Self-Archiving Mandates

[Forwarding from SPARC via Peter Suber's SPARC OA Forum. Also see the PDF edition.]

An Open Letter to the U.S. Congress
Signed by 33 Nobel Prize Winners
September 9th, 2008

Dear Members of Congress:

As scientists and Nobel Laureates we are writing today to support the NIH Public Access Policy that was instituted earlier this year as a Congressional mandate. This is one of the most important public access initiatives ever undertaken. Finally, scientists, physicians, health care workers, libraries, students, researchers and thousands of academic institutions and companies will have access to the published work of scientists who have been supported by NIH.

For scientists working at the cutting edge of knowledge, it is essential that they have unhindered access to the world's scientific literature. Increasingly, scientists and researchers at all but the most well-financed universities are finding it difficult to pay the escalating costs of subscriptions to the journals that provide their life blood. A major result of the NIH public access initiative is that increasing amounts of scientific knowledge are being made freely available to those who need to use it and through the internet the dissemination of that knowledge is now facile.

The clientele for this knowledge are not just an esoteric group of university scientists and researchers who are pushing forward the frontiers of knowledge. Increasingly, high school students preparing for their science fairs need access to this material so that they too can feel the thrill of research. Teachers preparing courses also need access to the most up-to-date science to augment the inevitably out-of-date textbooks. Most importantly, the lay public wants to know about research findings that may be pertinent to their own health diagnoses and treatment modalities.

The scientific literature is our communal heritage. It has been assembled by the painstaking work of hundreds of thousands of research scientists and the results are essential to the pursuit of science. The research breakthroughs that can lead to new treatments for disease, to better diagnostics or to innovative industrial applications depend completely on access not just to specialized literature, but rather to the complete published literature. A small finding in one field combined with a second finding in some completely unrelated field often triggers that "Eureka" moment that leads to a groundbreaking scientific advance. Public access makes this possible.

The current move by the publishers is wrong. The NIH came through with an enlightened policy that serves the best interest of science, the scientists who practice it, the students who read about it and the taxpayers who pay for it. The legislators who mandated this policy should be applauded and any attempts to weaken or reverse this policy should be halted.

Name, Category of Nobel Prize Awarded, Year

David Baltimore, Physiology or Medicine, 1975
Paul Berg, Chemistry, 1980
Michael Bishop, Physiology or Medicine, 1989
Gunter Blobel, Physiology or Medicine, 1999
Paul Boyer, Chemistry, 1997
Sydney Brenner, Physiology or Medicine, 2002
Mario Cappechi, Physiology or Medicine, 2007
Thomas Cech, Chemistry, 1989
Stanley Cohen, Physiology or Medicine, 1986
Robert Curl, Chemistry, 1996
Johann Deisenhofer, Chemistry, 1988
John Fenn, Chemistry, 2002
Edmond Fischer, Physiology or Medicine, 1992
Paul Greengard, Physiology or Medicine, 2000
Roger Guillemin, Physiology or Medicine, 1977
Leland Hartwell, Physiology or Medicine, 2001
Dudley Herschbach, Chemistry, 1986
Roald Hoffman, Chemistry, 1981
H. Robert Horvitz, Physiology or Medicine, 2002
Roger Kornberg, Chemistry, 2006
Harold Kroto, Chemistry, 1996
Roderick MacKinnon, Chemistry, 2003
Craig Mello, Physiology or Medicine, 2006
Kary Mullis, Chemistry, 1993
Joseph Murray, Physiology or Medicine, 1990
Marshall Nirenberg, Physiology or Medicine, 1968
Paul Nurse, Physiology or Medicine, 2001
Stanley Prusiner, Physiology or Medicine, 1997
Richard Roberts, Physiology or Medicine, 1993
Susumu Tonegawa, Physiology or Medicine, 1987
Hamilton Smith, Physiology or Medicine, 1978
Harold Varmus, Physiology or Medicine, 1989
James Watson, Physiology or Medicine, 1962

Press Contact:
Sir Richard Roberts
(Nobel Prize in Physiology or Medicine,1993)
Tel: (978) 380-7405
Fax: (978) 380-7406
Email: roberts -- neb.com

Fred Friend (JISC) wrote:

"Under your Plan B, what would stop publishers increasing the Closed Access embargo period to two years?"

