Source: Istanbul University
Source: Istanbul University

A couple of years ago, a brouhaha erupted in the wake of a public statement from the American Historical Association regarding the online dissemination of student dissertations in the field of history. Whereas student authors across disciplines are often allowed to embargo their dissertations for up to three years, the AHA proposed that history students be allowed to impose embargoes of up to six years on their work while they try to get it formally published — provided that they place printed copies of their dissertations in their granting institutions’ libraries and make those copies available to researchers immediately.

The firestorm of reaction to this proposal was truly remarkable: it included public name-calling, ridicule, and general vituperation of a kind that doesn’t often occur when the topic of discussion is graduate-school policies. The public criticism also prominently featured radical misrepresentations of the AHA statement, strongly suggesting that many of its critics had not bothered to read the statement before issuing their responses. I discussed some of these at the time in a posting here in the Kitchen and also in a lecture I gave at the Smithsonian Libraries.

Over and throughout this public conversation (or shouting match) lurked a larger issue, one that was never really addressed in a direct way: to what degree is it appropriate for graduate schools to require students to give up control over the dissemination of their theses and dissertations?

I would like to suggest that this is a difficult and complex question, and that the right answer to it is not obvious.

For most of the past 200 years or so, this hasn’t been a particularly fraught issue. Pretty much every graduate student understood that, upon completion and acceptance, her thesis or dissertation would be printed and bound in several copies, and that one of those copies would be made publicly available in the library stacks. Except in unusual circumstances regarding sensitive intellectual property (such as papers based on patentable products or processes), students generally didn’t have any desire to restrict access to the printed copies of their papers. Most students would likely have welcomed a broad readership, and in cases where the dissertation’s content might be sensitive in some way and wholesale dissemination or redistribution might be problematic, the organic unwieldiness of the print format was usually sufficient to prevent it. And, of course, until recently there was no expectation that students would abdicate the traditional copyright prerogatives in their theses or dissertations; they retained both their copyrights and the legal control over their work that copyright law provides.

In the current information environment, obviously, things are very different. Dissertations that would, in the past, have languished in obscurity on local library shelves (and, more recently, been available digitally only behind a Digital Dissertations paywall) are now routinely made publicly and freely available online, in institutional repositories.

I don’t know anyone who objects to this in principle. I think most of us agree that the practice of making theses and dissertations freely available online provides significant benefits to the student author, to the degree-granting institution, and to scholars generally. However, as we all know, making good policy decisions isn’t just about identifying benefits — life is more complicated than that, and even the best policies and most well-intentioned practices are likely to raise difficult questions, especially when those policies and practices involve the disposition of intellectual property within a system explicitly designed to provide unfettered worldwide access and unrestricted copying.

In the case of online theses and dissertations, difficult questions that arise include these three:

  • To what degree should or shouldn’t the scholarly products of a graduate school education be treated like conventional intellectual property, with both copyright and control remaining in the hands of the author?
  • In situations where the institution makes a claim of control on dissertations and theses (whether by claiming copyright or, more often, by requiring students to grant either the institution or the public a license of some kind) are matriculating students being clearly informed that such claims will be made on their subsequent work?
  • Should graduate students be allowed to opt out of granting licenses to their institutions or to the public, or be able to set the terms of those licenses within fairly broad parameters? (And if not, why not?)

These are controversial questions, and when raised, they tend to arouse very passionate responses. (Again, consider the furious reaction to the AHA statement — and consider the fact that the only change in current practice called for by that the statement was to give students the option of extending online embargoes from three to six years if they so choose.)

Why all the passion and fury? I think it’s because these questions go to very fundamental issues about values. These aren’t questions about the most effective way to make dissertations available, or about which repository platforms work best. They are not, in other words, implementation or efficiency or effectiveness questions. They are not is questions, but should questions: moral and ethical questions that involve the balancing of individual rights and collective responsibility. Assuming we all agree that scholarly authors have both a set of individual rights and a set of collective responsibilities with regard to their work, and given the reality that there might sometimes be real tension between what an author would prefer and what would best serve the collective good, how should that tension be resolved when it comes to theses and dissertations?

One possible way to resolve the tension is by compromise, in this case between individual and group interests. In order to do so, we would first have to figure out what fundamental principles should guide such a compromise, and then what the compromise would look like.

There’s another way to respond to this situation, of course, and that is to deny that the rights and obligations in question are really in conflict or tension at all. From this perspective, the author’s interests and the institution’s interests are actually in perfect harmony, and if they appear not to be — if, for example, a student author believes she would benefit from embargoing her dissertation for five years and the institution wants her to limit the embargo to one year — the appearance of conflict is an illusion, one that can be dispelled by education.

