The Occasional Pamphlet

A strange social contract has arisen in the scholarly publishing field, a kind of “don’t ask, don’t tell” approach to online distribution of articles by authors.  Publishers officially forbid online distribution, authors do it anyway without telling the publishers, and publishers don’t ask them to stop even though it violates contractual obligations. What happens when you refuse to play that game? Read on.

Publishing of research, in the sense of providing for its widespread dissemination, is the means by which new discoveries join the collective knowledge of humanity.  The means by which the distribution is implemented has been subscription-based publication via a publisher.  Indeed, until recently, this was the only practical means by which research could be distributed, since the cost of the dissemination, which involved printing and shipping of paper, showed economies of scale that individual researchers could not take advantage of.  Publishers made revenue by limiting access to the papers to paying customers.

Not coincidentally, scholars are motivated to publish because recognition of their research by this same system is the currency by which their career advancement is purchased.  They are thus forced, effectively as a condition of employment, to make their research available through publishers.

With the availability of the internet, the marginal cost of distribution of a scholarly paper has been reduced to essentially zero.  That particular economy of scale benefit that traditional print publishing relied on has disappeared.  Still, participation in the traditional publication system remains important for its peer-review-based vetting system.

However, recognition of scholars’ research is also measured by other metrics of impact, such as subjective judgements of influence on the field, sometimes formalized in proxies such as citation counts and impact factors.  To optimize these metrics, it behooves a scholar to distribute writings universally, to take advantage of the network and its low costs.  Study after study has shown that articles available freely online are more widely cited than those that are available only through publishers’ access-limited venues.

An author has a simple solution to the quandary of whether to distribute through a publisher’s access-limited mechanism or freely online: Do both.  Unfortunately, publishers typically restrict authors from this approach through contractual limitations stipulated in copyright assignment forms.

This brings us to the strange social contract.  What has arisen, perhaps surprisingly, is a kind of “don’t ask, don’t tell” approach to online distribution by authors.  Publishers officially forbid online distribution, authors do it anyway without telling the publishers, and publishers don’t ask them to stop even though it violates contractual obligations.

The standard system for scholarly communication is thus based on widespread contractual violation and fraud.

Why don’t publishers police their contractees more carefully, as the RIAA does with their customers who distribute copyrighted material online?  It would be a simple matter to find a small number of violators and send them take-down letters.  General publicity of this effort would presumably have a substantial chilling effect among academics against their routine violation of copyright assignments.

We can only speculate that the fear of upsetting their content providers trumps their need to maintain control over the content itself, given that there is no evidence that the online availability is hurting their revenues.  (The situation may be different with music, hence the RIAA’s different strategy.)  One can just imagine what kind of backlash an RIAA-style approach would have with academics, in addition to the desired chilling effect.

Nonetheless, individual authors still breach contracts regularly as they act to maximize their career advancement possibilities.  To many, including myself, this state of affairs is untenable.  I am not willing to routinely violate contracts in this way.  Consequently, I and others have for some time reconciled the two distribution mechanisms explicitly, by amending the contractual conditions of copyright assignments.  For many years, I have as a matter of course refused to sign copyright assignment forms that do not give me the right of noncommercial online distribution of my work. Originally, I would use alternative copyright assignments that I wrote myself.  More recently, I have been attaching the SPARC addendum to publishers’ assignment forms, and then the Science Commons addenda that superseded it.

In the many years that I have been routinely replacing or modifying copyright assignments, I have never had a complaint (or even an acknowledgement) from a publisher.  In retrospect, this may make sense.  Since the contractual modification applies only to a single article by a single author, it is unlikely that anyone looking for copyright clearance would even know that all copyright hadn’t been assigned to the publisher.  And in any case publishers must realize that authors act as if they have a noncommercial distribution license whether they formally retain one or not.

I say that I’ve never had a complaint from a publisher, and that has been true with one exception.  This post describes that singular case.  It may serve to illuminate several points:

• How journal publishers think about rights.
• How the rights landscape might be changing.
• How authors can recoup positive progress.

I describe my experience in challenging an irrational and detrimental license clause, and how it spiraled into a battle that resulted in the publisher changing its policy for the journal as a whole. My experience is certainly not unique but accounts are rare, so I encourage others to share their experiences with successful (and unsuccessful) rights retention negotiations with journal publishers in the comments section.

In May of 2004, I submitted a paper to the Blackwell philosophy journal Nôus.  The contents of the paper is immaterial^[1] and its trajectory through the normal review process is unexceptional. After a round of revisions, the paper was accepted for publication in the journal in November of 2004, some six months after submission, and my final copy was submitted the following May and a copy placed on my web site.

