ContentMine is going to mine the whole scholarly literature (10,000 articles every day). We’d hoped to do this some months ago and one of the reasons is the massive pushback from major publishers. Technically, legally , politically.

UK government note: You are about to spend about 40 M GBP each year with Elsevier. The real costs are about 2.5 M GBP according to Bjoern Brembs. A significant amount of the rest (even after the huge profit of ca 38% (yes!)) is spent on lobbyists, reps , lawyers, firewalls, captchas, etc. Much of their time is spent trying to make it as difficult as possible to create the Scholarly Commons [1] where we can read, use and re-use the literature without constantly looking up to worry about publishers.

So one of the aspects is legal agreements. We need legal agreement in all sorts of areas, buying houses, hiring staff, etc. These are often between two parties and they negotiate (e.g. on price and exactly what is included) and most of the time it’s relatively understood what the bargain is.

But not with Elsevier. Elsevier produce devious, complex, bespoke legal agreements unlike any other publisher. They neve use a standard form if they can complicate and mislead. You may think I’m being unfair and biassed, but I have spent many days challenging them over text and data mining. (TDM). They put in specific restrictions and clauses about what they hold onto. Despite the fact that it’s legal in UK, they try to persuade you that you have to make a separate agreement with them (an API). You don’t. It’s legal, probably, but it’s immoral and unethical.

Here’s the most recent unpleasantness. A common way to publish your work as Open Access is to pay the publisher (often a lot of money) to allow you to use a CC BY licence. And you retain all rights as author. Straightforward publishers like BMC have done this for 10 years and I have published with them perfectly happily

So when you hear that Elsevier’s licence is CC BY you think fine, I continue to own the paper and Elsevier have a non-exclusive right to use it.

But no. Elsevier has written weasel words into the small print. You no longer own the paper. It may be CC BY but it’s Elsevier’s. And the weasel words are there to look like you are getting what you paid for, but actually you have to be a lawyer to be sure that you have actually been fooled.

Does this matter? At first sight not. And if you trust Elsevier ,  maybe not. But I don’t, and nor does Heather Morrison and nor does Michael Eisen. So let’s listen to them:

Here’s one of the ubiquitous Elsevier staff trying to convince Michael, and here’s Michael’s repsonse. I’ll leave it there , the TL;DR is that this contract is misleading and should be rejected. Michael calls it “despicable”. I wish that Universities treated licences as serious and challenged them rather than letting Michael, Mike Taylor, Heather, Charles Oppenheim, Ross Mounce, me, etc. to to their work voluntarily. After all it’s the Universities who contract with the publishers, and they just don’t seem to care whether their money is well spent.

Read the following from MikeE. If you teach law students, set it as an exercise to pick holes in…

<quote from=”mikeEisen” >

http://www.michaeleisen.org/blog/?p=1915
Elsevier is tricking authors into surrendering their rights
By MICHAEL EISEN | Published: MAY 24, 2016
A recent post on the GOAL mailing list by Heather Morrison alerted me to the following sneaky aspect of Elsevier’s “open access” publishing practices.

To put it simply, Elsevier have distorted the widely recognized concept of open access, in which authors retain copyright in their work and give others permission to reuse it, and where publishers are a vehicle authors use to distribute their work, into “Elsevier access” in which Elsevier, and not authors, retain all rights not granted by the license. As a result, despite highlighting the “fact” that authors retain copyright, they have ceded all decisions about how their work is used, if and when to pursue legal action for misuse of their work and, crucially, if they use a non-commercial license they are making Elsevier is the sole beneficiary of commercial reuse of their “open access” content.

For some historical context, when PLOS and BioMed Central launched open access journals over a decade ago, they adopted the use of Creative Commons licenses in which authors retain copyright in their work, but grant in advance the right for others to republish and use that work subject to restrictions that differ according to the license used. PLOS and BMC and most true open access publishers use the CC-BY license, whose only condition is that any reuse must be accompanied by proper attribution.

When PLOS, BioMed Central and other true open access publishers began to enjoy financial success, established subscription publishers like Elsevier began to see a business opportunity in open access publishing, and began offering a variety of “open access” options, where authors pay an article-processing charge in order to make their work available under one of several licenses. The license choices at Elsevier include CC-BY, but also CC-BY-NC (which does not allow commercial reuse) and a bespoke Elsevier license that is even more limiting (nobody else can reuse or redistribute these works).

