The discussion of the new Elsevier policies about sharing and open access has continued at a brisk pace, as anyone following the lists, blogs and Twitter feeds will know.  On one of the most active lists, Elsevier officials have been regular contributors, trying to calm fears and offering rationales, often specious, for their new policy. If one of the stated reasons for their change was to make the policy simpler, the evidence of all these many “clarifying” statements indicates that it is already a dismal failure.

As I read one of the most recent messages from Dr. Alicia Wise of Elsevier, one key aspect of the new policy documents finally sunk in for me, and when I fully realized what Elsevier was doing, and what they clearly thought would be a welcome concession to the academics who create the content from which they make billions, my jaw dropped in amazement.

It appears that Elsevier is making a distinction between an author’s personal website or blog and the repository at the institution where that author works. Authors are, I think, able to post final manuscripts to the former for public access, but posting to the latter must be restricted only to internal users for the duration of the newly-imposed embargo periods. In the four column chart that was included in their original announcement, this disparate treatment of repositories and other sites is illustrated in the “After Acceptance” column, where it says that “author manuscripts can be shared… [o]n personal websites or blogs,” but that sharing must be done “privately” on institutional repositories. I think I missed this at first because the chart is so difficult to understand; it must be read from left to right and understood as cumulative, since by themselves the columns are incomplete and confusing.  But, in their publicity campaign around these new rules, Elsevier is placing a lot of weight on this distinction.

In a way, I guess this situation is a little better than what I thought when I first saw the policy. But really, I think I must have missed the distinction at first because it was so improbable that Elsevier would really try to treat individual websites and IRs differently. Now that I fully understand that intention, it provides clear evidence of just how out of touch with the real conditions of academic work Elsevier has become.

Questions abound. Many scientists, for example, maintain lab websites, and their personal profiles are often subordinate to those sites. Articles are most often linked, in these situations, from the main lab website.  Is this a personal website? Given the distinction Elsevier makes, I think it must be, but it is indicative of the fact that the real world does not conform to Elsevier’s attempt to make a simple distinction between “the Internet we think is OK” and “the Internet we are still afraid of.”

By the way, since the new policy allows authors to replace pre-prints on ArXive and RePEC — those two are specifically mentioned — with final author manuscripts, it is even clearer to see that this new policy is a direct attack on repositories, as the Chronicle of Higher Education perceives in this article.  Elsevier seems to want to broaden its ongoing attack on repositories, shifting from a focus on just those campuses that have an open access policy to now inhibiting green self-archiving on all university campuses.  But they are doing so using a distinction that ultimately makes no sense.

That distinction gets really messy when we try to apply it to the actual conditions of campus IT, something Elsevier apparently knows little about and did not consider as the wrote the new policy documents.  I am reminded that, in a conversation unrelated to the Elsevier policy change, a librarian told me recently that her campus Counsel’s Office had told her that she should treat the repository as an extension of faculty members’ personal sites.  Even before it was enshrined by Elsevier, this was clearly a distinction without a difference.

For one thing, when we consider how users access these copies of final authors’ manuscripts, the line between a personal website and a repository vanishes entirely. In both cases the manuscript would reside on the same servers, or, at least, in the same “cloud.” And our analytics tell us that most people find our repositories through an Internet search engine; they do not go through the “front door” of repository software. The result is that a manuscript will be found just as easily, in the same manner and by the same potential users, if it is on a personal website or in an institutional repository. A Google or Google Scholar search will still find the free copy, so trying to wall off institutional repositories is a truly foolish and futile move.

For many of our campuses, this effort becomes even more problematic as we adopt software that helps faculty members create and populate standardized web profiles. With this software – VIVO and Elements are examples that are becoming quite common — the open access copies that are presented on a faculty author’s individual profile page actually “reside” in the repository. Elsevier apparently views these two “places” – the repository and the faculty web site – as if they really were different rooms in a building, and they could control access to one while making the other open to the public. But that is simply not how the Internet works. After 30 years of experience with hypertext, and with all the money at their disposal, one would think that Elsevier should have gained a better grasp on the technological conditions that prevail on the campuses where the content they publish is created and disseminated. But this policy seems written to facilitate feel-good press releases while still keeping the affordances of the Internet at bay, rather than to provide practical guidelines or address any of the actual needs of researchers.