The question is a good one, a natural one, and a pertinent one:

(1) "Plan B" is a contingency plan, in case the Conyers Bill should defeat the current NIH OA Policy (i.e., "Plan A").

(2) If the Conyers Bill were to pass, not only Plan A but all protection from publisher embargoes would be dead in the water (hence embargoes could in principle be made infinitely long).

(3) Plan B is accordingly designed to free the NIH Mandate from any dependence at all on publishers to decide when research (postprints) may be deposited.

(4) Plan B is to make all research postprints OA as soon as they can be, but to require that the actual deposit of all postprints in authors' Institutional Repositories be made immediately upon acceptance for publication, and to rely on almost-OA (via the semi-automatic "email eprint request" Button) for deposits that for any reason cannot be made OA immediately.

(5) With Immediate-Deposit mandated universally, "Almost-OA" (via the Button) serves research and public needs almost as well as OA during any embargo period.

(6) Universal Deposit mandates, plus the resulting enormous growth in usage and impact via OA and Almost-OA, will make it harder and harder for publishers to justify embargoes, while at the same time making embargoes virtually ineffectual:

(7) Hence embargoes will die their natural and well-deserved deaths once universal Deposit (Plan B) is mandated, by all research institutions and funders, worldwide, paving the way for full, immediate OA.

(8) It is far less clear whether Plan A [Delayed, Post-Embargo Deposit] can or will be universally adopted, by all research institutions and funders, worldwide -- nor whether, even if it were, it would ever lead to immediate-OA.

(9) Even with the Button, Delayed-Post-Embargo Mandates cannot provide immediate almost-OA. (NIH requires immediate "submission" but it is deposited -- in PubMed Central -- only after the embargo.)

(10) Hence it is in fact Plan A that locks in publisher embargoes, not Plan B!

Stevan Harnad
American Scientist Open Access Forum

Peter Suber wrote in OA News:

PS: " I agree that [a Deposit Mandate [DM] plus the "email eprint request" Button] would be a good fallback in the (unlikely) event that the Conyers bill passes. But I can't agree that it would 'hasten universal OA more effectively than the current NIH mandate'"

There are three issues here, not one!

(1) Replacing the current NIH mandate with a DM if the current mandate is defeated.

(2) Adding DM to the current NIH mandate even if it is not defeated.

(3) And the question of what would have been the effect of adopting DM in the first place.

On all three, I think the answer is very definitely that DM (whether added or substituted) would hasten universal OA more effectively. (And I actually think Peter would agree about all three of these; the seeming disagreement may be just a verbal one.)

Note that what I said was that an NIH Deposit Mandate (Immediate Deposit, with optional Closed Access plus the "Almost-OA" Button) would hasten universal OA far more effectively than the current NIH (Delayed OA-Deposit) Mandate.

I said universal OA because I was not referring merely to OA for NIH-funded research, nor to the effect of the NIH mandate only on NIH-funded research:

It is far easier for other funders and institutions to reach agreement on adopting Deposit Mandates of their own (Immediate Deposit of all articles, with the option of Closed Access and the "email eprint request" Button during any publisher embargo) because that completely removes copyright concerns, publishers, and the publisher's lobby from the decision loop. For embargoed articles, DM is merely an internal record-keeping mandate, yet with the Button it can also provide almost-OA -- and will almost certainly lead to full OA once the practice becomes universal.

So the first reason an NIH DM would be (and would have been) more effective is that, being free of legal obstacles, it is universally adoptable. Many funders and even institutions have instead copied, or tried to copy, the NIH Embargoed OA Mandate (with deposit mandated only after the publisher OA embargo has elapsed).

These legally encumbered mandates have either failed to be adopted elsewhere altogether, because of unresolved legal concerns (I know of many that have been under debate for years), or they have been adopted, cloning the NIH model, with the loss of the opportunity for Almost-OA during the embargoes (and an uncertainty about whether and when deposit actually takes place).

That represents a (i) a loss of any mandate at all, among the would-be mandates that failed to be adopted because of the avoidable copyright concerns, (ii) a loss of a good deal of Almost-OA during the embargo periods for the adopted mandates, and (iii) continuing delay in reaching universal OA, for which universal deposit is a necessary precondition!