One challenge with this approach is that education goes both ways. Authors don’t have a perfect understanding of all the downstream impacts that will result from decisions they make about the disposition of their work, and institutions have a similarly imperfect understanding of all the downstream impacts produced by requirements that they impose on theit authors.

Another challenge is the difficulty of finding “educators” who aren’t advocates for one side or the other of the question. There’s nothing wrong with advocacy, of course, but you can’t expect advocates to be impartial providers of education — that’s not their job. Their job is proselytizing, which is not the same thing.

Another problem is that education can only go so far when the problem is genuine disagreement rather than ignorance or illusion. An author may simply disagree that her rights as a content creator ought to take a back seat to the institution’s interest in disseminating her work under the institution’s preferred terms. Not all disagreements are a function of a lack of mutual understanding, or of one party’s misperception of reality; sometimes disagreements arise because the two parties genuinely want mutually exclusive things.

I think the time has come for a more open and candid conversation on our campuses about the questions listed above and about the underlying issues that inform them — conversations that include all stakeholders and explicitly recognize their legitimate interests, their rights, and their responsibilities. Achieving a consensus agreement as to what those rights and responsibilities are may not be easy, but it seems to me like an absolutely essential first step towards establishing answers to these questions and outlining a feasible and equitable way forward.

Rick Anderson

Rick Anderson

Rick Anderson is University Librarian at Brigham Young University. He has worked previously as a bibliographer for YBP, Inc., as Head Acquisitions Librarian for the University of North Carolina, Greensboro, as Director of Resource Acquisition at the University of Nevada, Reno, and as Associate Dean for Collections & Scholarly Communication at the University of Utah.


73 Thoughts on "The Dissertation Mess: Balancing Rights and Responsibilities"

Rick, thanks for your good piece. My institution, Georgetown University, has wrestled with this issue and raised some distinctions around embargoes that I think are relevant for this discussion. While the university’s policy is not yet formalized, the policy committee has recognized three levels of embargoes: a restrictive embargo for a limited period involving a pending patent application; a common embargo, for up to two years, during which the text would be withheld from public distribution but would be available to anyone with a Georgetown NetID; and an extension of an embargo beyond two years, a case-by-case scenario involving discussions between the student, the dean of the grad school, and the student’s mentor. So there is some flexibility there, which is appropriate given the different needs of different disciplines.

Great discussion and your comments are insightful.

However, how the embargo effects not only the author but also the publisher. Why would I want to publish something that is readily available for free? My hesitancy to publish means that the author cannot get the dissertation, though revised, published and this means that the author could be denied tenure.

Would the solution be to just go back to the paper system? In other words, publish one copy for the university library. Should the author want wider distribution simply notify the institution that they are free to do with the dissertation what they want, in other words transfer the copyright to the institution,

Harvey has pointed to the key issue here. The only reason that immediate open access is not universally applauded is that the current tenure system is based (at least in most areas of the humanities and social sciences) on rewarding the publication of a first book, which almost always has to be a revision of the author’s dissertation, given the time limit imposed (tenure is usually decided at the end of six years, which no doubt accounts for why that number was chosen by the AHA). The real argument, then, revolves around whether or not publishers make decisions about publishing revised dissertations depending on whether or not the original dissertations are openly available or not. That is an empirical question that can only be decided by gathering information about how publishers make their decisions. While I would not go so far as to say that every publisher would reach a negative decision about publishing a revised dissertation if the original is available open access, I would bet that it factors into every publisher’s decisionmaking process and, at least for some, is a decisive consideration leading to rejection.

Harvey has certainly pointed to one important issue here, but it would be a mistake to assume it’s the only one, or even “the” key one. A more fundamental issue is simply the question of whether and to what degree authors of dissertations ought to be allowed to retain for themselves the prerogatives given them by copyright law–specifically, the right to say how and whether their work will be distributed and reused (beyond fair use). Questions about publisher practice are meaningful, but it seems to me that questions about authors’ rights are prior to them and have implications that go beyond the marketability of their books.

What reasons would authors of dissertations have for embargoing their dissertations if there were no issues involving their professional careers at stake? Don’t most people like to be read? I have a hard time believing that grad students will want to assert their rights under copyright law when so few of them know anything about copyright law to begin with.

I have a hard time believing that grad students will want to assert their rights under copyright law when so few of them know anything about copyright law to begin with.

