Tuesday, February 04, 2014

Guest Post: Charles Oppenheim on who owns the rights to scholarly articles

The recent decision by Elsevier to start sending take down notices to sites like Academia.edu, and to individual universities, demanding that they remove self-archived papers from their web sites has sparked a debate about the copyright status of different versions of a scholarly paper.

Last week, the Scholarly Communications Officer at Duke University in the US, Kevin Smith, published a blog post challenging a widely held assumption amongst OA advocates that when scholars transfer copyright in their papers they transfer only the final version of the article. This is not true, Smith argued.

As he put it:

Each version is a revision of the original, and the copyright is the same for all these derivatives.  When copyright is transferred to a publisher, the rights in the entire set of versions, as derivatives of one another, are included in the transfer.  Authors are not allowed to use their post-prints because the rights in that version are not covered in the transfer; they are allowed to use post-prints only because the right to do so, in specified situations, is licensed back to them as part of the publication agreement.

If correct, this would seem to have important implications for Green OA, not least because it would mean that publishers have greater control over self-archiving than OA advocates assume.

However Charles Oppenheim, a UK-based copyright specialist, believes that OA advocates are correct in thinking that when an author signs a copyright assignment only the rights in the final version of the paper are transferred, and so authors retain the rights to all earlier versions of their work, certainly under UK and EU law. As such, they are free to post earlier versions of their papers on the Web.

Charles Oppenheim explains his thinking on this below:

Charles Oppenheim

In this article, I will try to tease out the copyright ownership issues associated with scholarly articles.  Before I do so, I first have to explain certain terms used in this piece:

Assignment is when the current copyright owner of a work gives or sells that copyright to a third party.  Assignments cannot be done informally.  They require a signed contract (a formal contract or a letter).  The need for a signature means it is usual to do this by means of the post, face to face, or by fax, though it is possible to do so using e mail under some circumstances.

In contrast, a licence (spelt “license” in the USA) occurs where the copyright owner retains copyright in the work, but grants a third party (the licensee) certain permissions to do things with the work which would not normally be allowed under copyright law.  Licences are usually in writing (for example, the licences libraries sign for access to Elsevier’s Science Direct).  Licences may be charged for, or can be free of charge, e.g., Creative Commons licences. But unlike assignments, licences can exist without any formal agreement. 

Such licences are called implied licences.  In such cases, permissions are granted by custom and practice, rather than by formal agreement. A typical example is where an author submits a manuscript to a scholarly journal.  The author grants the journal editor and publisher an implied licence to forward a copy of the manuscript to peer reviewers and/or to store the article temporarily on its management systems.

Imagine I prepare an article that I would like published in a scholarly journal.  Let us call it D (for draft).  I send D to the editor of the journal, who also receives the implied licence from me mentioned above.  The editor forwards D to some peer reviewers.  In due course, the reviewers pass comments back to the editor, who in turn summarises them in a message to me.  Let us assume certain amendments are required to D.  I make those amendments, and submit the final manuscript (called F) back to the editor.  The editor then asks me to assign copyright in F to the publisher.  Let us assume I sign it.  What are the legal consequences of that signature?

1.      The publisher acquires the copyright in F.  This means no third party – including myself – is allowed to copy, disseminate or amend F without the publisher’s express permission – which may well not be forthcoming.  If I do copy F, say onto an Open Access repository without permission, I am infringing the publisher’s copyright, and the publisher is entitled to insist I take it down (as Elsevier has famously done recently), and could in principle sue me for damages because I have infringed its copyright.

2.      But crucially, I retain copyright in D.  How come?  Because the assignment I gave relates ONLY to F. Assignments precisely specify what is being assigned, and nowhere does the assignment I signed refer to “precursors of F”.  Indeed, it cannot, because that would include the very first stab I made at writing the article, perhaps just a few sentences written many months previously and bearing no relation to D or F. This crucial difference means I am free to do anything I like with D, including posting it on an OA repository.