So, yes, the NIH mandate would have been more effective, both for NIH OA and for universal OA, if it had been a DM: It would have generated immediate Almost-OA for NIH and more DMs worldwide. DM can still be added to the NIH mandate now, whether or not the Conyers Bill passes. If Conyers fails, that will make NIH an Immediate DM plus Embargoed OA Mandate -- which is much better than just an Embargoed OA Mandate. If Conyers passes, then it will make NIH just an Immediate DM, which is still better than no mandate for NIH, still provides Almost-OA during any embargo, and is far more conducive to consensus for universal adoption.

PS: " The NIH mandate provides (or will soon provide) OA to 100% of NIH-funded research, not OA to 63% and almost-OA to 37%"

If Conyers passes, then (as Peter agrees), DM is the right Plan B. But even if Conyers fails, why not add DM to the current NIH embargoed OA mandate, and have at least embargoed OA for 100% of NIH-funded research plus immediate almost-OA during the embargo (and a better mandate model for universal adoption)?

PS: " [Nor can I agree that] "there is no way to stop [an NIH immediate DM plus the Button] legally". ...if Congress wanted to, it could block closed-access deposits too."

Of course Congress can vote into law anything that the Supreme Court does not rule unconstitutional and the President does not veto. But it does seem a bit far-fetched to imagine that Congress would make a law to the effect that NIH is forbidden to require the deposit of a copy of the publications it funds, for internal record-keeping purposes. (And such a bizarre law would be sui generis, nothing to do with copyright.)

Nor does it seem likely that Congress would make a law that US researchers could no longer send reprints of their research to researchers requesting them, as researchers worldwide have been doing for a half century. (And this too would be an ad hoc law, though, at a stretch, it could be portrayed as a curb on Fair Use.)

PS: "However, I would like to see the NIH add the email request button even if the Conyers bill goes down in flames. Then the policy would provide embargoed OA to 100% of NIH-funded research, and almost-OA during the embargo period."

I think we are in complete agreement! Substitute DM + Button for the current NIH Mandate if Conyers fails, and add DM + Button even if it succeeds. (But that doesn't just mean the Button: It means adding the requirement to deposit immediately upon acceptance for publication.)

The only thing missing from this agreement is the one additional (but crucial) further facilitator of universal adoption of DMs: NIH should specify that its preferred mode of deposit is to deposit the postprint in the author's own Institutional Repository (if there is one) and then to port it automatically to NIH from there, via the SWORD protocol:

"One Small Step for NIH, One Giant Leap for Mankind"

Stevan Harnad
American Scientist Open Access Forum

    Re: Conyers Bill:  "Bill Would Block NIH Public Access Policy"
    (Science, 11 September)

Plan B for NIH Public Access Mandate: A Deposit Mandate
(And It's Stronger Than Plan A!)

Let us hope that the Conyers Bill, resulting from the publisher lobby's attempt to overturn the NIH Public Access Mandate, will not succeed. 

But in case it does, I would like to recommend making a small but far-reaching modification in the NIH mandate and its implementation that will effectively immunize it against any further publisher attempts to overturn it on legal grounds. And this Plan B will actually help hasten universal OA more effectively than the current NIH mandate:

(1) NIH should mandate deposit of the refereed final draft of all NIH-funded research, immediately upon acceptance for publication.

(2) But access to that deposited draft need only be made Open Access when there is no publisher embargo on making it Open Access; otherwise it may be made Closed Access.

(3) Open Access means that the full text of the deposited draft is freely accessible to anyone, webwide, immediately.

(4) Closed Access means that the full text of the deposited draft is visible and accessible only to the depositor and the depositor's employer and funder, for internal record-keeping and grant-fulfillment purposes. (Publishers have no say whatsoever in institutional and funder internal record-keeping.)

(5) For all deposits, however, both Open Access and Closed Access, the deposited article's metadata (author, title, journal, date. etc.) are Open Access, hence visible and accessible to anyone, webwide. 

(6) Now the essence of this strategy: NIH should also implement the "Email Eprint Request" Button, so that any would-be user, webwide, who reaches a link to a Closed Access article, can insert their email address in a box, indicate that a single copy of the postprint is being requested for research or health purposes, and click.

(7) That eprint request is then emailed by the repository software, automatically and immediately, to the author of the article, who receives an email with a URL in it that the author can then click if they wish to have the repository software automatically email one individual copy of that eprint to that individual requester.

(8) This is not Open Access (OA). But functionally, it is almost-OA. 