First of all, the same could be said about an awful lot of faculty members, in my direct experience.

Second of all, it really doesn’t matter whether you understand why a grad student would want to assert his rights under copyright law — nor, ultimately, does it matter whether grad students do or don’t fully understand copyright law. The author’s rights are the author’s rights regardless of whether he or she understands them fully. The question is whether and to what degree it’s appropriate for those rights to be taken from authors by those with power over their careers, such as funders or universities.

The same question could be asked of publishers who take the rights of authors as a condition of publishing their articles. Publishers likewise have “power over their careers,” do they not? Do you consider it “appropriate” for publishers to exercise their powers to compel rights to be relinquished as a condition of publication? If you do, then explain how this differs from universities exercising their power to grant degrees conditional upon the recipients relinquishing some of their rights.?

That’s a good question. A couple of answers:

1. Publishers that take an author’s copyright do so as part of a quid pro quo arrangement: the author trades her copyright for publishing services, which are otherwise provided to her at no charge. The university, however, isn’t giving its educational services to the graduate student in exchange for intellectual property rights. It’s giving them to the student in exchange for tuition, or labor, or both. For the institution to require its students both to pay tuition and to give up the effective rights in their original work strikes me as a meaningfully different arrangement from what traditionally exists between authors and publishers.

2. There’s also a big difference between assigning copyright to a publisher and assigning a license to the general public, and that difference has significant real-world implications for my rights as an author. If, for example, I publish an article with Elsevier, I’ll be required to give Elsevier the copyright in my work, which means that Elsevier (but no one else) can exercise copyright prerogatives in my work without asking my permission. If, however, I publish my article with PLOS, I have to do so under a CC BY license, which means that everyone in the world gets to exercise all copyright prerogatives in my work without asking my permission. And if I produce a dissertation at Duke, I’ll be required to make it available to the public under a CC BY NC ND license, which allows everyone in the world to redistribute and republish my work in any noncommercial venue they wish. In other words, if I publish with Elsevier I’ve given control of my work to one entity; if I publish under a CC license, I’ve given control of my work to everyone in the world. There are both costs and benefits in each scenario. One of the costs of the latter scenario is that I no longer have any effective control over how, where, and in what venues my work is used. Some believe that this is exactly as it should be and that authors who disagree are just ignorant, selfish, and/or irrationally fearful. I’m not so sure that’s obviously the case.

To be clear, I’m not asserting that it isn’t okay for institutions to require copyright assignment or a public license. What I’m saying is that this issue is complex and needs more (and more public) discussion. And I do feel very strongly that if an institution or graduate program is going to require a public license or the assignment of copyright, it should say so very clearly and explicitly to its graduate students before they matriculate so that the students make their choice in the full knowledge of its implications for their intellectual property rights.

Another analogy, perhaps even more direct, is between the requirement placed on grad students and the mandate for faculty to make their articles available via some “open access” mechanism, whether federal like PubMed or an institutional repository. Most librarians have strongly favored these mandates, indeed have been the primary advocates of implementing OA in this way. Do not those mandates similarly impinge on the rights of the authors involved? (They generally do have an “opt out” option, but so do grad students.)

If there really is an effective opt-out clause in the policy, then the program is voluntary and isn’t impinging on anyone’s rights.

That said: as far as I’m aware, unconditional opt-outs are a nearly universal feature of what are often misrepresented as “OA mandates” for faculty, especially at U.S. institutions. However, my impression (admittedly not supported by systematic research) is that opt-outs are not usually available to graduate students with regard to their dissertations.

I agree that faculty members don’t appear to understand it either.
This morning I looked through my university’s policies on open access for research higher degree theses. The policy summary is here

There only options are “open access” and “embargo”. There is no mention, or choice, of different OA licenses. Presumably this means it’s published under the least restrictive, CC-BY?

I also had a brief chat with my supervisors – who immediately said “just make it open access”. Again, there was no mention of types of license – or any idea that their are possible downsides.

If you ask 100 different academics for a definition of “open access” you will likely get 100 different definitions. I work with a lot of groups and some very small percentage understands the full BOAI definition, most just think it means there’s a free copy of it available, a significant portion think it means PLOS ONE-style peer review standards, and very few have any clue whatsoever about licensing and reuse requirements.

That’s the impression I get too. I was surprised though to find the university’s guidelines were so vague.

Sandy I am surprised you make this argument. It is the one Madoff relied upon to sell stock!

