I had thought that my two most recent posts reflecting on the future of research libraries would end the year for this blog. But I find two issues have arisen that I want to comment on. Since they both involve copyright — one is merely my observations and the other involves reporting on a recent court decision — I consider this post a year-end copyright roundup. Yee Haw!
I have commented before about the campaign that Elsevier has undertaken to send take down notices regarding PDFs of articles they have published that are posted on Academia.edu. At that time, I expressed the hope that the American Association of University Professors would get involved. The AAUP has been very diligent in defending the copyright interests of faculty, correctly perceiving that rights ownership can be an issue of academic freedom. Unfortunately, up till now, at least, this concern of the AAUP has had a huge blind spot — the threat to academic freedom that is posed when copyright is transferred to large commercial interests like Elsevier, usually gratis and on terms dictated by the assignee rather than the author, which do not share the values regarding scholarship that are held by most faculty authors and by the AAUP itself.
Now we have begun to see that this take down campaign from Elsevier has expanded, as we hear that notices to remove posted PDFs are also being sent in large numbers to universities, presumably directed at copies of articles that are posted on the individual profile or research sites of the faculty authors. These authors clearly believe that they have the right to post these articles — probably because they correctly consider the distinction between their final submitted manuscript and the published PDF to be meaningless — and Elsevier and Elsevier’s insistence that those articles be removed is a direct attack on research sharing, the advance of scholarship and, therefore, academic freedom. Why does the AAUP remain silent? In the past they have implied that they did not want to interfere with authors’ decisions about where to publish. But now we have clear evidence that many authors do not understand the rights situation and their ability to continue their research projects when they assign copyright to Elsevier. On individual campuses we work to educate them, but we desperately need the voice, and the gravitas, of the AAUP to step in and help faculty authors understand the consequences of these publication decisions. Thoughtless assignment of copyright is a grave threat to academic freedom, as Elseivier is working hard to demonstrate, and it is an issue that the AAUP cannot ignore without seeming to condone.
As for campus responses to these take down notices, I want to make one point. These are not ordinary DMCA take down notices; the DMCA, and section 512 of the copyright act that it amended, pretty clearly imagine that take down notices will be sent from the rights holder to remove material posted by some third-party user of an Internet service. They do not contemplate a situation where the actual ownership of the rights might be disputable. But that is the case here, where the “user” who posts the PDF is the original rights holder, the author. Elsevier is sending these notices as a putative assignee. But we know from long experience that publishers are not good at actually completing copyright transfers, faced with publication deadlines and author disregard of the final paperwork. So in this unique situation, we cannot be sure that Elsevier actually does hold the rights on the basis of which they are demanding takedown, and that the authors who posted the work are therefore not entitled to do so. In this situation, simple compliance with the law demands that universities, through their DMCA agents, insist on receiving evidence of a completed copyright assignment — the law requires a written and signed instrument — before they comply with Elsevier’s take down notices.
The case I want to comment on also involves the relationship between universities and authors who hold copyright, in this case the authors of dissertations. In Andrew Diversey v. The University of New Mexico, which was before the Tenth Circuit Court of Appeals and decided on December 23, we had a direct conflict between a dissertation author and his university. Briefly, Diversey alleged that he had a terrible time with his dissertation committee, finally sent a draft of his thesis to the Dissertation Coordinator for proofreading, and had it “confiscated” by the Graduate Studies Dean. Subsequently, Diversey discovered that copies of his draft had been sent to ProQuest and to two UNM libraries. The story strikes me as bizarre, especially since there is no indication that Diversey received a degree or that his dissertation won formal approval. So it is odd that it was apparently treated as a finished product; there may be more to the story that we do not know. But ultimately, Diversey sued the University of New Mexico for copyright infringement for copying and distributing his dissertation without authorization. His claim was dismissed by the trial court because that court said it was filed after the statute of limitations had run out. But the Court of Appeals reinstated part of the claim, for unauthorized distribution, and sent the case back to the District Court.
The crux of the case is whether the University infringed Diversey’s by making copies and putting them in their libraries. The Court of Appeals finds that it is at least plausible that it did. The Court recognizes Diversey’s copyright in his dissertation, of course, and suggests that any copying or distribution without his permission, even by the University at which he was a student, is infringing. Copies were made, according to the Court, but Diversey had notice of those copies far enough in the past that the statute of limitations had run out on his claim for unauthorized reproduction by the time it was filed. But the Court found that Diversey only became aware of the distribution, by which it means placing copies of the dissertation in the libraries and listing them in a catalog, within the three-year limitations period, so his claim on that alleged infringement should go forward. The Court also found that a fair use defense would not protect UNM in this set of circumstances.
To me this case is a warning to universities and schools that want to retrospectively digitize Doctoral or Masters’ theses and dissertations, and decide to skip the step of seeking permission from the authors. It is perfectly true that libraries can distribute “lawfully made” copies under the doctrine of first sale. But if the copies themselves were unauthorized, so too is the distribution of those copies. In other words, if the only copy that UNM placed in its library was the one Diversey originally submitted, they could have a first-sale defense. But the University clearly made more copies, since they sent one to ProQuest and had at least two copies in their libraries. It is the distribution of these unauthorized copies that the Court of Appeals thinks is possibly infringing. And that situation is very similar, it seems to me, to the making of digital copies of older dissertations and placing them in an institutional repository without permission from the author. I understand that some institutions decide that it is easier to ask forgiveness in this instance instead of permission. But to me that attitude is disrespectful of the authors of these works. And now we have a case that reminds us that that approach can also be expensive.