(9) Many journals (63%) already endorse immediate OA.

(10) Closed Access plus the Button will further provide almost-OA for the remaining 37%. 

(11) That means an NIH Deposit Mandate guarantees either immediate OA (63%) or almost-OA (37%) for 100% of NIH-funded research, and 100% of research's immediate-access needs are fulfilled, almost-immediately, during any embargo period, worldwide.

(12) In addition, an NIH Deposit Mandate will encourage universities in the US and worldwide to adopt Deposit Mandates too, for all of their research article output, not just NIH-funded biomedical research output. 

(13) The spread of such Deposit Mandates across institutions and funders globally will inevitably lead to a transition to universal OA for all research output as a natural matter of course, as the mandates ensure the universal practice of immediate deposit, and the resultant pressure for the benefits of instant access becomes irresistible.

(14) Lastly, to make the almost-OA Button even more powerful and easier to implement, NIH should stipulate that the preferred locus of deposit is the author's own Institutional Repository, which can then export the deposit to PubMed Central using the automatized SWORD protocol.

The fact is (and everyone will see this clearly in hindsight) that, all along, the online medium itself has made OA a foregone conclusion for research publications. There is no way to stop it legally. 

It is only technological short-sightedness that is making publishers and OA advocates alike imagine that the outcome is somehow a matter of law and legislation. It is not, and never has been. 

It is only because we have been taking an obsolete, paper-based view of it all that we have not realized that when authors wish it to be so, the Web itself has made it no longer possible to prevent researchers from freely distributing their own research findings, one way or the other. There is no law against an author giving away individual copies of his own writing for research purposes. Researchers have been doing it -- by mailing individual reprints to individual requesters -- for at least a half century.

And NIH need only mandate that authors deposit their (published research journal) writings: giving them away for free can be left to the individual author. The eventual outcome is obvious, optimal, natural, and inevitable.

I strongly urge OA advocates to united behind this back-up strategy. It will allow us to snatch victory from the jaws of defeat. 

(Thanks to Colin Steele for forwarding this and to Glen Newton for the excerpting.)

---

The Australian minister for Innovation, Industry, Science and Research, Kim Carr spoke about this report in a speech released yesterday. Full text here.

It is embodied in a series of recommendations aimed at unlocking public information and content, including the results of publicly funded research.

The review panel recommends making this material available under a creative commons licence through:

machine searchable repositories, especially for scientific papers and data

and the internet, where it would be freely available to the world.

...The arguments for stepping out first on open access are the same as the arguments for stepping out first on emissions trading – the more willing we are to show leadership on this, we more chance we have of persuading other countries to reciprocate.

This speech reflects a number of recommendations in the report:

Recommendation 7.7: Australia should establish a National Information Strategy to optimise the flow of information in the Australian economy. The fundamental aim of a National Information Strategy should be to: ·utilise the principles of targeted transparency and the development of auditable standards to maximise the flow of information in private markets about product quality; and ·maximise the flow of government generated information, research, and content for the benefit of users (including private sector resellers of information).

Recommendation 7.8: Australian governments should adopt international standards of open publishing as far as possible. Material released for public information by Australian governments should be released under a creative commons licence.

Recommendation 7.9: Funding models and institutional mandates should recognise the research and innovation role and contributions of cultural agencies and institutions responsible for information repositories, physical collections or creative content and fund them accordingly.

Recommendation 7.10: A specific strategy for ensuring the scientific knowledge produced in Australia is placed in machine searchable repositories be developed and implemented using public funding agencies and universities as drivers.

Recommendation 7.14: To the maximum extent practicable, information, research and content funded by Australian governments including national collections should be made freely available over the internet as part of the global public commons. This should be done whilst the Australian Government encourages other countries to reciprocate by making their own contributions to the global digital pubic commons.

---

As ROARMAP indicates, the world leader in Open Access, both in time and in absolute size, is indisputably the United Kingdom: A UK Parliamentary Select Committee was the world's first governmental recommendation to mandate OA, in 2004. However, the UK government, under pressure from the publishing lobby, did not accept its own committee's recommendation. Nevertheless, six of the seven RCUK research funding councils went on to mandate OA anyway. The UK now has a total of 18 university and funder OA mandates (including the world's first OA mandate at Southampton's School of Electronics and Computer Science in 2003).

