How tightly enforced are open-access embargoes?

Question

Many academic journals have copyright policies which forbid authors (possibly for a finite embargo time) to make their papers (either preprints, accepted manuscripts, or camera-ready versions) freely available (either on their personal websites or on repositories such as the arXiv). How tightly enforced are these policies? Are there known cases of publishers pursuing legal action against an author for posting copyrighted academic papers? Or would such cases normally be dealt with private requests to cease-and-desist?

This answer seems to indicate such cases are rare, but there could be privately-dealt with cases that are not visible. Or is there a large body of public-repository-published papers that possibly / probably / demonstrably have been publicly posted in breach of a copyright policy?

Answer 1

One data point:

I am a mathematician, and I have never heard of this happening to anyone.

Many mathematicians post preprints on their website and to the arXiv, and it is somewhat common for mathematicians to also post scanned copies of the published journal versions of their older papers.

Ethical judgements may vary, but I am unaware of any authors who have suffered negative practical consequences for this, or even who have been asked to remove their papers.

Answer 2

Update: Here's the new text of the German UrhG §38 Abs (4):

Der Urheber eines wissenschaftlichen Beitrags, der im Rahmen einer mindestens zur Hälfte mit öffentlichen Mitteln geförderten Forschungstätigkeit entstanden und in einer periodisch mindestens zweimal jährlich erscheinenden Sammlung erschienen ist, hat auch dann, wenn er dem Verleger oder Herausgeber ein ausschließliches Nutzungsrecht eingeräumt hat, das Recht, den Beitrag nach Ablauf von zwölf Monaten seit der Erstveröffentlichung in der akzeptierten Manuskriptversion öffentlich zugänglich zu machen, soweit dies keinem gewerblichen Zweck dient. Die Quelle der Erstveröffentlichung ist anzugeben. Eine zum Nachteil des Urhebers abweichende Vereinbarung ist unwirksam.

rough translation:

The author of a scientific work

• which was produced during scientific reseach that funded at least half by public money and
• which was published in a periodical that is issued at least twice per year

retains the right to make the accpted manuscript publicly available

• after an embargo period of 12 months from the date of the first publication
• as long as this is not for commercial purposes.

The source of the first publication must be given.
This right is retained also in case of a complete copyright transfer to the publisher of editor.
Agreements on this topic to the disadvantage of the author are void.

I guess the non-commercial clause will raise some questions and difficulties.

---

old answer:

Here in Germany, the parliament (Bundestag) actually voted to change §38 UrhG
to (among other changes):

Einführung eines unabdingbaren Zweitverwertungsrechtes für Autoren von mit überwiegend aus öffentlichen Mitteln geförderten Beiträgen in Periodika 12 Monate nach Erstveröffentlichung;

rough translation: introduction of an unalienable right of secondary use* for authors of contributions to journals which are mainly supported mainly by public grants 12 months after the primary publication.

* I'm not quite sure how to translate "Zweitverwertungsrecht" - it is the right for secondary use/exploitation(?) of a work. Meaning that agreements that the publisher get the exclusive rights to the work will be valid in Germany only for 12 months, thereafter the authors have the right to make these papers publicly available.

The new text does not yet show up (the voting took place only 2 weeks ago) in the law texts in internet, it will become §38 (4) UrhG.

While this still means that only papers with public funding are covered, and the embargo period may be annoying, I see this as an important step into the right direction. And it definitively means that there won't be any possibility for publishers to enforce anything after a year if a German author made the work publicly available.

In any case, our version of the "fair use" rights mean that I'm always allowed to send single copies of scientific papers which are needed for collaboration to my colleagues. So within the first year after publication, you'll still have to email me for the paper, but thereafter, you can blame me for not making the manuscript available e.g. via arXiv.

Answer 3

This seems to be changing. A recent article in the Washington Post, How one publisher is stopping academics from sharing their research (2013-12-19), describes a recent 'spree' of takedown notices sent by Elsevier to Academia.edu, the University of Calgary, the University of California-Irvine, and Harvard University. This seems to be a new development:

The letter to Harvard identified 23 articles that it requested be taken down from University-hosted pages in mid November. "We had not received takedown notices for scholarly articles before this, as far as we know," says Peter Suber, the Director of the Harvard Office for Scholarly Communication and the Harvard Open Access Project.

Academia.edu, which is a for-profit company, and also a competitor of Elsevier-owned Mendeley, got about 2,800 requests.

This story is also covered by Wired and The Chronicle of Higher Education.

Answer 4

Here is one event that can qualify as such. In 2011, Russell O'Connor submitted a paper to the ACM Workshop for Generic Programming 2011 after having submitted it to the arXiv under a Public Domain licence. At press time, the printer asked O'Connor for a letter of permission from the ACM in lieu of an exclusive copyright transfer which he could not grant. Although he had informed the conference chairs of the arXiv upload when he submitted, and they were OK with it, ACM legal refused to publish the paper as it was "already published." The paper is only mentioned briefly in the proceedings in a note:

We note that one of the papers presented in the workshop is not included in the proceedings. This paper, ‘Functor is to Lens as Applicative is to Biplate: Introducing Multiplate’ by Russell O’Connor, is accessible as arXiv:1103.2841v2 [cs.PL].

This is explained in more detail in O'Connor's blog post The ACM and Me.

Answer 5

Fully agreeing with eykanal's comment, I can provide the following thought on reasons: as with all copyrighted materials originating from commercial sources, it becomes a business decision on when and how to hunt down those who break the rules. If some authors put their manuscripts out on their own web sites it is probably not worth the expenses to prevent it. If everyone did it systematically, I am sure things would look different. If publishers see a decline in subscription rates from libraries due to the fact that authors make stuff available, then I think we will see something akin to what happened with downloadable music. For a publisher, an author is also what makes the business go around, so hunting authors may not be a good business practice. But, with more and more published work being made available in formally illegal ways, I am sure it is only a matter of time before clamps are tightened. So it is a bit like teasing a sleeping bear, fine until it wakes up (and I do not mean "fine" as in agreeing with posting material illegally).

In addition to putting the published paper online, it might be worth noting that in many if not most cases, it is perfectly legal to post the original manuscript (before review and typesetting) on the web since this is not the copyrighted material that is in the publications. Unless you have signed off on immaterial rights you still have the rights to that original work.

The bottom line is it is illegal and many probably do not really understand what they have signed when they published their papers.