I’m aware of graduates objecting – strongly – to OA dissertations being offered for sale by a third party. ProQuest used to have an agreement with Amazon to do this. Many graduates were shocked to find their dissertation for sale, even though they had legally agreed to it in their contract with ProQuest. (On the plus side, they were entitled to royalties from any sales.) There are a few anecdotes floating around blogs and chat forums of grads finding their thesis being sold by an unknown publisher, but I don’t know how common this is. The complaint is usually along the lines of “it was used without my knowledge or consent”, pointing to a serious misunderstanding about what OA licenses mean.

A newer issue I came across yesterday was that Turnitin [] are using OA dissertations their commercial plagiarism detection software. Courts in the US have apparently ruled this is “fair use”. Graduates obviously get no royalties for it. The main complaint seems to be that the graduate didn’t “authorize” use of their dissertation in this way… (even though they probably did through whatever OA license they selected).

The Fourth Circuit Court of Appeals in A.V. vs iParadigms (the company that owns Turnitin) in its 1988 decision found Turnitin’s use of student papers to be “transformative” fair use–the same ground on which the recent Google case was decided in favor of Google in the Second Circuit Court of Appeals. Given this decision, it is highly unlikely that any graduate students could win a case based on Turnitin’s use of their dissertation.

Thanks, Sandy. That was my understanding of the situation. However I think my points still stand. First, most grad students I’ve spoken to (and faculty for that matter) seem completely unaware that their OA dissertation is (or could be) used in this way.
Second, most seem to believe that anyone using their OA dissertation (especially for commercial gain) have some obligation – be it legal, ethical, or moral – to inform them or ask permission. They seem truly shocked that there is no such obligation, especially so if the reuse is for commercial gain.

The embargo may have nothing to do with copyright or professional careers. The dissertation / thesis *document* is supposed to be a full record of the research conducted. It may contain material that cannot or should not be made public. Inclusion of patent-able material is one example already mentioned. Another could be culturally sensitive material. I have colleagues whose theses are embargoed as they contain photos of deceased persons of Aboriginal descent. My own undergraduate engineering thesis contained material that was commercial-in-confidence. It was marked and then kept in a locked cabinet in my supervisor’s office for a while.

My understanding is the dissertation (or thesis) is the document containing a full record of the research undertaken by a candidate. It may therefore contain far more material than would normally be made public (or published) to facilitate the examination process.

Rick and others have already mentioned patent-able material. An embargo could also be needed if the dissertation includes material that is classified, legally suppressed, commercial-in-confidence, or culturally sensitive. Two examples from my area of health would be photos of a deceased person of Aboriginal descent, or personal information in child protection cases that could not be adequately rendered non-identifiable.

On the other hand, publishing a revised dissertation that is openly available may give you an advantage as the author has already built a market for her work.

DRL the argument that publishing something twice is providing a built in market is specious. If I have read it once for free why would I read it again for a price? We are talking about markets that basically number in the hundreds if that.

If it’s OA, *anyone* can republish it, without the author’s knowledge or consent. The dissertation could be published by multiple third parties, which could actually reduce the potential market for the work.

A thoughtful and useful discussion (and timely for me: I just finished drafting the “dissertations” section of an Economics of Access essay, and will add this to it) in a difficult area. (My metadiscussion does not have clear heroes and villains.) I would pick two nits, one with you, one with Sandy.

One nit: when it comes to “intellectual property” (a terrible term) for people at universities, it’s not that simple, as I think you know: many (most?) universities require that another form of “intellectual property,” patents, be at least shared with the institution if clearly associated with the person’s work. Presumably, faculty sign employment agreements to that effect; do grad students (who are frequently also employees as TAs and the like)?

The other nit: Sandy, I *know* you participated in the previous discussion, and I *know* you’re aware of the C&RL survey that says your bet is wrong, at least according to explicit statements from some university presses. (The first half of the bet that is.) Here’s the link in case you’ve forgotten:

One nit: when it comes to “intellectual property” (a terrible term) for people at universities, it’s not that simple, as I think you know: many (most?) universities require that another form of “intellectual property,” patents, be at least shared with the institution if clearly associated with the person’s work.

Agreed, but this essay isn’t about patents, which entail a related but separate set of issues. My concern in this posting is with intellectual property of the more garden-variety, textual kind — the kind of which most dissertations mainly consist. When it comes to this kind of intellectual property, the underlying issue really is pretty simple: the law gives the author copyright in the original work. An institution of higher ed might, as a condition of matriculation or graduation, require student authors to license some or all of the rights associated with copyright to the institution or to the public. Such polices do exist, though they’re not universal, and they raise difficult and complex questions. The appropriateness of such policies is my concern here.

