creators with a legitimate interest in protecting their work. (There’s a lot of debate as to what kinds and degree of protection are legitimate, of course.)
I think it is a fair point, Rick, that OA does not necessarily entail CC-BY, and one I probably skated over too quickly in trying to draft a brief post. However, while you could have OA with retention of copyright, this does compromise the potential for authors to extract value from the sale of secondary rights in publication. This is not particularly relevant in the STEM fields that have driven a lot of OA debate, where patents are the really valuable intellectual property and need not be compromised by OA to scientific publications. The STEM folks can still make their millions while giving away their journal papers. In SSH, copyright is the intellectual property with value. Mostly it may only be worth low hundreds to low thousands of dollars per year but it does, for example, allow for occasional windfalls where work is picked up for TV or movie purposes.
As for the librarian profession, we can simply agree to differ. Like any other, it contains many decent and honourable people as well as its quota of cranks and zealots. I have worked with some great representatives of your community over the last 45 years – and some less pleasant ones. That’s just how it goes.
Let me explain, Rick, why this decision is not a win for publishers. The key fact for me in this decision is that it was written by Pierre Leval, who is credited with inventing the idea of “transformative use” in his 1990 Harvard Law Review article, which was cited more times by the Supreme Court in the Campbell case than any other source and became the linchpin of this SC ruling, as it did for many subsequent court decisions about fair use. My reading of Leval’s original article led me to believe that the key to understanding fair use is the act of human creativity involved in adding new meaning and value to a previous work used for this transformative purpose. That is exactly what was at issue in the Campbell case involving a parody of a song. The Ninth Circuit, however, starting with cases involving thumbnail images used to create an index, relied on a situation where the index was created by an algorithm in a computer’s operation, no more creative than the act of pushing a button on a photocopy machine. In the Texaco case, where Leval was the district court judge, the judge on the appeals court who wrote the decision finding infringement specifically acknowledged the “social utility” afforded by photocopying but argued that, as a mechanical operation, it had “nothing to do” with fair use as traditionally understood (because that traditional understanding always focused on an act of human creativity). In the Perfect 10 case the dissenting judge took his colleagues on the Ninth Circuit to task for taking it upon themselves to decide whether the “social utility” of something should be used to create public policy, since that is usually the role of a legislative body like Congress. Before the HathiTrust case came along, all of the decisions in the Second Circuit using transformative use involved acts of human creativity (Bill Graham Archives, Richard Prince, etc.). But Judge Baer, citing the Ninth Circuit decisions, was the first in that circuit to import the quite different approach of the Ninth into the Second Circuit’s jurisprudence. Then Judge Chin followed up with his ruling in the Google case at the district court level (where he said that the scanning of books “allows for” creative uses, which is a tacit admission that the scanning and indexing themselves are not creative but only preconditions for creative activity), and now Leval himself has endorsed that interpretation. One consequence is that there is no longer an apparent split between circuits that would give reason for the Supreme Court to consider an appeal. Another consequence is that it is now more open than ever for the groups that supported the Code of Best Practices on Fair Use for Academic Libraries to continue using Jonathan Band’s argument that the re-purposing that is involved in using digitized journal articles and monographs, originally written to advance research and therefore aimed at other scholars, for a new pedagogical purpose in teaching students in undergraduate classrooms. This argument sanctions the provision of sometimes entire texts of articles and monographs (and yes, even novels since their primary audience is the general reader) if the teaching purpose can so justify. That involves the fourth factor, of course, but Band et al. would argue that the primary market for such works being other scholars, it is not affected by using such works for the classroom; the only works prescribed from such use would be actual textbooks and other materials published for use in the classroom in the first instance. I know (from private correspondence several years ago) that Leval himself would not accept this extension of transformative use, but since it is unlikely Leval would be the hudge assigned a case involving such use, it is quite possible for another judge (like Baer, who thought making digital copies for the physically handicapped is “transformative”–a part of his ruling that was overturned by the appeals court) to accept the Band-type argument. The result then would be the collapse of the paperback market for monographs, which for the press I used to direct at Penn State constituted 40% of our revenue. So, I respectfully disagree that this is a win for publishers. However, you should be aware that the AAP and other groups representing content owners do support the Copyright Office’s proposal for separate legislation to deal with mass digitization (and also orphans works) rather than allowing these complex issues to be approached only through the very broad (and often fickle and unpredictable) avenue of fair use. It would appear, unfortunately, that this “victory” is only going to lead library groups and some others to resist such legislative proposals and want to continue relying on judicial solutions.