At PLOS, authors do not need to transfer any rights to the publisher, since the agreement of authors to license their work under CC-BY grants PLOS (and anyone else) all the rights they need to publish the work. However, this is not true with more restrictive licenses like CC-BY-NC, which, by itself, does not give Elsevier the right to publish works. Thus, Elsevier if either CC-BY-NC or Elsevier’s own license are used, the authors have to grant publishing rights to Elsevier.

However, as Morrison points out, the publishing agreement that Elsevier open access authors sign is far more restrictive. Instead of just granting Elsevier the right to publish their work:

Authors sign an exclusive license agreement, where authors have copyright but license exclusive rights in their article to the publisher**.

**This includes the right for the publisher to make and authorize commercial use, please see “Rights granted to Elsevier” for more details.

(Text from Elsevier’s page on Copyright).

This is not a subtle distinction. Elsevier and other publishers that offer it routinely push CC-BY-NC to authors under the premise that they don’t want to allow people to use their work for commercial purposes without their permission. Normally this would be the case with a work licensed under CC-BY-NC. But because exclusive rights to publish works licensed with CC-BY-NC are transferred to Elsevier, the company, and not the authors, are the ones who determine what commercial reuse is permissible. And, of course, it is Elsevier who profit from granting these rights.

It’s bad enough that Elsevier plays on misplaced fears of commercial reuse to convince authors not to grant the right to commercial reuse, which violates the spirit and goals of open access. But to convince people that they should retain the right to veto commercial reuses of their work, and then seize all those rights for themselves, is despicable.

– See more at: http://www.michaeleisen.org/blog/?p=1915#sthash.hvxs8VLo.V4v0Dea6.dpuf

[1] Maryann Martone’s phrase.

TL;DR. I continue to try to get public data out of Elsevier. I think I should be able to – every other publisher has no problem. After some not-very-useful replies Elsevier simply give up answering me.

Over the last 7-8 years I have had major issues with Elsevier on many aspects – licensing, paywalls , availability etc.  It’s normally impossible to find anyone who gives me a straight answer. I believe that a modern company should have a clear channel of communication – where requests are handled formally and there is accountability when things go wrong.

Indeed some do. Cambridge and Oxford University Presses. They are parts of the Universities and so have to abide by Freedom Of Information request rules – give clear public answers to  questions within a given period of time (20 working days). – and I have used this. By contrast Elsevier find many ways of not answering questions.

So I try again – and I leave it to you to decide whether this is a company that the UK should give 40 M GBP of taxpayers’ money to. When I go to meetings about scholarly publishing and informatics there are increasingly representatives from Elsevier who mingle with the other delegates.  They are “friendly” and “want to help us”. So here’s the first story – there may be more. Always remember that we are paying them.

Background: Everyone (except Elsevier) thinks that non-sensitive scientific data accompanying an article should be in the public domain. This is critical because the data:

• is there to support the claims in the article.
• can be re-used by others for many purposes (data-driven science, deriving parameters, aggregation, simulation – a huge list). I have spent much of my scientific career re-using public data.

I’m going to take crystallographic data – my field, but also central to much modern science. Its publication has been supported by International Unions, CODATA, and many other respected scientific bodies.

And almost all publishers make it public. American Chemical Society, Royal Society of Chemistry, Acta Crystallographica and many more.

But not Elsevier. They either hide it behind a paywall, or send it to the Cambridge Crystallographic Data centre, who provide it under a subscription licence (a trivial amount – probably < 1% is available for free, but NOT for re-use). I wrote to the “Director of Universal Access” some years ago and got waffle.

I and many others think this is outrageous. It’s public data, not Elsevier’s . The science in the paper is seriously diminished without the data. I help run the Crystallography Open Database (COD) which has hundreds of thousands of structures. Will Elsevier give these back to the public?

The only route that I have are the “helpful” reps I meet. So a month ago I met one. He agreed to take my concern into Elsevier. At least I would get a clear answer…

I am including all the letters. I have removed the name of the Elsevier rep.

[1,2,3] TL;DR he agreed to take something on. It wasn’t his department. He’s sent it off.

[4,5] He discovers that authors can send their files to [1] CCDC or [2] Elsevier (behind the paywall). This defines the scope of the question. (It doesn’t tell me anything I didn’t already know). Most of the data are in the second category.

[6] PMR reiterates that by hiding data Elsevier is going against all other responsible parties in the field.

[7] Elsevier replies that they assume I only want data of type 2 and that they can make them available “openly” behind a Mendeley login.

[8] PMR replies that we want all the data (this is consistent with every other publisher) and that data behind a Mendeley login are not open. PMR lists 6 questions that he would like answered.

I have not had the courtesy of a reply even after 19 days, so I can only assume Elsevier regard me as not worth continuing to answer.