Sounds like a good candidate for an issue tree analysis. I wonder how big the tree is?

Walt: I just read the article cited and the discussion on page 376 seems to back the argument of publishers being hesitant to publish indexed dissertations because libraries do not want to duplicate that which is in their holdings.

David: It seems hypocrisy is an institutional trait. As La Rochefoucauld said: Hypocrisy is the homage vice pays to virtue.

Institutions that are more liberal with IP and patent ownership have a competitive advantage in recruiting top faculty over those with stricter policies (and I know of at least one that uses this to great advantage).

The other end of the spectrum would be to treat researchers as employees and everything they do at a university as “work for hire”. Similar angles have been suggested for research grants, turning the grants into contracts and the researcher into an employee of the granting agency, with all outputs owned and controlled by the funder. This would then solve all the stick compliance issues around access to papers and making data publicly available. Unsurprisingly, this sort of approach is not popular among researchers.

Making a researcher an employee of a granting agency would be a HR nightmare. Depending on the field, it would put the entire research workforce into short-term, contracted positions, forcing researchers to change employers every few months or years. The impact on basic entitlements – annual leave, sick leave, health insurance, etc – would surely be absolutely catastrophic and unworkable.

As to the outputs, I don’t see how it would solve the problems of access to papers. As David Wojick and I often point out, the processes for scholarly output (e.g. papers) rumble along well after the funding has ended. For a tenured professor, this may not be a huge issue because they are still drawing a salary and presumably fit it around other work. A research on a short term contract may not even be employed in the same organisation. If they are, their new role may not allow time to work on publications unrelated to the current funding.

Oh, I agree that it’s an undesirable way to go, but given the increasing ownership that universities and funders are placing on research outputs, we seem to be creeping closer and closer toward it. Besides, in the age of Uber, why bother with HR? All researchers will be independent contractors who rent space from universities, responsible for their own benefits. Problem solved (albeit in perhaps the worst way imaginable).

Do you have a sense for how many countries have laws similar to the USA regarding researchers and IP? I was under the impression that the USA was something of an anomaly, and many researchers internationally were already in “work for hire” arrangements.

It’s a good question. We have the Bayh-Dole Act in the US, which is considered one of the most effective pieces of legislation of the last 35 or so years, and responsible for enormous levels of economic growth. I’m not sure of similar legislation anywhere else, but every major funding agency that I’ve checked (RCUK, Wellcome, HHMI, etc.) has a similar policy regarding IP. I suspect it is something of a competitive advantage in recruiting/retaining smart people, as well as an effective incentive system for driving success.

Ah, I think I misunderstood. I thought the Act specifically allowed an individual to keep the IP, whereas it really is just means that IP for federally -funded research isn’t assigned to the federal gov’t. An Institution (e.g. university) could still have its own “work for hire” arrangements with its research employees, right?

Right, it’s left up to the grantee and their institution. But any work for hire requirement would be a massive disincentive for job recruiting. Schools that are more liberal and supportive gain a big advantage for top researchers.

This points to a very big difference between the higher-ed systems in the US and those in the UK and Europe. In the UK, for example, it’s possible to impose a single policy that applies to all public universities, because there’s only one public higher-ed system. In the US, we have 50 such systems (51 if you count Cal and Cal State separately). This means that US and UK policy makers have to worry in very different terms about competition between institutions. If the UK decides to define academic work as work for hire, it will affect all public universities in the country similarly, but if Kansas decides to do the same thing it runs the risk of a brain drain to Nebraska.

Actually David C., in the US Federal case at least, grants are contracts. They even look a bit like work for hire because the proposal, which is incorporated into the contract by reference, includes a detailed breakdown of hours to be worked by everyone funded via the grant. But a work for hire contract normally does not make one an employee. And in any case who owns what is a function of the terms and conditions of the contract.

In the US the IP derived from federal grants is designated by law to belong to the grantee and their institution, regardless.

On the contrary, the Feds get a Federal use (copyright) license to any “works developed” under a grant. This license is the basis for the US Public Access program, whereby grantees are required to deliver (surrender?) a copy of their accepted manuscript reporting the research, even if the writing of the article itself was not paid for from the grant. The Feds can then do whatever they like with that copy, including making it freely available without restriction.