After the article sat in the journal’s backlog for eighteen months, I received notification in November of 2006 that the paper’s publication in volume 41, number 1 of Nôus (the March, 2007 issue) was imminent. Blackwell would be sending me proofs around December 19 (in the event, they arrived December 21) and I was to send any corrections by January 5 coincident with the winter holidays to perform the work.  At that time the managing editor of the journal (an employee of Brown University, not the publisher) stated that Blackwell had changed the copyright form that they require from authors from a Copyright Transfer Form to an Exclusive License Form (ELF).  She attached a copy for me to sign and return.

The copyright policy of Blackwell was only alluded to on their web site. Their public statement on open access stated that “Our copyright assignment policy allows authors to self-archive their final version of their article on personal websites or institutional repositories.” Historically, Blackwell has been quite progressive when it comes to copyright policies and author rights.  For all these reasons, I had little reason to expect a problem with distributing my paper online.

However, Blackwell’s Exclusive License Form turned out to be a mixed bag.  In most ways it was progressive, specifically allowing prepublication distribution on the author’s web site and afterwards as well, with the exception of a twelve-month hiatus immediately after publication, what is termed in the parlance an “embargo period”. With the exception of the embargo period, the ELF seemed relatively unobjectionable.  As was my usual practice, I attached the SPARC addendum and returned the ELF to Blackwell shortly after my final corrections to the proofs in early January 2007.

On February 1, 2007, an assistant journals editor at Blackwell (who I will refer to as “L—” to retain anonymity) informed me by email that

We noticed that you submitted a signed SPARC Author's Addendum along with the signed Blackwell Exclusive License Form (ELF) for your article 'The Turing Test as Interactive Proof', which will appear in an upcoming issue of Nous. I have attached a copy of these forms for your reference.^[2] According to the terms of the SPARC agreement the Blackwell ELF takes precedence. However, I wanted to make you aware that several of the points in the SPARC agreement differ with the rights you agreed to in the Blackwell ELF.

This email struck me, frankly, as bizarre, and the Office of the General Counsel at my home institution seemed puzzled as well.  I replied on February 7.

L— --

Thanks for bringing up a possible misconstrual of the copyright provisions.  I and many others appreciate the expansiveness of the Blackwell copyright assignment.

I recognize that there are some differences between the terms of the Blackwell ELF and the SPARC addendum.  My understanding of the SPARC addendum is that, where there is conflict between the addendum and the Blackwell ELF, the addendum takes precedence, based on the following language in the addendum:

"The parties agree that wherever there is any conflict between this Addendum and the Publication Agreement, the provisions of this Addendum are paramount and the Publication Agreement shall be construed accordingly."

I don't understand your statement that "According to the terms of the SPARC agreement the Blackwell ELF takes precedence."

Regards,

Perhaps this was an attempt on the part of Blackwell to have on record my acquiescence to the Blackwell ELF as taking precedence over the addendum despite the addendum’s plain language to the contrary.

On February 9, L— now adjusted her claim saying that the SPARC form “does appear to be unclear” as to precedence.  In any case,

Blackwell and the Nous Editorial office are not in agreement with the terms of the SPARC Addendum, especially the lack of embargo period, which could jeopardize the journal.

This email for the first time was copied to Professor Ernest Sosa, the editor-in-chief of Nôus.

This began a period of emails back and forth in which I tried to find a minimal change to the ELF that Blackwell would accept.  Blackwell viewed their job as “to be the guardian of Nous content”; they must “protect the journal”.  They worried that my posting my paper on my web site over the next twelve months (where it had been for the last several years) “could jeopardize the journal”.

Meanwhile, fearing the worst, I started contacting Nôus board members for advice on the matter.  By February 28, an email from L— reminded me that time was running out and a resolution had to be reached by March 2 or the paper would have to be pulled from its intended issue. As a final attempt, I sent a copy of the ELF with just four words elided to drop the embargo period, the first four words of the clause “12 months after publication you may post an electronic version of the Articles on your own personal website, on your employer’s website/repository and on free public servers in your subject area.”

L— requested a phone call.

On March 1, I spoke to L— by phone.  She was quite pleasant and professional, but stated definitively that the embargo period was non-negotiable.

We talked about various issues regarding the policy.  First, I was interested in the reasons for the embargo period.  She provided three, serially, as I argued against each.  I turn to them below.

I asked if the embargo policy was a Blackwell policy or a policy of the Nôus editorial board.  She said the latter, and reported that she had spoken with the editor-in-chief who she said was supportive of Blackwell’s stance.

In the end, neither of us would budge, so we agreed to disagree and L— said she would have to pull the paper.  I emphasize that she was gracious and understanding throughout.

During the phone call, L— put forth three arguments for the embargo period, none of which was well founded.