If D is identical to F – in other words, the reviewers and the editor are so happy with my manuscript that no changes are needed, what follows below does not apply.  However, I suspect in the vast majority of cases, D is different to F.  Note that it is irrelevant if F is very similar to D, or is greatly different, to what follows below.

Posting D on an OA repository is the so-called “Harnad-Oppenheim” solution, first proposed by Stevan Harnad and me more than 10 years ago. 

When the solution was first enunciated, publishers dismissed it for two reasons:  firstly, why would anyone want to read a draft when the final perfect version can be obtained via the publisher? And secondly, it would be difficult to track down a copy of D anyway.  Their comments remain valid today, though the second one is not as strong because of services such as Google Scholar.  But no publisher suggested that the solution was illegal because publishers owned the copyright to D, and they were right not to do so.  The law is clear that I own the copyright in D. That is why I am so puzzled that some recent non-publisher commentators seem to think publishers own the copyright in D. 

Another idea going the rounds is to post F on a repository before signing anything with the publisher, so the publisher is forced to accept that the item has already appeared.  The problem with that approach is that publishers’ licence or assignment terms require the author to confirm that F has not been published before, or is not being considered for publication elsewhere. That’s why it must be D that is posted, rather than F.

I don’t particularly recommend the Harnad-Oppenheim solution, for the reasons publishers gave when the solution was first propounded.  The approaches authors should be taking, in order of preference, are:

1.      Offer the article only to an OA journal or some other OA vehicle

2.      Offer the article to a subscription-based journal which is happy that you give the publisher a sole licence to publish, leaving you free to put F on a repository, possibly after an embargo period

3.      Offer the article to a subscription-based journal, which nominally requires assignment, but will back off and let you insist on a licence if you stick to your guns. (Elsevier is a good example)

4.      Agree to assign copyright to the publisher, and then use the Harnad-Oppenheim solution.

5.      Agree to assign copyright and don’t do anything more.

For nearly 20 years, I have only used options 1, 2 or 3 without problems. I don’t use option 4 because I’ve not needed to. (Disclaimer: where my article has been co-authored with someone from another institution and they are happy to use option 5, I have gone along with it – but that’s been very rare). 

N.B. This commentary is based upon my understanding of UK and EU law; the laws of other countries may be different, but I believe that a copyright assignment in every major country requires a written, signed agreement that refers precisely to the work being assigned, and only that work.


This text is made available under a CC BY ND licence.  As such anyone is free to re-use it as long as Charles Oppenheim’s name is retained and it is not amended in any way.


Charles Oppenheim was, until he retired in 2009, Professor of Information Science and Head of the Department of Information Science at Loughborough University. He is currently a Professor at Robert Gordon University, Aberdeen, and a Visiting Professor at Cass Business School, part of The City University, London.  Previously he held posts in other academic institutions, and for twelve years worked in the electronic publishing industry. 

Charles is an Honorary Fellow of the Chartered Institute of Library and Information Profession.

He has been involved in, given talks on, and published on the library and information professions, Intellectual Property Rights and other legal issues relevant to the library and information professions, bibliometrics, evaluation of research quality, and on scholarly publishing trends.

He can be followed on Twitter: @CharlesOppenh


Leslie Carr said...

Publishers are caught on the horns of a dilemma here: if they claim that D and F are "the same" and that the assigned copyright applies to the previous version, the corollary is that the publisher has added no noticeable value to F. And that would be a strong admission that the case for Open Access is correct.

Toma said...

I was under the impression that the publisher only receives copyright to the final peer-reviewed AND copyedited version (i.e., a postprint), but that the authors retain the rights to the peer-reviewed but author-formatted manuscript (i.e., a preprint). But apparently this is not the case...

But isn't the peer-reviewed but author-formatted version what we are supposed to deposit in repositories in Green Open Access?

Charles Oppenheim said...

Toma, it all depends on the precise wording of the assignment the author signs, and also the permissions, if any, the publisher gives back to the author. But the standard wording just refers to copyright in "the article" being assigned.