If I’m following the various threads of your argument here, Sandy, I think I hear you making two points in support of your belief that this decision is bad for publishers:
1. The Ninth Circuit doesn’t have an accurate understanding of fair use, but it does have influence on other courts.
2. This decision will embolden people who agree with Jonathan Band about what constitutes transformative use, and if they make such an argument in court it might be with a judge other than Leval (who you believe disagrees with Band).
Is that about right? If so, I have to say that point #1 seems largely irrelevant to this case, and point #2 seems like (in the words of Judge Leval) “sheer speculation.” But I’m fully prepared to be told that I’ve missed something in your essay.
Well, I wouldn’t put the points in quite the way you do, but basically you got the gist. On (1) I would rather say that the Ninth Circuit has a wrongheaded approach that conceptually muddies the waters in a way that I had thought Leval had previously made clear (but Leval doesn’t see it the way I do, obviously, now that he has sided with the Ninth Circuit). The loss to conceptual clarity stands, however, in my opinion. It is the fact that Leval could have separated himself from the Ninth circuit that makes the fact that he didn’t relevant to this and future cases. On (2) I do not see how Band et al. cannot be encouraged by this decision because it provides conceptual support for their position. Whether a case along the lines of what the Code recommends ever comes to court, of course, is mere speculation. No one can predict the future in that way. What is predictable is that fair use cases will continue to crop up and they will be difficult to predict as to outcomes just because of the conceptual confusion that exists.
While I don’t agree that anything that benefits authors necessarily benefits publishers… that is a stretch even for an academic… I truly find it difficult to trust Google. Let me explain why. many years ago when Google was first ‘exhibiting’ books on their site in ‘snippets’… I was sent instructions on how to hack the book site and get anywhere from 90 to 100% of any title I wanted to read. Since I never tried to ‘take’ any title away from the site, I cannot respond to whether the titles were downloadable at that point, but reading 90-100% of a book for free doesn’t seem to be the firewall publishers need to rely on.
When I went to BEA one year and discussed the issue with Google engineers and sales people, they told me it wasn’t possible. I took out the instructions and did it in front of their eyes. THIS finally got them on their toes (after repeated attempts via email and telephone) and after, this ‘hole’ was closed. I am hopeful this was a one-off…but again, I hardly would trust my cache of titles to Google’s “altruistic interests” (is that an oxymoron for Google?).
I truly hope they will act as the court believes they will. If not, we have just handed the keys to the hen house to the fox.
Years ago publishers were permitted to ‘opt out’ and not permit Google to digitize their titles. I haven’t followed this phase as carefully, so i have no idea if that firewall is still in place to protect skeptical publishers. But if it is, those of us who still don’t like the world having unfettered access to that which we have paid handsomely to create and market… should file the necessary documents to ‘opt out’.
I think we can all agree that Google has a strong obligation to safeguard the intellectual property that it has appropriated with this project. If it fails to do so, then those whose interests are hurt by that failure should sue and Google should be held responsible.
You must be joking. Google allowed between 50 and 100 MILLION dollars to win this lawsuit when ti started. They were prepared, according to some internal unnamed sources, to spend in excess of 250 MILLION dollars…that is how important this case was. You actually believe ANY publisher short of the Big Five has the resources to fight this behemoth? You must be joking, and with a short memory. When this case started and it was only the Publishers fighting Google… there was no settlement… until the US Justice Dept stepped in… the only people with more money and resources than Google. THEN… Google was ready to settle…and we see how THAT has worked out for the publishers now. Yes, of course, I just sue Google and get bankrupted trying to win justice from an unlawful gang whose motto is ” It is better to ask forgiveness than permission.”
You must be drinking Google Kool-Aid. I advise you take a look at the hard facts.
It cost over $3 million in legal fees for publishers to win the Texaco case about photocopying–and that was about 20 years ago!