==================================

[1]  PMR 2016-04-19
I thank you for our conversation yesterday, where you agreed that all factual supplemental crystallographic data published with papers in Elsevier journals should be made available without restrictions (effectively CC0). You agreed that you would work with your technical colleagues to see how this could be done as soon as possible (“flipping a switch”). You agreed that this would bring Elsevier into line with most other major publishers (ACS, RSC, IUCr, Nature) who have for many years released all their crystallographic data (“CIF”s) into public view on their websites, without restrictions. This data would include both current published data (probably back to about 1990) and all future data.

The Crystallography Open Database (COD) (PMR is a board member) has a 10-year record of accepting, validating CIFs from all domains (organic, organometallic, inorganic, metals and alloys) and then offering to the world for re-use under effectively CC0 licence. It also provides a variety of modern search and analysis software.

I ask that you commit to this publicly now and am confident that COD will be willing to host the data if Elsevier does not wish to mount them on its web pages.

[2] Elsevier 2016-04-19 =======

To be very clear – what we agreed was that I would look into this and get back to you with a clear response one way or the other. I made no commitment as this is not in my area of responsibility. I’d appreciate in the interests of establishing trust between us that we are both careful in reporting our conversations accurately.

Per our conversation I have already reached out to my colleagues to understand the current situation w.r.t. Crystallographic supplemental files in our journals. I will let you know as soon as hear back.

[3] PMR: 2016-04-19 =========

I wanted to let you know that I am making some progress in discussions with internal colleagues w.r.t. how we currently treat CIF files, but I haven’t got fully to the bottom of the story.A couple of facts I have discovered:

• We give authors the choice as to whether they deposit their CIF files with an external database and provide us with a link to the file, or have us host their files as Supplementary Material. You can see examples of the two cases as follows

1. Articles linking out to CCDC using data banner links (e.g. http://www.sciencedirect.com/science/article/pii/S0040402009001021)
2. Articles with CIF files delivered as supplementary material (e.g. http://www.sciencedirect.com/science/article/pii/S2352340915000906)

• The linking arrangement is CCDC is not exclusive – we have a large number of these (see https://www.elsevier.com/books-and-journals/content-innovation/data-base-linking/supported-data-repositories) and I believe one of my colleagues is currently evaluating a request to establish a linking arrangement with COD.
• For articles with CIF files that we host as supplementary material, we are still evaluating both from the technical point of view and the legal point of view the feasibility of making these available openly via our new Mendeley data platform (http://data.mendeley.com).

I will keep you informed as I learn more. However, for confirmation I assume that your main interest is in securing open access to files in category 2 above?

PMR: [6] 2016-05-05 ============

To clarify our relationship. I am acting as a board member of the Crystallography Open Database, a Public Interest Research Organization (PIRO),  and copying them. You are a formal representative of Elsevier. / RELX. I regard our correspondence as in the public interest and intend to publish all of it.

I list below a number of direct, simple questions to which I request answers. These are ones that I would expect organizations subject to Freedom Of Information requests (including, for example OUP and CUP, and myself) to have to answer. Although Elsevier is not subject to FOI I am expect the same comprehensive and clarity of response. There is also a request for crystallographic data.
I have set out our expectations. In summary: All major publishers except Elsevier make their crystallographic data fully and publicly available, effectively CC0. Elsevier’s policy is in direct conflict with national and international science organizations such as CODATA, ICSU and the International Union of Crystallography. Elsevier’s position in withholding scientific data is in direct opposition to the  norms and expectations of the scientific world.

[7] Elsevier 2016-05-05 ==========

I wanted to let you know that I am making some progress in discussions with internal colleagues w.r.t. how we currently treat CIF files, but I haven’t got fully to the bottom of the story.

1. Articles linking out to CCDC using data banner links (e.g. http://www.sciencedirect.com/science/article/pii/S0040402009001021)
2. Articles with CIF files delivered as supplementary material (e.g. http://www.sciencedirect.com/science/article/pii/S2352340915000906)

• The linking arrangement is CCDC is not exclusive – we have a large number of these (see https://www.elsevier.com/books-and-journals/content-innovation/data-base-linking/supported-data-repositories) and I believe one of my colleagues is currently evaluating a request to establish a linking arrangement with COD.
• For articles with CIF files that we host as supplementary material, we are still evaluating both from the technical point of view and the legal point of view the feasibility of making these available openly via our new Mendeley data platform (http://data.mendeley.com).

I will keep you informed as I learn more. However, for confirmation I assume that your main interest is in securing open access to files in category 2 above?