Australia, however, adopted the world's first university-wide OA mandate in 2004, and with its current total of 7 university and funder mandates along with this vigorous governmental support from Minister Carr, Australia is the world's relative, if not absolute leader in OA, by size as well as timing. And it is about to consolidate that leadership with an international Open Access and Research Conference in Brisbane next week, convened by Tom Cochrane, the DVC who engineered the world's first university OA mandate.

By way of comparison, the US, the country with the world's largest research output, has only five OA mandates (though that includes one from the NIH, the world's biggest biomedical research funder, as well as Faculty mandates from Harvard and Stanford). Universities are the sleeping giants, and the council of the European Universities Association (EUA) has unanimously recommended the adoption of an OA mandate by its 791 member universities in 46 countries -- but that mandate has not been adopted yet (although Professor Bernard Rentier, Rector of University of Liege is working on it, with EurOpenScholar).

But Australia looks poised now to be the one that sets all the dominoes falling worldwide.

Stevan Harnad
American Scientist Open Access Forum

---

SUMMARY: SHERPA/RoMEO is listing the paid OA charges of non-Green publishers. On no account should any author or institution ever have to pay money to a publisher in order to be able to comply with a mandate to provide Open Access (OA). That would be a perverse distortion of the purpose of both OA and OA mandates; it would also profoundly discourage funders and institutions from mandating OA, and authors from complying with OA mandates.
    SHERPA has an outstanding record for supporting and promoting OA, worldwide. The OA movement and the global research community are greatly in their debt. However, SHERPA alas also has a history of amplifying arbitrary, irrelevant and even absurd details and noise associated with publisher policies and practices, instead of focusing on what makes sense and is essential to the understanding and progress of OA.
    I urge SHERPA to focus on what the research community needs to hear, understand and do in order to reach 100% OA as soon as possible -- not on advertising publisher options that are not only unnecessary but counterproductive to the growth of OA and OA mandates. Authors and funders who are foolish enough to squander their money on paying non-green publishers for OA (instead of just depositing their postprints anyway and relying on the IR's "email eprint request" Button during any embargo) can find out the prices for themselves. SHERPA/ROMEO is not a publishers' price catalogue.

---

The SHERPA/RoMEO site says:

"Where a publishers' standard policy does not allow an author to comply with their funding agency's mandate, paid open access options may enable an author to comply."

On no account should any author have to comply with any mandate to provide Open Access (OA) by having to pay money to a publisher.

That would be a grotesque distortion of the purpose of both OA and OA mandates.

It would also profoundly discourage funders and institutions from mandating OA, and authors from complying with OA mandates.

If a journal is not one of the 63% of journals that are already Green on immediate OA self-archiving then the right strategy for the author is to deposit the refereed final draft in their institutional repository anyway, immediately upon acceptance for publication.

Access to that deposit can then be set as Closed Access instead of Open Access during the publisher embargo, if the author wishes. The repository's semi-automatic "email eprint request" Button can then provide all would-be users with almost-OA during the embargo.

Most OA mandates tolerate an embargo of 6-12 months. Once immediate deposit is universally mandated by 100% of funders and institutions, that will provide at least 63% immediate-OA plus at most 37% almost-OA, immediately, for a universal total of 100% immediate-OA plus almost-OA.

After OA mandates are adopted universally, the increasingly palpable benefits of the resulting OA for research, researchers, and the tax-paying public will ensure that the rest of the dominos will inevitably fall of their own accord: Access embargoes will soon die their natural (and well-deserved) deaths, yielding 100% immediate-OA.

SHERPA has an outstanding record for supporting and promoting OA, worldwide. The OA movement and the global research community are greatly in their debt. However, SHERPA alas also has a history of amplifying arbitrary, irrelevant and even absurd details and noise associated with publisher policies and practices, instead of focusing on what makes sense and is essential to the understanding and progress of OA.

I urge SHERPA to focus on what the research community needs to hear, understand and do in order to reach 100% OA as soon as possible -- not on advertising publisher options that are not only unnecessary but counterproductive to the growth of OA and OA mandates.

---

Charles Oppenheim, U. Loughborough, replied:

"Stevan misunderstands the purpose of SHERPA/ROMEO.  It is there to report publishers' terms and conditions, to help authors decide where to place their articles. To argue that it should not list those publishers that are not "green" is akin to asking an abstracting service not to record those articles that the editor happens not to agree with.
 