I hope it was clear, when I wrote about the press release from Elsevier addressing their new approach to authors’ rights and self-archiving, that I believe the fundamental issue is control.  In a comment to my original post, Mark Seeley, who is Elsevier’s General Counsel, objected to the language I used about control.  Nevertheless, the point he made, about how publishers want people to access “their content,” but in a way that “ensures that their business has continuity” actually re-enforced that the language I used was right on the mark.

My colleague Paolo Mangiafico has suggested that what these new policies are really about is capturing the ecosystem for scholarly sharing under Elsevier’s control.  As Paolo points out, these new policies, which impose long embargo periods on do-it-yourself sharing by authors but offer limited opportunities to share articles when a link or API provided by Elsevier is used, should be seen alongside the company’s purchase of Mendeley; both provide Elsevier an opportunity to capture data about how works are used and re-used, and both  reflect an effort to grab the reins over scholarly sharing to ensure that it is more difficult to share outside of Elsevier’s walled garden than it is inside that enclosure.

I deliberately quote Mr. Seeley’s phrase about “their content” because it is characteristic of how publishers seem to think about what they publish.  I believe it may even be a nearly unconscious gesture of denial of the evident fact that academic publishers rely on others — faculty authors, editors and reviewers — to do most of the work, while the publisher collects all of the profit and fights the authors for subsequent control of the works those authors have created. That denial must be resisted, however, because it is in that gesture that the desire for control becomes outright disrespect for the authors that publishing is supposed to serve.

Nowhere is this disrespect more evident than in publisher claims that the works they publish are “work made for hire,” which means, in legal terms, that the publisher IS the author.  The faculty member who puts pen to paper is completely erased from the transaction.  To be clear, as far as I know Elsevier is not making such a claim with its new policies.  But these work made for hire assertions are growing in academic publishing.

Three years ago I wrote about an author agreement from Oxford University Press that claimed work made for hire over book chapters; that agreement is still in use as far as I am aware.  At the time, I pointed out two reasons why I thought OUP might want to make that claim.  First, if something is a work made for hire, the provision in U.S. copyright law that allows an author or her heirs to terminate any license or transfer after 35 years simply does not apply.  More significantly, an open access license, such as is created by many university policies, probably is not effective if the work is considered made for hire.  This should be pretty obvious, since our law employs the legal fiction that says the employer, not the actual writer, is the author from the very moment of creation in work made for hire situations.  So we should read these claims, when we find them in author agreements, as pretty direct assaults on an author’s ability to comply with an open access policy, no matter how much she may want to.

As disturbing as the Oxford agreement is, however, it should be said that it makes some legal sense.  When a work is created by an independent contractor (and it is not clear to me if an academic author should be defined that way), there are only selected types of works that can even be considered work made for hire; one of them is “contribution[s] to a collective work.”  So a chapter in an edited book is at least plausible as a work made for hire, although the other requirement — an explicit agreement, which some courts have said must predate the creation of the work — may still not be met.  In any case, the situation is much worse with the publication agreement from the American Society of Mechanical Engineers (ASME), which was recently brought to my attention.

ASME takes as its motto the phrase “Setting the Standard,” and with this publication agreement they may well set the standard for contemptuous maltreatment of their authors, many of whom are undoubtedly also members of the society.  A couple of points should be noted here.  First, the contract does claim that the works in question were prepared as work made for hire.  It attempts to “back date” this claim by beginning with an “acknowledgement” that the paper was “specially ordered and commissioned as a work made for hire and, accordingly, ASME is the author of the Paper.”  This acknowledgement is almost certainly untrue in many, if not most, cases, especially since it appears to apply even to conference presentations, which are most certainly not “specially commissioned.”  The legal fiction behind work made for hire has been pushed into the realm of pure fantasy here.

What’s more, later in the agreement the “author” agrees to waive all moral rights, which means that they surrender the right to be attributed as the author of the paper and to protect its integrity.  Basically, an author who is foolish enough to sign this agreement has no relationship at all to the work, once the agreement is in place.  They are given back a very limited set of permissions to use the work internally within their organization and to create some, but not all, forms of derivative works from it (they cannot produce or allow a translation, for example).  Apparently ASME has recently started to disallow some students who publish with them to use the published paper as part of a dissertation, since most dissertations are now online and ASME does not permit the  writer to deposit the article, even in such revised form, in an open repository.