[8] PMR 2016-05-05 ============

Thank you. I have allocated you the same length of time (one month/20 working days) as for a UK FOI request to provide information. If you are also, as I expect, working towards a change in Elsevier policy and practice, then it will be necessary at the end of the month to detail what you have set in motion and with what expected timescale. Until 21st May I accept that I will not make our discussions public.

>ELS> For articles with CIF files that we host as supplementary material, we are still evaluating both from the technical point of view and the legal point of view the feasibility of making these available openly via our new Mendeley data platform (http://data.mendeley.com).

The word “openly” is imprecise.  I note that Mendeley requires a login which is inconsistent with Openness. I assume therefore that Mendeley will impose its own terms and conditions, which by definition will be inconsistent with CC0.

Note that under the new 2014 UK exception to Copyright I can legally mine the data associated with any Elsevier publication that I have the right to read. Since the data itself is uncopyrightable, and since a journal is not a database covered by European sui generis database rights,  I can therefore download all CIFs as part of my personal non-commercial research and I can publish the data from that research. There is no benefit in using Mendeley. From Elsevier’s point of view it would possibly be preferable to bundle this historic data now and ship it to COD and we would be happy to make this technically possible. Otherwise I shall extract it under the UK law.

>ELS> I will keep you informed as I learn more. However, for confirmation I assume that your main interest is in securing open access to files in category 2 above?

>PMR> **NO**. Our interest is in all supporting ALL crystallographic data that is associated with ALL Elsevier publications, in line with all other major publishers.

I would therefore like answers to the following questions and will publish answers when the month is up. I have framed them so that many can be answered with Yes/No/DeclineToAnswer/. I use the phrase “NonOpen database) to refer to databases such as those run by CCDC, ICSD and other organizations which do not make the total data available under CC0.  (Note that if these questions were submitted to OUP I would expect them all to be fully answered under FOI).

1. Does Elsevier hold copies of ALL raw CIFs associated with Elsevier publications, or if not can it obtain these CIFs?
2. Please provide a complete list of all NonOpen databases that Elsevier requires or allows authors to submit crystallographic data to. Please indicate whether Elsevier has the right to obtain ALL the crystallographic data in BULK associated with their publications.3. Does Elsevier have formal contractual relations with these NonOpen databases. Please indicate what these contracts allow and forbid.
3. Please indicate how Elsevier decides on its policy on crystallographic data. Does it consult with ICSU, CODATA or IUCr? When was the policy last reviewed? What is the mechanism for PIROs to formally request changes in policy?
4. Please provide a list of all files of Type 1 and Type 2 and a service where updates of these lists can be obtained.

These are requests for information.

Our request for crystallographic data, which is consistent will all major scientific bodies, funding bodies and all major publishers other than yourselves, for the files themselves is:

5.Please provide all files of type 1 and 2 , or an open mechanism (e.g. an API) where all these files can be obtained. Please confirm that redistribution of the files is permitted without further permission.

6. Please indicate that Elsevier is committing to changing the policy to make supplemental data files publicly, freely and openly available. Please indicate the process that has been initiated and how it will report back to the world.

I will publish the correspondence, unedited, on 21st May.

=====================

19 days have elapsed without the courtesy of a reply

When I go the airport by taxi [1] I try to get the same taxi driver, Ken [2]. Ken is a shining example of why every citizen of the world needs access to the whole scholarly literature – open and for free.

You often hear publishers (and some academics) say “ordinary people wouldn’t understand the science”.  This is appallingly arrogant , and blatantly untrue.   In the taxi we are discussing whether people listen more to scientists from Cambridge than less-well-known universities:

PMR: Doctors in a Western Australian hospital struggled for many years to convince the medical profession of the true cause of stomach ulcers…

KEN: You mean Campylobacter.

This is the point. Ken has no University education. But he knows the cause of ulcers and he knows the precise scientific name [3] (Read the story of Barry_Marshall and Robin Warren; everyone should be able to follow it. They published their results [4] in The Lancet a well-known medical journal.)

Oh, dear, Ken. I’m sorry you can’t read this unless classic paper unless you fork out 36 USD – and you would then have just 24 hours to read it. And you can’t show it to your mates – that’s copyright violation. Oh, and all the money goes to Elsevier – none to the authors.

Taxi drivers are an underclass. Only academics in Cambridge are allowed to read about Helicobacter. What? It was published in 1983? Yes,  that’s far too recent to make it Open and Free for taxi-drivers. One of the most important papers in science? won a Nobel prize? You expect taxi-driver tax-payers to be allowed to read work they fund??  Sorry.  Just keep driving taxis.