No one seems to be exactly clear on what a “federal use” license means. It’s interesting that at least in my experience, the agencies are only trying to exert that over the author’s manuscript version of papers and not the version of record. Further, I suspect some uses would draw a legal challenge (and some current uses might as well). Also of interest is that the recently released Department of Transportation public access policy requires grantees to give the agency:
“… a comprehensive non-exclusive, paid-up, royalty-free copyright license to the DOT and the submission of any Publications to the NTL digital repository. The copyright license must include “all rights under copyright,” including, but not limited to:
 Right to copy;
 Right to distribute;
 Right to prepare derivative works;  Right to display; and
 Right to perform in public.”

Indeed, the DOT language is the topic of my Inside Public Access newsletter this week. To my knowledge it is the first attempt to define the terms of a Federal use license. Note that it includes “all rights under copyright.” I call this a non-exclusive copyright, not a use license. The government admits of no restrictions whatever.

Apropo of this post, if a dissertation involves Federal funding of the research then it too would seem to be a work developed under a grant, so the Feds could glom and publish it.

I suspect many journals may object to giving someone else the right to sell copies of the article. The DOT may have inadvertently set up an unfunded OA policy, where the only way one can comply would be to publish via CC BY.

Emma: I think the major Federally funded Unis are setting up compliance offices to track all of this. Note that it could also apply to the accepted manuscript of the resulting book or monograph. Nor is there any time limit.

What counts as Federally funded research for the purposes of the Public Access program certainly has yet to be defined, with many grey areas. It is a massive program.

I’m aware of the article, Walt, but it doesn’t answer the most import question, which is how the open availability of dissertations affects the final decisions publishers make. They may be willing to consider revised dissertations, but they will surely take into account in their final decisions whether or not the potential market might be affected by ready accessibility of the dissertation. As Alex Holzman, then director of Temple University Press said in a discussion back in October 2014 on this same topic on Liblicense: “my own press definitely considers revised dissertations, but the
final decision to publish or not can include the forecast sales. And while the decline in library sales for already-available electronic dissertations may seem marginal to YBP, university presses these days often have zero wiggle room. That’s part of a whole different problem for discussion another time.” Librarian Robin Champieux said in the same thread: “More data would go a long way to helping us disentangle the real and false tensions between ETDs and later publication.  This includes data on actual publication decisions – not just publisher attitudes about ETDs.  For example, are openly available thesis and dissertations being published at the same or different rates than their closed and embargoed counterparts?” So the issue was hardly put to rest by the article you cite.

“put to rest,” no, certainly not that, nor can I imagine any set of real-world studies that would put the issue to rest. Given that libraries these days also have “zero wiggle room,” I almost wonder whether more data would actually help. I think the survey provided strong evidence that open access to the dissertation wasn’t generally a deal-killer for an honestly revised/expanded/improved monograph, but “generally” and “deal-killer” are both waffle words.

I worked in both the public and private sectors and I would never publish that which was freely available as an ETD publication regardless of the amount of revision. Revision does not remove the central thesis of the argument, it only adds more detail as to why the position is taken.

In both worlds one is required to make money – a horrible word except for those who have to make it. In short, one presents a P&L statement with each proposal to publish. It has now come to the point that since one cannot make money on academic monographs that OA supported books are being experimented with. If those experiments work the only thing that has changed is that someone else pays instead of the consumer who pays in an indirect fashion through taxes that are passed on to the university in the form of the state budget.

Alas there is no free lunch.

It might be an issue if someone took the OA dissertation and republished it in some way, before the original author had a chance to do it.

In Australia, a graduate student owns the IP for research they do for their degree / award (unless there is another agreement with e.g. external funder). The university will generally own the IP for research by staff. Any IP generated as a TA-equivalent (some employed on a casual or sessional basis), and therefore not their degree / award, would generally be owned by the university (unless there is another agreement in place).

I think what this post gets to is the fundamental balance that must be struck between the needs of the individual researcher and the common good. If we don’t expect researchers to live as ascetic monks, we must provide ways for them to earn a good living, own a house, feed a family, etc., otherwise all the smart people will leave academia for more reasonable careers. At the same time, the academic research community has a common goal of improving the world’s knowledge, and through that, improving the world. Care must be taken when asking one to sacrifice for the other. These concerns run through so much of the academic world, from publishing to patenting and so much more.

Here you have researchers looking out for their own personal best interests. Yes, it would be nice to immediately share that information with the world, but does doing so put one’s career (and future research that could be shared with the world) at risk? Does a popular published book on the subject do a better job of disseminating that information than an unembargoed dissertation and hence serve both missions better? One factor to perhaps consider is the urgency of the material–if the dissertation could lead to a cure for malaria, is that of higher urgency for release than a dissertation on Plato’s writings, which presumably isn’t going to save lives.