And Texaco had only a tiny bit of ‘skin’ in the game, with little profit motive to pursue. Google’s position is far more long term and profit motivated. 100 Million is a drop in the bucket if you look at One Billion Dollars as the potential down line.
You actually believe ANY publisher short of the Big Five has the resources to fight this behemoth?
Well, yes — because that’s what just happened. The plaintiffs in this case had only a tiny fraction of the resources that Google has available, and yet they did fight Google and got their case heard in federal court. So unless it’s your belief that Judge Leval was in the bag for Google, I think there are only two possible explanations for the final outcome: either Judge Leval made a good-faith ruling based on his genuine understanding of copyright law (indicating that the system worked as it should), or he was bamboozled by the fancy talk of Google’s expensive, big-city lawyers, against whom the Authors Guild’s cut-rate representation stood no chance.
If your view is that Google won the day because they could afford better attorneys, then you can probably tell us what the errors are in Judge Leval’s reasoning. That would make for a constructive contribution to this discussion.
My concerns are two, both relating to the ‘slippery slope’ of the decision. Obviously, Google spent many millions to orchestrate a step by step plan to achieve their goal, once they couldn’t get it done directly, ten years ago. Commenting on the judge’s decision only discusses where we are at a single point in time.
Be aware that this is a fine decision for those of YOUR profession… in fact I might be correct in assuming you would prefer copyright be eliminated altogether as it is an interference to your work.
Ok, now that we have that out of the way…
The two real issues here are: What happens when Google is hacked, and it will be, and 20 or 30 or 50 million books are strewn across the Internet much as Edward Snowden’s work was? Ooops? Is simply not an acceptable response from Google. The other issue, not addressed in the decision is some prohibition on Google from being able to monetize this process in any way shape or form. As in, we will grant this permission providing… but that isn’t within the purview of this type of decision. As I said above, this is a step by step process for Google as their internal information, on a pay basis, suggests once done, this “service” will be worth One Billion Dollars to Google. But, again, there is no prohibition against Google eventually converting this to a pay service, is there? Just their ‘word’…which, as history shows dozens of times, is worth…well…
I’m done. I won’t be able to convince those who believe publishers aren’t entitled to make a prof** (explicative deleted) that this is a bad decision and simply greases the slippery slope for Google.
I am working hard to beg publishers to ‘opt out’, is all.
Be aware that this is a fine decision for those of YOUR profession… in fact I might be correct in assuming you would prefer copyright be eliminated altogether as it is an interference to your work.
No, you would be incorrect in assuming that. Are there librarians who are on the extreme fringes of the “copyleft” movement and would like to see copyright (as traditionally understood and defined in US law) abolished? Sure. Do they represent more than a tiny minority of the profession? No. It’s a great oversimplification to say that copyright constitutes “an interference” in librarianship; our relationship with copyright is much more complex and nuanced than that.
What happens when Google is hacked, and it will be, and 20 or 30 or 50 million books are strewn across the Internet much as Edward Snowden’s work was?
Why are you so confident that it will be? Given the likely cost of hacking through the industrial-strength protections that Google has put in place around its confidential and proprietary information, and the relatively low market value of most of what is in the books database, it seems to me like the probability of a hack along the lines you describe is relatively low. But let’s say it does happen: “What will Google do?” is a reasonable question. Obviously, it would depend on whether the hack was facilitated by negligence on Google’s part. Negligence would have to be determined in court, of course. And contrary to your assertions above, Google’s opponents in the publishing world have demonstrated a robust ability to bring Google to court.
The other issue, not addressed in the decision is some prohibition on Google from being able to monetize this process in any way shape or form… but that isn’t within the purview of this type of decision.
It’s more than outside the purview of this decision — such a prohibition would, I believe, have no basis in law. There is such a thing as commercial fair use, as Judge Leval noted in his decision.
I am working hard to beg publishers to ‘opt out’, is all.
If I recall correctly, opting out was something publishers would have been able to do under the proposed settlement. However, the settlement was set aside pending a determination on the fair use question. That question seems now to have been settled. If Google’s use of the copyrighted materials is fair, then “opting out” becomes a non-issue — copyright holders don’t get to decide whether or not others will make fair use of their work.