It’s a moral imperative to publish science for everyone. Not just academics but also taxi drivers. Next time you are in a taxi, don’t sit back but ask your driver: “are you interested in science?”. Not everyone is, but everyone who is interested in science can be a scientist. It’s a matter of attitude and philosophy, not a white coat.

PMR: So we had a meeting last week to discuss negotiations with Elsevier.

KEN: Elsevier , the publisher?… (Ken is interested in politics , and science/Cambridge. He knows about Elsevier.)

PMR Yes…

And I continue to set the scene:

A week ago a small selected group of concerned Cambridge academics (including PMR), and library staff, met with Jisc (who are advising HEFCE – who fund English universities), to find out about Jisc’s negotiations with Elsevier about university subscriptions.  Almost 40M GBP  year  of taxpayer and student money for academics to read journals. Until now I didn’t even realise there was a negotiation – it has been kept very quiet indeed. The deal has to be concluded by 24:00 2016-12-31 ; if not the subscriptions are cancelled and even Cambridge academics won’t be able to read The Lancet.  (The journal which Ken still can’t anyway read, thus bringing Cambridge academics and taxi-drivers even closer together).

The meeting opened my eyes to the massive and visceral resistance that Elsevier was putting up against any normal “negotiations”. (I’ve negotiated with equipment suppliers before – one deal was >2M GBP in today’s money.) What came over to me very clearly was that this is not about price. It’s a battle for control. Who makes the decisions about the dissemination of scholarly knowledge, whether or not paid for by the taxpayers and student fees?

Elsevier (along with Digital Science from Nature/Springer, etc.) are rapidly taking over our academic infrastructure. Last week Elsevier they bought SSRN, the social sciences repository. They now control preprints in significant part of academia. They are selling the Universities PURE – a system of repositories that are under Elsevier control and where I fear the we are the product as well as the customers. What does Open matter if the gateway is controlled by a publisher who Openwashes the language to legitimize its control?

It’s not because Elsevier are the biggest, it’s because they are the most ruthless, arrogant, publisher. I have been dealing with them for ca 8 years. They treat me as a nobody – a nuisance. I’m far from the only one. It’s critical that we wrest back control. After all it’s us who are paying the money. And it’s every taxi driver in the world who ultimately suffers.

The UK is not the first country to negotiate with Elsevier. Last year (2015) the Dutch did. They announced that unless they got a set of nonnegotiable demands they would walk away from Elsevier and cancel subscriptions.

What actually happened?

I don’t know. The Dutch agreed something with Elsevier. What? Don’t know , because Elsevier requires secrecy and the Dutch agreed. You can read reports before the deal and after the deal. Maybe, Ken, you can make more of them than me.
The worst possible thing would be an announcement on 2017-01-01:

“The UK and Elsevier have concluded a 40 M agreement about purchase of Elsevier publications and services. The details are commercially sensitive but [some important person] says: ‘This is a good deal for the UK …'”

The thinking world will say: “The Dutch stuck their heels in a bit but finally Elsevier won. The same has happened in UK”.

The unthinking will say: “I didn’t even realise that UK and Elsevier were negotiating. Ah well, I don’t read the scientific literature because I can’t afford it.”

KEN: had already said – his words – “The theft of Knowledge“.

This has to be fought in public, starting now. Elsevier are effectively a monopoly. The people of Europe – 250,000 – fought software patents and won. The people of many countries took on Microsoft and won. The people of UK, and later Europe, should take on Elsevier and win. It won’t be easy but it has to be done.

And if the taxi drivers take to the streets it can happen.

And in case you are wondering why we are paying 40 Million GBP for electronic content – which Elsevier neither authors nor referees , here is Bjoern Brembs  with the breakdown of costs.  Bjoern – a neuroscientist – asks  Why haven’t we already canceled all subscriptions?

The question in the title is serious: of the ~US$10 billion we collectively pay publishers annually world-wide to hide publicly funded research behind paywalls, we already know that only between 200-800 million go towards actual costs. The rest goes towards profits (~3-4 billion) and paywalls/other inefficiencies (~5 billion).

So the UK will be paying 20M GBP to Elsevier for gross inefficiencies (which from my own experience I can confirm) and technology to stop Ken, and Chris Hartgerink (“Elsevier stopped me doing my research”) , reading science.

And we should pay them just 3M GBP at the most.

Or, much better, do it ourselves. We’d do it cheaper, better, faster. Bjoern, a senior scientist, says so and I agree.

=========================