The Shieber Model Publication Agreement: an Analysis

As mentioned in the previous post, Stuart Shieber at Harvard proposed a model OA journal publication agreement on his blog.  In many ways, it is comparable to the simplified agreements made available by the University of Michigan, and Shieber points to another example as a starting point under the label of an "Online Guide to Open Access Journals Publishing", which some cursory Googling found mostly dead links and a few references to a project from 2010 that may have changed its name or merged with another project since.

That starting point agreement is essentially a CC-BY license with an additional clause granting first publication rights to the journal.  In essence, the journal executing that agreement would become the "official" citation.  It's simple, painless, and also likely inadequate for many journals--even very OA-centric ones.  The UM license is more in the ballpark of a robust, generally OA publication agreement. But while it accomplishes OA ends, it's written as an arrangement between an author and a publisher that permits open access, rather than an agreement for the use of a work that is open access.  Shieber's model is similar, but contains some helpful added specificity.



In the previous post, I mentioned the consequences of releasing a manuscript via CC-BY and the result that journals would not need a publication agreement simply to edit or publish the work in that case. Since the value of a CC-BY article to a journal would be the propagation of the citation in other works (broadly equivalent to "first publication rights", though that's a loaded term in academic publishing) and potentially some clearer legal warranty/indemnification from the author, that's where I feel the publication agreement should focus in the world of open access.  What follows is a clause-by-clause review of the Shieber model and, in my next post a proposal for a variant that addresses these concerns.

PUBLICATION AGREEMENT

• Header: "[This is version 1.2 of 8 January 2014. It is made available under a Creative Commons Attribution 4.0 License. The license's attribution requirement can be satisfied by including language conveying at least the following: “The language of this publication agreement is based on Stuart Shieber’s model open-access journal publication agreement, version 1.2, available at http://bit.ly/1m9UsNt.” If you intend to use this agreement, I request as a courtesy that you notify me so that I can track usage.]"

How meta...a license for a license! This is increasingly common and certainly useful in version tracking and cross-referencing.

• Preamble: This is a publication agreement (“this agreement”) regarding a written manuscript currently entitled […] (manuscript title) (“the article”) to be published in […] (“the journal”). The parties to this Agreement are: […] (the corresponding author on behalf of any other authors, collectively “you”) and […], (“the publisher”).

Pretty standard stuff. Defined terms don't need definite articles and do need to be capitalized or stylistically set off from other words in the agreement.  Because all caps is standard form for warranty language and because capitalizing only the first letter starts to make English contracts look like German, I personally prefer the "small caps" style, but that is frustratingly difficult to achieve on the web, so I'll go with sentence case and bold print instead in my sample below (so "'this agreement'" will become "'Agreement'").

The point of trouble here is the "on behalf of any other authors" part, which also appears in term 1:

• By signing this form, you warrant that you are signing on behalf of all authors of the article, and that you have the authority to act as their agent for the purpose of entering into this agreement.

Where there are multiple authors, you're dealing either with a compilation or a joint work. If a compilation, the submitting author should have agreements in place with the other contributors allowing this. But if it's a joint work, it's not strictly necessary to obtain written permission from coauthors. Each joint author is an owner for the purposes of copyright law, and each may exercise copyrights. Where things get dicey is when offering exclusive rights without the permission of all joint owners, but in an OA publication agreement without exclusive rights, the submitting author doesn't need to act on behalf of any other authors.

• 2: You hereby grant a Creative Commons copyright license in the article to the general public, in particular a Creative Commons Attribution 4.0 International License, which is incorporated herein by reference and is further specified at http://creativecommons.org/licenses/by/4.0/legalcode (human readable summary at http://creativecommons.org/licenses/by/4.0).

"Hereby grant" is the wrong mechanism for accomplishing this, because it makes the CC release contingent upon execution of the agreement.  It would be more accurate to say that the article in question is available under CC-BY, irrespective of the status of the publication agreement.

• 3: You agree to require that a citation to the original publication of the article in the journal be included in any attribution statement satisfying the attribution requirement of the Creative Commons license of paragraph 2.

I think this is trying to establish that the attribution form in the CC-BY release of the article will include the journal citation.  Wording could be more clear.

• 4: You retain ownership of all rights under copyright in all versions of the article, and all rights not expressly granted in this agreement.

Works for me.

• 5: To the extent that any edits made by the publisher to make the article suitable for publication in the journal amount to copyrightable works of authorship, the publisher hereby assigns all right, title, and interest in such edits to you. The publisher agrees to verify with you any such edits that are substantive. You agree that the license of paragraph 2 covers such edits.

This is an important omission from the UM version and I'm glad to see it here. Although there is a comment on the original blog that a publisher should "never" make changes that constitute a copyrightable work, that is only potentially true of the manuscript text. The formatted version of the article is subject to copyright (hence the preprint/published version distinction) and when artwork or print-ready versions of graphs are prepared, those too are copyrightable. Not clarifying that any new material generated in the course of publication may be used by the author would be a mistake in my opinion. However, I am not sure that an assignment is the appropriate mechanism.  If you won't assign your copyright to the journal, why should the journal assign its copyright to you?

Term 6 - Warranties

• 6a: The article is original, has not been formally published in any other peer-reviewed journal or in a book or edited collection, and is not under consideration for any such publication.

You can really only make this representation if you have not yet released a substantially similar version under a CC license or posted the paper online.  Once it's out there in basically final form, you truthfully have no way of knowing its publication status (unless you use the NC license variants) since no one is under any obligation to seek your permission before submitting it for peer review or publication. To do so would be bad form, and most journals are unlikely to accept submission from non-authors, but it wouldn't be a violation of the law or the license terms.

• 6b: You are the sole author(s) of the article, and that you have a complete and unencumbered right to make the grants you make.

This is fine, except for the "sole author" part, which contradicts the multiple author arrangement contemplated up top.

• 6c: The article does not libel anyone, invade anyone’s copyright or otherwise violate any statutory or common law right of anyone, and that you have made all reasonable efforts to ensure the accuracy of any factual information contained in the article. You agree to indemnify the publisher against any claim or action alleging facts which, if true, constitute a breach of any of the foregoing warranties or other provisions of this agreement, as well as against any related damages, losses, liabilities, and expenses incurred by the publisher

Here is a big one and the real value of the agreement for the journal apart from the citation.  The ideas are sound, but the language is nonstandard.  Indemnification really should be broken out into its own term, and that language as written puts the author on the hook in a bigger way than is really necessary.

• 7: This is the entire agreement between you and the publisher, and it may be modified only in writing. It will be governed by the laws of […the Commonwealth of Massachusetts…]. It will bind and benefit our respective assigns and successors in interest, including your heirs. It will terminate if the publisher does not publish, in any medium, the article within one year of the date of your signature.

An integration clause is important to have in any contract (one of the vulnerabilities of some of the other samples), and the time limit on publication is theoretically sound, but technically problematic because circumstances outside the journal's control (including conduct of the author) may run down the clock. If I were a journal, I would want more flexible language to refocus on the underlying aim of the provision: to ensure that journals work diligently toward actually publishing the article.