If an author discovers her dissertation in an institutional repository and is very unhappy about it, she will have three years from the time she either knew or should have known about the unauthorized copy to bring suit. The fact that the institution might remove the copy from distribution once it learns of the objection will not protect them from such a lawsuit if the author is determined. The expense of defending that lawsuit would be very high regardless of the outcome, and even higher if the institution loses or has to settle (which is now very likely for UNM in this case). For these reasons, this case reinforces my long-held belief that systematic digitization of older dissertations or theses is a situation where permission is the better course, since forgiveness might be doubtful and, without it, the cost could be much too high.
Kevin,
Might your expressed concerns about retro digitization of legacy dissertations without permission omit the possibility that many of these works are in the PD due to failure to comply with formalities of Copyright Statute? Substantive research on this matter was published in a referred journal (see link below) and has prompted University Counsels at research universities such as Univ Florida, Georgia Tech and others to proceed in retro dig projects without author permission.
http://muse.jhu.edu/journals/pla/summary/v011/11.3.clement.html
Gail is correct, of course, that my warning is directed only to materials within a retrospective digitization project that continue to be protected by copyright. No such worry would exist for materials in the public domain, and I probably should have said that explicitly. But I have to add that I am less confident than the authors of the article Gail cites about relying on the idea that all pre-1978 dissertations were published without notice and therefore not subject to copyright (under the 1909 copyright law). Even though there may have been a strong consensus with the library and academic communities that dissertations were published, I worry that courts might not make that determination when confronted in a specific case with a plaintiff who would lose his or her right to be heard by the court if the court ruled that the work had been so published. Courts have been all over the map in their definitions of publication, but seem especially reluctant to hold that a work has been published when doing so would divest an otherwise qualified author of his or her copyright.
Kevin,
A few clarifications are in order as concerns your comments on my lawsuit against the administrators of the University of New Mexico (UNM). The confiscated copy of my draft dissertation was in fact a temporary loan copy for the sole purpose of correction and was thus never “provided”, as the 10th Circuit loosely qualified it, but only lent to UNM with the promise that it be returned within two weeks. Because there must be, as you well know, lawful ownership, and not mere possession, of a copyrighted work before the first sale doctrine can apply, UNM was not eligible for section 109 or 108 privileges under the Copyright Act. Nor did they qualify, as you correctly stated, for section 107 fair use since the dissertation was an unpublished work and since UNM copied and distributed it in its entirety. Keep in mind, however, that if UNM had lawfully acquired a copy of my dissertation, they would have had the right pursuant to section 108(b), which pertains to unpublished works, to make preservation copies of it and to distribute it to library patrons for scholarly research. UNM initially made the untenable claim in district court that they had 108(b) privileges without ownership of a copy of my dissertation on the grounds that the 108(b)(1) requirement that the reproduced copy of the work in question be “currently in the collections of the library or archives” could be wholly satisfied by an unlawful transfer. This ridiculous argument was shut down permanently by my citing to Association of American Medical Colleges v. Carey, 728 F. Supp. 873, 889 (ND NY 1990). Finally, I should add a few extra facts not mentioned in the 10th Circuit opinion that will help to make sense of this admittedly bizarre case. Having neglected to perform their required duties, UNM faculty and administrators fraudulently approved my dissertation without having reviewed it first, which is a flagrant violation of UNM catalog policy as well as of national and international academic norms and standards for doctoral programs. Once this was revealed by way of my official protests, the UNM Dean of Graduate Studies confiscated the loan copy of my draft dissertation and proceeded to copy and publish it against my explicit will in an attempt to cover up faculty and administrator fraud (including his own) and to force upon me an invalid Ph.D. degree, which I refused to accept.
I wonder how much this really applies to the dissertations and theses with which libraries will be working. Perhaps there is more information in the record that I haven’t seen, but the facts of the case as presented in the opinion are so strange as to seem virtually useless in applying the case to other circumstances: a draft dissertation which was never approved (and the degree, presumably, not granted), the copy having been confiscated for reasons which the opinion does not discuss, and then is for some reason deposited in the library, and even sent to ProQuest, despite never having been approved or defended. Further, the only infringement for which the Court found liability was the distribution; had it been a legally-acquired copy, this would have been a text book application of Sec. 109(a).
In addition, though I think under the circumstances they came to the right conclusion, I think the court’s fair use analysis is highly problematic. They gave almost no attention to the third factor, treating it as a simpler question than it really is; had they bothered to site any existing law, they might have found it relevant that other circuits having found room for fair use of the full work. Their discussion of the fourth factor hinged heavily on the fact that Diversey was never able to complete the process properly and receive his degree (and can’t do so elsewhere). And though I think the discussion of the second factor was essentially correct, the circumstances of the case—the only copy having been essentially stolen from him and then routed through a process which is only supposed to happen when a thesis has been successfully defended—are not facts which I expect will be germane to most cases.
Given that the facts are so strange and the liability normally exempted under the First Sale doctrine, I’m not seeing how this is setting much of a precedent. Is there something I’m missing, here?
I don’t think you have the DMCA details quite right. 512(1)(c)(1)c [I likely have the citation screwed up] talks about “upon notification of claimed infringement as described in paragraph (3), responds expeditiously to remove, or disable access to, the material that is claimed to be infringing . . . ” And 512(1)(c)3(v) only requires a statement that “the complaining party has a good faith belief that the use of the material in the manner complained of is not authorized by the copyright owner . . .”
So there is no requirement the complaint cite real infringement of anyone’s copyright. The ISP is required to take the cited work down in order to avoid liability for secondary infringement. It is only after taking it down that we could send a counter notice and put the paper back up after 10 days.
As we follow the never ending Lenz case, this lack of a requirement that the complaint reflect the real world becomes clearer even though the notice is also supposed to include a statement that the information included be accurate. Apparently you only have to claim accuracy, not be accurate. Oh, and you have good faith.