Les, totally agree with your point.

Dan said...

Using your nomenclature, F is derivative of the copyrighted work D. Under U.S. law, creating a derivative work is an exclusive right of the copyright owner. Following your reasoning, I could sue the publisher for copying and distributing F even though I had assigned it to the publisher as they infringe my rights in D.

Unknown said...

Toma, my working understanding has been that the drafted, submitted version (prior to peer-review)is the pre-print; and that the post-print is the accepted, post peer-reviewed article, which has yet to be copy-edited and formatted by the publisher. So there are 3 versions; submitted (pre-print), accepted (post-print), and Final Published (publisher's version)

Jan Velterop said...

In reality, the 'final' manuscript (in which changes as a result of reviewers' comments have been incorporated) differs very little – if at all – from the published version. Significant subsequent copy-editing is a rarity.
Charles may well be right, but it means that by having secured copyright, publishers have secured precisely nothing of any value to them. According to his interpretation, an entrepreneurial new open access publisher could offer to publish the 'final' author's manuscript under a CC-BY licence, give it a DOI, and even make a nice PDF version, all for a small fee (not much investment is involved), with a footnote saying that "the copyright to a practically identical version of this article has been transferred to publisher X and it has been published in journal Y, volume N, issue M, date Z", so securing an association with any prestige journal Y might have.

Charles Oppenheim said...

Jan, I didn't say that. I was referring to the draft (D) as submitted to the publisher *before* peer-review. Copyright in the version following peer review has been assigned to the publisher, who then produces a finalised typeset version of a document to which they already own the copyright.

As I indicated, the author owns the copyright to D, so can put it on an OA server, or - as you suggest - offer it to another publisher, but why would that publisher want to publish such a document?

Dan, I don't follow your logic. If I have freely assigned copyright in F to publisher, I have given up all rights to F. The fact that F is derived from D becomes irrelevant.

Mike Taylor said...

Note that certain classes of work carry no copyright in the first place -- for example, papers written by employees in the US Federal Government in the course of their work. Publishers, knowing this, typically have an alternative clause in their copyright transfer agreements that allows the author to state "this work is in the public domain, and so the publisher is free to publish it".

By analogy, I encourage all authors who feel they need to submit their work to barrier-based journals simply to dedicate their manuscript to the public domain before submitting it, and sign that arm of the publishing agreement. This is easily done using the CC0 tool.

Stevan Harnad said...


1. Sixty percent of journals (including Elsevier) state formally in their copyright agreements that their authors retain the right to make their final, peer-reviewed, revised and accepted version (Green) Open Access (OA) immediately, without embargo, by self-archiving them in their institutional repositories.

2. The Elsevier take-down notices did not pertain to the author’s final version but to the publisher’s version of record (and in the case of 3rd party sites like academia.edu they concerned not only the version but the location).

3. The IDOA (immediate-deposit, optional-access) mandate is formally immune to take-down notices, because it separates deposit from OA:

4. For articles published in the 60% of journals in which authors formally retain their right to provide immediate, unembargoed Green OA, they can be self-archived immediately in the institutional repository and also made OA immediately.

5. For articles published in the 40% of journals that formally embargo OA, if authors wish to comply with the publisher’s embargo, the final, peer-reviewed, revised and accepted version can still be deposited immediately in the institutional repository, with access set as Closed Access (CA) during any embargo: only the title and abstract are accessible to all users; the full text is accessible only to the author.


Stevan Harnad said...


6. For CA deposits, institutional repositories have an email-eprint-request Button with which individual users can launch an automated email request to the author for an individual copy for research purposes, with one click; the author can then decide, on an individual case by case basis, with one click, whether or not the repository software should email a copy to that requestor.