Some funders, such as Wellcome,  encourage the applicant for funding to include the cost of paying a "gold" journal in their funding bid. If it is to perform a useful information function, SHERPA/ROMEO has to reflect current reality, not ideal future scenarios."

I can only disagree (profoundly) with my comrade-at-arms, Charles Oppenheim, on this important strategic point!

I certainly did not say that SHERPA/ROMEO should only list Green Publishers! It should list all publishers (and, more relevantly, all their individual journals).

But along with all the journals, SHERPA/ROMEO should only list and classify the journal policy details that are relevant to OA, OA mandates, and the growth of OA.

Those four relevant journal policy details are these:

(1) Does the journal endorse immediate OA self-archiving of the refereed postprint? If so, the journal is GREEN.

(2) Does the journal endorse the immediate OA self-archiving of the unrefereed preprint only? If so, the journal is PALE-GREEN.

(3) Is the journal neither GREEN nor PALE-GREEN? If so then the journal is GRAY.

(4) If the journal is PALE-GREEN or GRAY, does it endorse OA self-archiving after an embargo? If so, how long?

That's it. All the rest of the details that SHERPA/ROMEO is currently canonizing are irrelevant amplifications of noise that merely confuse instead of informing, clarifying and facilitating OA-relevant policy and decisions on the part of authors, institutions and funders.

Amongst the irrelevant and confusing idiosyncratic publisher details that SHERPA/ROMEO is currently amplifying (and there are many!), there are two that might be worth retaining as a footnote, as long as it is made clear that they are not fundamental for policy or practice, but merely details for two special cases:

(i) What version is endorsed for OA self-archiving: the author's final draft or the publisher's PDF?

(ii) Where does the journal endorse self-archiving: the author's institutional repository and/or central repositories?

The reason these details are inessential is that the default option in both cases is already known a priori:

(i) Self-archiving the author's final draft is the default option. A publisher that endorses self-archiving the publisher's PDF also authorizes, a fortiori, the self-archiving of the author's final draft. (IP pedants and pundits might have some fun thrashing this one back and forth, citing all sorts of formalisms and legalisms, but in the end, sense would prevail: Once the publisher has formally authorized making the published article OA, Pandora's box is open [sic], and residual matters concerning authors' prior versions or subsequent updates are all moot [as they should be].)

The default option of self-archiving the postprint is sufficient for OA, hence the PDF side-show is a needless distraction.

(ii)  Self-archiving in the author's institutional repository is the default option. A publisher that endorses self-archiving in a central repository also endorses, a fortiori, self-archiving in the author's own institutional repository.

The default option of self-archiving in the institutional repository is sufficient for OA, hence the matter of central deposit is a needless distraction. (Where direct central deposit is mandated by a funder, this can and will be implemented by automatic SWORD-based export to central repositories, of either the metadata and full-text or merely the metadata and the link to the full-text.)

Hence (i) and (ii) are minor details that need only be consulted by those who, for some reason, are particularly concerned about the PDF, or those who need to comply with a funder mandate that (neelessly) specifies direct central deposit.

There is absolutely no call for SHERPA/ROMEO to advertise the price lists of GRAY publishers for paid OA! I can only repeat that that is grotesque. Let authors and funders who are foolish enough to squander their money on paying those non-GREEN publishers (instead of just relying on their tolerated embargo limits plus the Button) find out the prices for themselves. (SHERPA/ROMEO is not an abstracting service; nor is it a publishers' price catalogue!)

---

Peter Millington, SHERPA, replied:

PM: "It is a pity that Prof. Harnad is only interested in "default" and "sufficient" options, and not in the best options, or indeed the most appropriate options. While author's final post-refereed draft is sufficient and acceptable for open access and research purposes, it is not the best."

OA is the best for research purposes. We don't yet have it. And it's long overdue.

I'm not sure whose purposes the publisher's PDF is best for, but whoever they are, their purposes are getting in the way of what is best for research purposes.

PM: "The best is the published version (publisher's PDF if you will). At the very least, this is the authoritative version vis-à-vis page numbers for quoted extracts and the like."