The citation argument: The first claim was that the embargo period is necessary so that there are not multiple versions available that could cause citation confusion.  However, Blackwell’s own agreement allows posting of papers both before and after the embargo period, which they apparently do not believe leads to citation confusion.  It is unclear why there would be confusion generated just within the embargo period.  In any case, with or without an embargo period, the agreement already requires that the posted version provide correct citation information for the definitive version, so that anyone downloading the paper is provided accurate information.

The economic argument: Second, L— claimed that the embargo period was necessary for the financial health of the journal. Presumably, the worry is that if papers were available for free without a subscription during the embargo period, libraries would stop subscribing, eliminating the main source of Blackwell’s revenue from the journal.  In fact, although the journal officially has this embargo policy, it seems to be honored more in the breach.  A quick study revealed that over 80 percent of the Nôus articles within the embargo period at the time were available from the authors’ web sites or other open access repositories. (In fact, some authors were even posting the publisher’s version.) Thus, over 80 percent of Nôus articles were already in violation of the copyright agreement that the publisher asserts is critical for the journal’s financial health.  Furthermore, there is ample empirical evidence from fields such as physics that enjoy essentially 100 percent open access to preprints that availability of preprints does not lead to reductions in subscriptions.

The standards argument: Finally, L— claimed that the embargo period was “standard policy in the industry”.  This is false.  For instance, Elsevier allows posting of authors’ final versions of papers on authors’ websites without an embargo period. Springer-Kluwer has a similar policy (as reported by SHERPA/RoMEO). Apparently even among Blackwell journals embargo periods varied with some having no embargo period. But even if the claim were true, the fact that an embargo period is standard is no argument for retaining the restriction if there is no other good argument for it.

In summary, Blackwell was singling me out for special treatment in not allowing me to post a copy of my paper on my web site, whereas all other Nôus authors are apparently allowed to do so with impunity. As far as I can tell, I was singled out solely because I was attempting to do legally what other authors were willing to do illegally.  No plausible argument for the restriction was provided. Indeed, the fact that violation of the contractual condition is so overwhelmingly widespread demonstrates that the contractual condition is unnecessary.

At this point, I gave up on Blackwell making an exception for me, and moved to try to get Nôus to change its policy overall, following on L—’s assurance that the embargo period was a policy of the editorial team of the journal.  I enlisted help from a Nôus board member to ask the editor-in-chief Sosa to reconsider the policy.  Eventually, he did so, stating that “something needs to be done” and that the “situation with Nous seems unsustainable”.  He passed on a request to his Blackwell contacts to change the Nôus copyright policy.

On May 5, 2007, Sosa reported that the matter was in the hands of the Blackwell legal department, but that Blackwell’s pending merger with Wiley may slow things down.

Finally, on May 24, 2007, Sosa informed me that Blackwell had agreed to drop the embargo requirement from the Nôus copyright form, and that my article could now reenter the publication stream.  The paper appeared, finally, in volume 41, number 4 of Nôus, and remained on my web site with updated citation information.

What is the moral of this story?  First, all participants — including the editorial board and editor-in-chief of the journal, the managing editor, the publisher’s staff — were people of good will. They all acted in ways they thought in the best interest of the institutions they represented and the larger missions of those institutions.  However, to a great extent, they may not have fully thought out the connections between the policies they acted under and the missions.  The editorial board may not have realized that the journal’s policy embargoed author distribution; certainly the journal’s contributors didn’t, or chose to ignore it.  The publisher may not have realized the inconsistencies between the journal policies and the facts-on-the-ground.

But it is also apparent that authors are far too acquiescent in the process of rights retention with publishers.  We are overly willing to accept the rulings of publishers as a fait accompli.  Despite the fact that publishers assert that their policies are supported by their editorial boards, editorial boards are in fact responsive to reasoned arguments.  And although a negotiation for rights retention between an author and a large commercial publishing company asymmetrically disfavors the author, one in which the author is supported by the editorial board is a different matter entirely.  This example calls for taking advantage of rights retention negotiations to enlist editors and editorial boards in the process of expanding access to scholarly articles in a way consonant with law, moving past the “don’t ask, don’t tell” social contract.

Update 6/19/11: Jeffrey Pomerantz has an interesting story about his own “copyfight” with Taylor and Francis, ending in the publisher refusing to vary the conditions of the publication agreement and his decision to withdraw the article, providing it via unilateral open access instead. He raises the only partially tongue-in-cheek possibility of this as a new scholarly communication strategy. Worth thinking about.

^[1] But since you asked, the paper concerns the status of the Turing Test as a philosophically tenable criterion for attributing intelligence.  The paper is freely available from my web page and other sites.

^[2]She didn’t and the email appears to have been truncated.