7. It is the IDOA + Button Strategy that is the update of the “Harnad-Oppenheim Prepint + Corrigenda” Strategy.

8. But of course even the IDOA + Button Strategy is unnecessary, as is definitively demonstrated by what I would like to dub the “Computer Science + Physics Strategy”:

9. Computer scientists since the 1980’s and Physicists since the 1990’s have been making both their preprints and their final drafts freely accessibly online immediately, without embargo (the former in institutional FTP archives and then institutional websites, and the latter in Arxiv, a 3rd-party website) without any take-down notices (and, after over a quarter century, even the mention of the prospect of author take-down notices for these papers is rightly considered ludicrous).

10. I accordingly recommend the following: Let realistic authors authors practice the Computer Science + Physics Strategy and let formalistic authors practice the IDOA + Button Strategy — but let them all deposit their their final, peer-reviewed, revised and accepted versions immediately.

Sale, A., Couture, M., Rodrigues, E., Carr, L. and Harnad, S. (2012) Open Access Mandates and the "Fair Dealing" Button. In: Dynamic Fair Dealing: Creating Canadian Culture Online (Rosemary J. Coombe & Darren Wershler, Eds.)

Jan Velterop said...

Charles, I accept immediately that my reading of the situation is not quite what you said. But the idea of a 'draft' remains vague, at least in my view. If I look at what many – perhaps most – 'green' advocates say, it is that the 'final' manuscript is the one that should, and can, be deposited without © problems. Indeed, many a mandate requires 'final' manuscripts. The NIH formulates it as follows: "[The mandate] requires scientists to submit final peer-reviewed journal manuscripts that arise from NIH funds to the digital archive PubMed Central immediately upon acceptance for publication." Clearly, the 'final' manuscript is the one that is submitted after (perhaps more than one round of) peer review, possibly even mediated by other journals than the one it is eventually submitted to, after a rejection or two. There is an allowance for a 12-months embargo in this case, but the difference between the 'final' manuscript and the published article is basically reduced to some formatting and the like.

If I follow you correctly, it is in the publisher's interest that version D needs a lot of work, after peer review. Accepting it pretty much 'as is' makes D too similar to F and destroys the potential for 'added value', and therefore the value of the © for the publisher.

The whole system of © transfer in exchange for publishing services is messy and it relies on an unambihuous definition of what you call the D version. The existing ambiguity in copyright interpretations, however, makes it into a "who-blinks-first" system. Usually the author blinks first, of course, as simply having to defend yourself in court is costly and individuals have much fewer funds available for litigation than publishers, in the main.

The 'gold' method of open access publishing, whether paid for by author-payments or subsidies (e.g. SciELO.org), is much cleaner, transparent and straightforward. We can probably agree on that.

Charle Oppenheim said...

I'm happy with both Stevan's comments, Mike's comments and Jan's latest comments! We seem to haver reached consensus.....

Mike Taylor said...

"I'm happy with both Stevan's comments, Mike's comments and Jan's latest comments! We seem to haver reached consensus....."



Richard Poynder said...

Below is a comment from Michael Carroll, Professor of Law and Director of the Program on Information Justice and Intellectual Property at American University Washington College of Law:

This is an old issue. Kevin Smith is correct. Here is my version of why from 2006.

The way to understand this is to forget about the sequence by which an article is produced and think only about the rights that a copyright owner has. On Professor Oppenheim's view, the copyright owner's exclusive right of reproduction would be limited to controlling only verbatim copies. If that were true, I would be free to republish the entire corpus of Elsevier publications if I make only small changes to the articles similar to the differences between a final draft and the final publication. Needless to say, if this were the law, some clever publisher would have done just as I suggest. But, this is not the law in the US or in the UK.

So even if the publisher were to be assigned rights only in the final version of an article and most publication agreements are not this limited the scope of those rights would preclude posting of substantially similar versions whether those versions were created before or after the published version is produced. (US law uses the term "substantially similar" whereas
UK law asks whether the copyright work has been copied "in substantial part" but it effectively means the same thing in this context. See here.