To me, this agreement is the epitome of disrespect for scholarly authors.  Your job, authors are told, is not to spread knowledge, not to teach, not to be part of a wider scholarly conversation.  It is to produce content for us, which we will own and you will have nothing to say about.  You are, as nearly as possible, just “chopped liver.”  It is mind-boggling to me that any self-respecting author would sign this blatant slap in their own face, and that a member-based organization could get away with demanding it.  The best explanation I can think of is that most people do not read the agreements they sign.  But authors — they are authors, darn it, in spite of the work for hire fiction — deserve more respect from publishers who rely on them for content (free content, in fact; the ASME agreement is explicit that writers are paid nothing and are responsible for their own expenses related to the paper).  Indeed, authors should have more respect for themselves, and for the traditions of academic freedom, than to agree to this outlandish publication contract.

The announcement from Elsevier about its new policies regarding author rights was a masterpiece of doublespeak, proclaiming that the company was “unleashing the power of sharing” while in fact tying up sharing in as many leashes as they could.  This is a retreat from open access, and it needs to be called out for what it is.

For context, since 2004 Elsevier has allowed authors to self-archive the final accepted manuscripts of their articles in an institutional repository without delay.  In 2012 they added a foolish and forgettable attempt to punish institutions that adopted an open access policy by purporting to revoke self-archiving rights from authors at such institutions.  This was a vain effort to undermine OA policies; clearly Elsevier was hoping that their sanctions would discourage adoption.  This did not prove to be the case.  Faculty authors continued to vote for green open access as the default policy for scholarship.  In just a week at the end of last month the University of North Carolina, Chapel Hill, Penn State, and Dartmouth all adopted such policies.

Attempting to catch up to reality, Elsevier announced last week that it was doing away with its punitive restriction that applied only to authors whose institutions had the temerity to support open access. They now call that policy “complex” — it was really just ambiguous and unenforceable — and assert that they are “simplifying” matters for Elsevier authors.  In reality they are simply punishing any authors who are foolish enough to publish under these terms.

Two major features of this retreat from openness need to be highlighted.  First, it imposes an embargo of at least one year on all self-archiving of final authors’ manuscripts, and those embargoes can be as long as four years.  Second, when the time finally does roll around when an author can make her own work available through an institutional repository, Elsevier now dictates how that access is to be controlled, mandating the most restrictive form of Creative Commons license, the CC-BY-NC-ND license for all green open access.

These embargoes are the principal feature of this new policy, and they are both complicated and draconian.  Far from making life simpler for authors, they now must navigate through several web pages to finally find the list of different embargo periods.  The list itself is 50 pages long, since each journal has its own embargo, but an effort to greatly extend the default expectation is obvious.  Many U.S. and European journals have embargoes of 24, 36 and even 48 months.  There are lots of 12 month embargoes, and one suspects that that delay is imposed because those journals that are deposited in PubMed Central, for which 12 months is the maximum embargo permitted.  Now that maximum embargo is also being imposed on individual authors.  For many others an even longer embargo, which is entirely unsupported by any evidence that it is needed to maintain journal viability, is now the rule.  And there is a handful of journals, all from Latin America, Africa, and the Middle East, as far as I can see, where no embargo is imposed; I wonder if that is the result of country-specific rules or simply a cynical calculation of the actual frequency of self-archiving from those journals.

The other effort to micromanage self-archiving in this new policy is the requirement that all authors who persevere and wish, after the embargo period, to deposit their final manuscript in a repository, must apply a non-commercial and no derivative works limitation on the license for each article.  This, of course, further limits the usefulness of these articles for real sharing and scholarly advancement.  It is one more way in which the new policy is exactly a reverse of what Elsevier calls it; it is a retreat from sharing and an effort to hamstring the movement toward more open scholarship.

The rapid growth of open access policies at U.S. institutions and around the world suggests that more and more scholarly authors want to make their work as accessible as possible.  Elsevier is pushing hard in the opposite direction, trying to delay and restrict scholarly sharing as much as they can.  It seems clear that they are hoping to control the terms of such sharing, in order to both restrict it putative impact on their business model and ultimately to turn it to their profit, if possible.  This latter goal may be a bigger threat to open access than the details of embargoes and licenses are. In any case, it is time, I believe, to look again at the boycott of Elsevier that was undertaken by many scholarly authors a few years ago; with this new salvo fired against the values of open scholarship, it is even more impossible to imagine a responsible author deciding to publish with Elsevier.