This issue has been much discussed in these pages: OA is needed (urgently) for all those users who can't afford paid access to the publisher's PDF. What these would-be users lack is access to the text, not a means of quoting extracts. Extracts can be quoted by paragraph number. Pages are on the way out anyway. What is urgently needed is access to the text. Publishers are far more willing to endorse self-archiving of the author postprint than the publisher's proprietary PDF. Hence author postprint self-archiving is the default option (if maximal OA, now, is the goal).

PM: "Also, it significantly expedites deposition to be able to use the publisher's PDF rather than having to generate your own, with all the complications that that may entail."

Significantly expedites deposition of what, where, by whom? I have deposited nearly 300 of my papers in the Southampton ECS IR. It takes me 1 keystroke and 1 second to generate PDF from TeX or to generate PDF or HTML from RTF. What complications do you have in mind?

PM: "In my view, the publishers who permit the use of their PDFs deserve to be applauded for their far-sightedness. Other publishers should be encouraged to do likewise."

The publishers to be applauded are the ones that are Green on immediate OA deposit of the postprint, regardless of whether they specify the author's postprint or the publisher's PDF. That's the line separating who is and isn't on the side of the angels regarding OA. The rest is trivial and irrelevant.

PM: "SHERPA therefore makes no apologies for having published our 'good list'"

SHERPA ROMEO could do a far, far greater service in informing authors and institutions, and in promoting OA, if it at long last got rid of all its superfluous categories and colour codes (yellow/preprint, blue/postprint, green/preprint+postprint, white/neither, and now "good"/PDF) and simply published a clear list of all the journals that endorse postprint self-archiving, regardless of whether the postprint is author-draft or publisher PDF and regardless of whether the journal also happens to endorse unrefereed preprint self-archiving -- and call that GREEN.

That, after all, is what OA is all about, and for.

PM: "(Some may concur with Prof. Harnad in regarding the Paid OA list as the "bad list", but I couldn't possibly comment.)"

The only thing authors need to know about these journals is that they are GRAY (and perhaps also how long they embargo access-provision).

PM: "As for where material may be deposited, Prof. Harnad states that permission to deposit in institutional repositories should be the default, implying that this would be sufficient. However, as before, institutional repositories alone are not the best option. Surely the best policy must be to be able to deposit in any open access repository - institutional and/or disciplinary."

No, the best policy is to allow deposit in any OA repository and to explicitly prefer IR deposit wherever possible. That is the way to integrate institutional and funder OA mandates so as to generate a convergence and synergy that will systematically cover all of OA space quickly and completely.

PM: "In any case, SHERPA/RoMEO has no choice but to reflect/quote the terminology for repository types used in the publishers' open access policies, CTAs, and related documentation. These are often wanting in clarity and are not always fully thought through. If the publishers do better, it follows that SHERPA/RoMEO's data will also improve."

Wouldn't it be nice, though, if SHERPA/ROMEO could lead publishers toward clarity, rather than just following and amplifying their obscurity (and their often deliberate obscurantism)?

This is all said in the spirit of unabating appreciation for all that SHERPA does do for OA -- but with an equally unabating frustration at what SHERPA persists in not doing for (and sometimes unwittingly doing against) OA, even though it would be ever so easy to fix.

This continuing insistence upon amplifying incoherent publisher noise simply because it is there cannot be described as a service to OA. And SHERPA does have a choice: It can do better for the research community without waiting for publishers to improve.

---

Andria McGrath replied:

AM: "It may be foolishness on the part of the funders, but I'm afraid it is the case that ALL the UK medical funders do insist that articles reporting research funded by them are posted on UK PubMedCentral within 6 months."

I have a simple solution, both for individual authors and for institutions who are trying to comply with a funder mandate to self-archive centrally articles that are published by journals that only endorse institutional OA self-archiving:

(1) Deposit the (refereed) postprint institutionally, immediately upon acceptance for publication.

(2) If the journal is Green on immediate OA, make the deposit immediately OA (otherwise rely on the "email eprint request" Button).

(3) If setting access to the deposit to OA is embargoed, set the access to OA when the embargo is over.

(4) If a funder requires deposit in UK PubMedCentral, set up SWORD for your IR so it will export the deposit to UKPMC at the requisite time -- and then let UKPMC worry about access-setting for the UKPMC version.

The author has complied with the funder mandate by depositing in his IR immediately upon acceptance for publication, and by setting access to the IR deposit as OA at the end of the publisher embargo. 