Now, Steven Harnad is quite correct that in a majority of publication agreements, the publisher receives exclusive rights either by assignment or by exclusive license and then grants back to the author the non-exclusive rights to post some earlier version of the article. And, he is quite correct that not enough authors exercise the rights they have under their existing publication agreements. He is also right that they may (and should!) deposit copies of their final manuscripts in institutional or disciplinary repositories because the act of deposit is covered by the exceptions and limitations to copyright such as fair use or fair dealing in a substantial number of countries around the world.

Charles Oppenheim said...

Professor Carroll has completely misrepresented me. As I made clear in my point 1, no-one can adapt or amend F without the publisher's permission. His misrepresentation must have been as a result of one of three things: (a) Prof Carroll never read my piece; (b) he read it and deliberately misconstrued what I wrote; or (c) he read it and did not understand what I wrote. I was totally clear that the author has rights to D, but cannot do anything with F. I now expect Prof Carroll to apologise for misrepresenting me and to explain which of (a), (b) or (c) was the reason.

He also notes the UK IPO statement that infringement covers copying all or a substantial part of a copyright work. I agree, but such copying has to be AFTER F is published. One cannot copy something before it is made! So he is again referring to my case 1, and misrepresenting it.

Finally, Prof Caroll claims that most copyright assignments refer to assignment of the article and more. Not in my experience as an author of hundreds of articles, and being in charge of assignments and licences in the 12 years I worked for scholarly publishers. So I further invite Prof Carroll to give me actual examples of such assignment wording. I asked Kevin, but he did not reply....

Jan Velterop said...

We should realise that in the context of scholarly publishing, the transfer of © has nothing to do with making scholarly research results more widely available or with being an incentive to researchers to create more publications describing scientific discovery or knowledge for the good of science or society at large. It has everything to do with the need and desire of publishers to sell scholarly content and © transfer by the author serves as a means of payment to the publisher so that the latter can actually sell the content.

This was all well and good in the print era – there was little alternative – but completely anachronistic in the web era. Assuming publishers add value to the publication process, they need and deserve to be paid straightforwardly for adding any value. © transfer is not an appropriate means of payment. The print era necessity of selling content doesn't exist in the web era any longer and the cumbersome way of getting paid by © transfer in order to be able to sell content later, only serves to limit dissemination.

Anonymous said...

Prof. Carroll did not misunderstand or misrepresent you. You just happen to be wrong. You are stuck on a formalistic understanding of what version--what artifact--has copyright protection. Prof. Carrol points out that this is not the issue, that the true issue is whether what you want to do--copy/publish/archive/make publicly available "D"--infringes a copyright held by the publisher. If "D" is substantially similar (US law) to "F", then your use may be infringing on the publisher's copyright in "F". It is completely irrelevant whether "D" was created before or after "F." Once you have assigned the copyright in "F" (absent a license provision to the contrary), you lose the right to do whatever you might otherwise want with "D". If you want to retain control of "D", you'll need to bargain for it as part of the contractual assignment of "F", or you'll need to license back the rights later (i.e., pay for them). The solutions here have to involve either better (more informed) bargaining by authors or judicial recognition of broad fair use for OA.

kg said...

For German law see http://archiv.twoday.net/stories/664972316/ (in German)

(i) In Germany (nor in the US) there is no typographic copyright like in the UK.

(ii) F is a derivative work of D and vice versa. Chronology doesn't matter. D and F are expressions of the same work and have the same copyright.

(iii) For the OA movement ("green" direction) it is worthless to insist that D might be not covered of the copyright of F, because it is recommended to self-archive the final draft (F).

(iv) According German law it depends whether the contract speaks about dervative use rights.

(v) As long as the publisher has exlusive copy rights (Vervielfältigungsrecht) - this is mainly the case according German law for 1 year! - he can forbidd posting F or D or other derivative versions.

Laurence Bebbington said...

In my view the position stated by Charles that an assignment of copyright in the latest version of a work cannot be taken to include transferring rights in the earlier works is wrong both in UK law and US law for the reasons given by Professor Carroll, and Kevin Smith at Duke's whose recent posts have rekindled this debate.