Similar concerns come up in questions of data availability. Science might progress faster if all were completely open with their experiments and data as they are performed, but it is in a researcher’s best interests to be secretive with that data until it is fully understood and ready for publication.

It’s also perhaps hypocritical for a university to take this “greater good” stance while still patenting research results and keeping them from the public. Any school that requires immediate/early release of dissertations or data should, to be consistent, dissolve their technology transfer department and release all patents to the public domain.

But this assumes there is a conflict between personal and communal good in choosing to provide broad access to dissertations prior to turning it into a book or chaptering it out. It is entirely possible, perhaps likely, that broad availability serves both interests. Your assumption that hiding my work, even for a limited time, serves my interest is certainly worth questioning.

So DRL, would you support a policy that allows authors to choose for themselves whether to make their dissertations immediately and publicly available?

Rick I think this is a possible solution. However, the author should be made aware of the advantages and pitfalls of providing immediate free distribution.

Sure, and also of the advantages and pitfalls of failing to do so. What we need is more dispassionate analysis in this space.

If we are to take policies that recognize student ownership of their own scholarship seriously, I think we need to leave it up to them. The problem is that we all are making decisions under uncertainty and we don’t know the impact of our decisions either way. Will opening my work up help build a reputation or will it limit publishing options? Am I deciding out of fear of what might happen or the likelihood of a particular result? Most often, I find, decisions to limit access are made out of fear rather then evidence.

Will opening my work up help build a reputation or will it limit publishing options?

Why assume that these are mutually exclusive possibilities? The reality may be that making your work freely available does some of both, or does neither. You’re right that we are all making decisions under uncertainty–and this goes for institutional policy makers as well as for authors. No one knows what all the downstream impacts of his or her decisions will be.

Am I deciding out of fear of what might happen or the likelihood of a particular result? Most often, I find, decisions to limit access are made out of fear rather then evidence.

Or at least, that’s how those pushing for more openness tend to characterize authors’ decisions to limit access. It’s a good way to sneak an ad hominem argument into the discussion: “You disagree with me not on the merits of my argument, but because you’re afraid.” But accusations of fearmongering can cut both ways: why are advocates of openness so afraid to let authors make their own decisions?

I’ve asked people who choose to embargo. So this claim is based on evidence.

Severely limited evidence, though. And you’re still not taking into account the fact that “fear” and rational disagreement can coexist. I might have genuine concerns about what will happen if I make my dissertation freely available to the world, and at the same time have principled reservations about a policy requiring me to do so. All too often, in my own direct (though also limited) experience, advocates want to focus on the “fear” instead of engaging with the substantive issues.

An author always has the option of asking the publishers he or she would most like to have a revised version published with whether its open availability would affect a decision about its publication at those publishing houses. The answers, if negative, would then provide the fear with empirical grounds.

I think context is important here too. We’re talking about graduate students being asked to make a decision on OA at one of the busier and more stressful times in the candidature, i.e. very close to or at thesis submission and/or viva. From the discussion here, it appears that they’re making a decision with little objective information or evidence.
At this point, grad students may have been living on very limited incomes for a while (some even close to or below the poverty line). They’re about to enter a very difficult and highly competitive job market, characterised by short term, unsecure employment. They may not be as receptive to arguments for “the common good” over their individual needs at this point in time.

It is possible that both needs can be met (see my example of a popular book above), but often it is not the case. That’s why nuance is needed in policy and self-determination made available to researchers. If the author’s needs are served by making everything freely available immediately, this is an available option.

Thanks for this post. You have touched on the key questions often raised by many while providing open access to theses and dissertations. I think hybrid approach of print and online access with embargo would best serve the interests of authors, institutions and the public.

What I feel is missing here is a decent evidence base from which to challenge assumptions or to confirm the reality for students. In our research for the Academic Book of the Future we have asked 12 major scholarly publishers whether the existence of a PhD online would have any effect on their decision to publish a book and the answer from all of them was no it would make no difference to the decision. This is because they are “book publishers not PhD publishers” – they want to see a good book proposal and they want to publish a good scholarly book. This bit of evidence contradicts what I hear from colleagues who state anecdotes that students are told by editors that if their PhD is freely available online then that might affect the chances of them getting published later especially if the book would be dependant upon a bit of the PhD that is a particularly juicy new discovery. Supervisors are regularly advising students to take up the embargo option.
I think we need to find out more to really come up with a proper opinion here – we have two sides the students and the publishers who seem to be saying different things. It may well be a nuanced difference that just magnifies the valid concerns of student supervisors rather than being a reality for the students themselves. But I only have suspicions not solid evidence as yet.