That's all there is to it. Funders cannot mandate any more of an author. And if the funder wants to pay publishers for the right to make the central UKPMC version OA, let them pay the publisher themselves.

The funder mandates are deposit mandates, not payment mandates. Comply by depositing institutionally, providing OA institutionally, and exporting the deposit to UKPMC. That's all there is to it.

AM: "I have just been going through Romeo trying to determine which of the major publishers allow this without the payment of article processing charges and there are very few. So far I have come up with BMJ Publishing, CUP, Company of Bioloigsts and Nat. Acad. Sci. that do allow this."

Fine. When those IR deposits are exported via SWORD to UKPMC, there will be no charge to be paid. For publishers other than those four,  there may be; that is not the problem of the author or his institution. And anyone construing the funder mandates as implying that it is the problem of the author or his institution, and that the mandate entails any further expense to the author or his institution, is profoundly misconstruing the mandate, the rationale for the mandate, and the rationale for OA.

Andria, you will have to get used to the fact that steps have been taken without carefully thinking them through. The funder OA mandates were very timely and welcome, and extremely important historically. But some (by no means all!) of them were also vague, careless, and, to a great extent "monkey see, monkey do" (many taking their cue from NIH and the Wellcome Trust, who themselves had not thought it through carefully enough). 

MRC simply adopted the wrong (because inchoate) mandate model. Other RCUK councils (such as, ARC, BBSRC, STFC) picked a better one. So did Europe's ERC, and now the EC, based on the EURAB model, which is the IDOA mandate model, the optimal one, and leads to none of these nonsensical consequences.

Good sense will eventually prevail, but until then, those who are trying to implement the existing mandates should not try to put themselves through impossible hoops -- and on no account should they lead their authors and institutions into grotesque and gratuitous expenses or constraints that were never the intention of either OA or OA mandates.

Just follow the sensible steps (1) - (4) above, and the rest will take care of itself as a matter of natural course.

AM: "As far as I can tell, Elsevier, Humana, Int Med Press, Wiley, Karger, Kluwer, Royal Soc and Springer do not allow self archiving in UK PMC by authors within 6 months, so that all authors funded by the medical charities are going to be forced into paying article processing charges to comply with their funders requirements if they publish in these publishers journals or in fully open access journals that make charges, like BMC."

Not only is it pure absurdity to imagine that the funder mandates were actually mandates for authors and their institutions to pay publishers for paid OA, but it is equally absurd to imagine that they were mandates for authors to publish only with publishers who endorse central self-archiving! 

Every single one of the eight publishers you list is on the side of the angels as regards OA: They are all Green on immediate deposit in the author's institutional repository, and the immediate setting of access to that deposit as OA.

Did anyone really imagine that OA was about more than that? That it further required publishers to consent to deposit in central repositories, for someone's capricious reasons? (The saga is made even sillier by the fact that if the blinkered centralists had sensibly targeted universal IR deposit first, then the dominos would -- and will -- fall for central repositories soon enough anyway! But instead they are creating gratuitous obstacles for OA itself, by putting centrality itself before OA -- and for absolutely no good reason, since all OA IRs are fully interoperable and harvestable anyway.)

And don't even get me started on the fatuousness of having decided to copycat PMC with a UKPMC! As if there were another category of biomedical research, consisting of UK biomedical research, requiring a central repository of its own: "Let me see now, what is it that British researchers -- and British researchers alone -- have found discovered about AIDS." (I hope no one replies that "one can search across PMC and UKPMC jointly," because that is the whole point! Search is done across distributed contents, not by going to -- or requiring -- one particular locus-of-deposit. Think OAIster, Citeseer or Google Scholar, not UKPMC! At most, UKPMC could simply be a harvester of UK biomedical output, for actuarial purposes, wherever its physical locus might happen to be.)

AM: "In view of this I do find it useful to have the extra information that Romeo is adding, and I would welcome even more specific info about publisher's policies re PMC. If I have any of this wrong I would be very grateful if people would let me know."

I think you have a good deal of it very, very wrong -- but it's not your fault, and you are not alone. I just wonder whether we will persist in bumbling in this misdirection for a few more years, yet again, until we discover we have goofed, or we will manage -- mirabile dictu -- to rally the good sense to fix it in advance...

Your weary and fast-wizening archivangelist,

Stevan Harnad
American Scientist Open Access Forum