It is normal to give legal authority for adopting a view on a legal problem, or to reason directly from what the legislation actually says and means. Kevin and Professor Carroll have done so.

In the UK, Metzler & Co (1920) Ltd v Curwen (J.) & Sons Ltd Mac.C.C. 127 is authority for establishing that an assignment of copyright in a later work must generally be taken to include an assignment of the copyright in the preceding works to the extent that they are found in the final work. In referring to Metzler, the leading practitioner text on UK copyright law Copinger and Skone James (16th ed. Section 5-93: Separate assignments of derivative works) states:

“Where the first assignment in time is of the later work, then…prima facie an assignee of the copyright in the later work can maintain an action for infringement against a person using the first work to the extent it can be said to reproduce the later work. This is on the basis that the assignment of the second work must generally be taken to include an assignment of the copyright in the preceding works to the extent that they are found in the final work. Otherwise the assignment would be valueless if confined only to what was new in the final work.”

This is basically the position adopted by Professor Carroll and Kevin.

An assignment is a contract. If a publisher is happy to accept a clause in the contract which licenses back to the author the right to use a previous version – then so be it. Under the doctrine of freedom of contract both parties are free to agree to that. But the very fact that this has to be done by means of this back licence merely strengthens the case for saying that the basic position in law is that where the works are basically the same, then an assignment of copyright in the final version will generally operate, as a matter of law, to transfer rights in the earlier ones.

As Kevin points out in the latest post on his blog the current debate “does not necessarily have any significant impact on what we do to enhance and encourage self-archiving on our campuses.”

Those who say that different drafts automatically acquire different copyrights which can be treated or disposed of separately need to produce legal authority which challenge Metzler; or (given that it is a fairly old precedent) they need to provide convincing argument as to why Metzler would not be applied today.

Richard Poynder said...

Kevin Smith's three posts on this topic can be read here, here, and here

Also of relevance to this discussion is a blog post by Nancy Sims here.

Richard Poynder said...

Michael Carroll asked me to post this response to Charles Oppenheim's comment above:

I'm afraid there's an option (d), which is that I did read Professor Oppenheim's post, and I think it misstates the legal situation in both the United States and the United Kingdom when an author assigns the exclusive rights under copyright in the final version of an article. Specifically, the statement that an assignment of rights in the final version of an article leaves the author with rights in the draft, i.e., "This crucial difference means I am free to do anything I like with D, including posting it on an OA repository." I am not aware of any legal authority that would support this understanding of the situation.

I take no pleasure in pointing this out, since Professor Oppenheim has done a lot of good work on behalf of OA. But, it's not helpful to understate the legal consequences of a copyright assignment or the grant of an exclusive license. I realize that Professor Oppenheim did not say that a publisher could do as I suggest, and I apologize for using a common form of argument among attorneys to point out the consequences of one's statement of the law. Let me rephrase to say, if Professor Oppenheim's statement of the law were correct . . . .

With respect to Kevin Smith's and Dave Hansen's posts, I think we're all roughly in agreement, but we're deeper in the weeds than we need to be. Dave, this really is about the law and not about contract interpretation because we are talking about the legal consequences of a standard transfer of the exclusive right of reproduction, which is ubiquitous in publication agreements that require either an assignment or an exclusive license. See for example, the IEEE agreement as an example.

But, let's assume for a moment that the publication agreement assigns only the exclusive rights in the final version of the article. The fundamental misunderstanding is about what the author has given up. The author initially owns the exclusive rights under copyright in the work of authorship, which is the author's original expression contained in the final article. This means the author has the legal authority to exclude others from making exact copies or copies of the work that are substantially similar to the original. When the author transfers this exclusive right to the publisher, the author now has the legal status of any other member of the public with respect to the final version of the article. [Let's leave aside for the moment any provisions of the contract that may give the author rights to post a draft online.] Assuming that fair use, fair dealing or other user rights do not apply, posting a substantially similar version of the final article online would infringe the publisher's right of reproduction regardless of whether the person posting is the author or any other member of the public and whether the substantially similar version is a prior draft or a variation created after publication. That's what it means to give up your rights under copyright, and that's why the record label was able to make a claim against John Fogerty (referenced in my 2006 post) for allegedly infringing the rights in a song that he had previously written but to which he no longer owned the copyright.