Whether or not making one’s dissertation freely available online has an impact on its marketability to publishers is a good and interesting question, and it’s inevitably where the conversation seems to go whenever this issue comes up. But it’s a secondary question. The more fundamental questions are the ones I bulleted in my posting.

Rick: There may be two reasons for the focus on what you call a secondary question. First the issue is far simpler than your bullet issues and blogs tend to favor simple issues. Your bullet issues are pretty hairy. Second, there may be an implication that if thesis copyright is of no practical value then your bullet issues may be moot. That is, if thesis publication does not affect book publication, how does the copyright issue even arise? In this sense the group may have zeroed in on the central issue.

The problem with treating practical marketing questions as the central issue is that requiring authors to give up their copyright prerogatives has implications that go way beyond the author’s ability to sell his or her manuscript to a publisher. A few examples off the top of my head: if you are required to make your dissertation available on a CC BY basis, people can reuse it, in its totality, in ways that you find objectionable or in support of agendas with which you disagree, without having to ask your permission. People can package your work in commercial products (or simply create commercial versions of your work and sell them) without your permission. People can translate your work without permission and distribute it as your work — regardless of whether you agree that the translation is good or even accurate. In these situations, copyright has significant practical value even if it doesn’t have direct market value.

These are issues that have little, if anything, to do with the author’s ability to sell his or her manuscript to a publisher. Of course, there are some who disagree that being concerned about these issues is reasonable. (Some even characterize it as “fearmongering.”) My inclination is to want to know what authors think, and to give deference to their views on whether their rights as authors should be honored.

And yes, the questions I identified are hairy, but I’m not sure what relevance that has to their importance relative to marketing concerns.

I agree that the type of CC license used is critical. The questions Rick raises here are ones that prompted me to argue for CC BY-NC-ND as the best license for OA monograph publishing. It is the license being used by the Amherst College Press, and it is the default license for Penn State’s institutional repository where I have over 80 of my articles posted OA.

Rick: As an issue analyst I was just looking for explanations for the behavior that you complained about. I am studying the logic of the discussion, not its content.

I offered two conjectures, one that marketability was the central issue. Your examples suggest that this may not be the case, although a lot of the discussion makes it sound like marketability may be the dominant factor in these issues, compared to things like CC-BY reuse.

The second possibility is that blog discussions tend to migrate to simple issues. Hence the relevance of the hairiness of your questions. This is not related to their importance relative to marketing concerns. It is related to the observed behavior of the participants here, that is which lines of reasoning they choose to follow. Where they go in the issue tree.

I think it’s a difficult thing to measure by survey. What people say versus what they do is likely to be different here. It’s kind of like the situation Rick has discussed in the past where pro-OA librarians won’t publicly state that they’d stop paying for materials that are freely available because they don’t want to hurt a cause they support. Here and editor saying they’re open to it may be very different from what they do in practice.

Perhaps the methodology would be to look at monograph publishing and see what percentage of what gets published (and where) comes from unembargoed dissertations versus that percentage that was embargoed. Or to compare the populations, X percent of embargoed dissertations see book publication versus Y percent of unembargoed dissertations.

But there would still be way too many confounding factors you’d need to control for.

Did your small survey ask whether the publishers would consider a manuscript based on a dissertation or whether it would enter into their decisionmaking about publishing the book? That is a crucial distinction, as Alex Holzman’s comment I quoted earlier underlines. Were the 12 publishers you asked commercial academic publishers or university presses? That also would make a difference.

My university forced me to submit my dissertation to ProQuest with a maximum two-year embargo as a condition of graduation. As soon as my degree was conferred I contacted ProQuest and had them put my dissertation on permanent embargo. They did so without any problem, but also notified my university’s graduate recorder, who sent a stern email reminding me that nonstandard embargoes required strong justification and approval from the Dean. I ignored the email and the permanent embargo has remained in place for several years now. Other than revoking my degree I can’t imagine what the university can do about it.

I believe the government is saying in its requirement that it can use the document internally. In short, there is no intention to make the document available for commercial purposes. They have the CRADA for that purpose.

But there does seem a difference between the “federal use license” claimed in general and the DOT’s new demand to be granted non-exclusive copyright to the work.

Harvey: This broad DOT language is from their Public Access Plan. Every accepted manuscript will be made public under these terms, if they hold. (It is just a plan, not a final policy.)

Comments are closed.