It is simply not the law that an author who has transferred the exclusive rights under copyright in the final version of an article still owns some residual rights in a prior draft that would allow the author to post it online if the draft and the final version are substantially similar to each other. And, the courts have defined the zone of substantial similarity to be fairly broad. Even if only half of the draft corresponds word-for-word with the final version, that would be substantially similar and could not be posted.

I imagine this thread will continue, but I don't expect to contribute further.

Richard Poynder said...

The post by Dave Hansen that Michael Carroll mentions can be read here.

Charles Oppenheim said...

In reply to "Anonymous", Professor Carroll has withdrawn his misinterpretation of what I said. I thank him for that and suggest that you reconsider your comment.

He and Bebbington argue that if, say, 50% of words in a Final Article (F) also appear in a draft (D), then assignment has been given in D as well to the publisher. So how come so many theses are available in OA repositories when I suspect in many cases, at least 50% of the words in an article based on a doctoral thesis also appear in the thesis? This says institutions that post theses into a repository are infringing the publisher's rights. Do Prof Carroll and Mr Bebbington agree that such theses should be withdrawn because of the risk of an infringement action? And what if they have already been made available under a CC licence?

We differ to a degree only. In my view, if the peer reviewers request only trivial changes ("The correct year for Smith's reference in 2005, not 2006"), then D is sufficiently similar to F to not enjoy separate copyright owned by the author. But if it has major changes ("you have completely missed Jones' major work on the topic and need to refer to it at length in your introduction", "you must explain how you chose the sample", "you must justify in rigorous way why you think the correlation means a causal relationship", then F is sufficiently different from D to justify my position (and indeed, explains why some peer reviewers are arguing that in such cases they should be identified as joint authors of F, but of course they wouldn't dream of arguing that they are joint authors of D).

Note how the authoritative "Clark's Publishing Agreements" model academic journal assignment states that the assignment relates to the article as published by the journal and to nothing else.

But what really disappoints me about this debate is the failure to address the pragmatics, which is what I am primarily interested in:

1. No publisher has ever claimed that the Harnad-Oppenheim (HO) solution is illegal. Instead, they have (rightly) said it is impractical - which, by implication, means they accept it may well be legal. Publishers have access to some pretty good IP lawyers…..

2. I don't recommend the HO solution for the reasons publishers gave. I regard the solution as a desperate poor quality last resort. There are much better ways to ensure OA thrives in the scholarly environment.

3. I hope Prof Carroll and Mr Bebbington agree with me that authors are foolish to assign (or sign an exclusive licence) copyright to publishers and that they should adopt one of the tactics I mentioned, or Mike Taylor's idea of putting F into an OA repository and then telling the publisher they cannot sign the assignment/exclusive licence because F is already out there.

Laurence Bebbington said...

I did not enter this debate to agree or disagree, nor please or disappoint Professor Oppenheim on any OA “pragmatics” which concern him.

I did so to offer what I think is an accurate statement on the narrow point of the actual law on the original post and I have offered the relevant authority for my position (which has not been rebutted). To quote exactly from Metzler:

“…when an author writes two works the second of which is derivative from and reproduces original matter contained in the first, and then assigns the copyright in his second work nominatim, he thereby assigns the copyright not only in his second work but in every such part of his first work as was incorporated in his second work.”

The legal position cannot be clearer.

There is nothing new in Professor Oppenheim’s latest reply and it contains nothing which is not covered by reading and reasoning from the previous posts and the above quotation which deal with the actual legal issues.

People can read the posts and make up their own minds. I’ve stated my own position, it’s supported by legal authority and that’s where I’m happy to leave my contribution on this.