Copyright

---

[Update: See new definition of "Weak" and "Strong" OA, 29/4/2008]

SUMMARY: The BBB "definitions" of OA still need some tweaking to get them right. OA means free online access to the full-text of refereed journal articles. With that comes the capability of linking, reading, downloading, storing, printing off, and data-mining, and that is what is urgently needed by researchers today. Let the Green OA mandates provide that, and the rest will all come naturally of its own accord. We'll never get self-archiving mandates adopted if we insist in advance that they include 3rd-party re-use rights. Nor does demanding machine-readable XML make sense when most authors still aren't even doing the few keystrokes it takes to deposit the drafts they already have. Let's not risk getting less by needlessly insisting on more.

---

I think it would be a big strategic mistake if today, when the cupboards are still 85% bare, we were to start insisting that deposits must all be Cordon Bleu ****.

OA just means free online access to the full-text of refereed journal articles. Please let's not risk getting less by needlessly insisting on more. The rest will come in due time, but what is urgently needed today, and what is still 85% overdue by more than 10 years today, is free online access. Let the Green OA mandates provide that, and the rest will all come naturally with the territory soon enough of its own accord.

But over-reach gratuitously now, and we will just delay the optimal and inevitable, already within our reach, still longer.

Ceterum Censeo: The BBB "definitions" (which were not brought down to us by Moses from On High, but puttered together by muddled mortals, including myself) are not etched in stone, and need some tweaking to get them right.

"Time to Update the BBB Definition of Open Access"

OA is free online access. With that comes, automatically, the individual capability of linking, reading, downloading, storing, printing off, and data-mining (locally).

The further "rights" for 3rd-party databases to data-mine and re-publish will come after universal Green OA mandates generate universal OA (free online access). But you'll never get universal Green OA mandates if you insist in advance that the 3rd-party re-use rights must be part of the mandate! (Notice that the Harvard mandate has an opt-out, which means it's not a mandate.)

"On Patience, and Letting (Human) Nature Take Its Course"

And as to demanding machine-readable XML from authors: 85% of authors cannot now be bothered to do even the few keystrokes it takes to deposit the drafts they already have: Does this sound like a reasonable time to ask them to upgrade their drafts to Cordon Bleu XML?

Stevan Harnad
American Scientist Open Access Forum

   [See also the original posting to which this is a follow-up: Part I.]

---

SUMMARY: I agree completely with Jonathan Oppenheim's and Bill Unruh's ends: (1) that authors should be able to publish and post derivative works and (2) that they should be able to adopt the Creative Commons license of their choice. I disagree only with their means: The American Physical Society (APS), the most progressive and adaptable of traditional journal publishers and the first to have adopted a formal policy endorsing Green OA self-archiving after the practice had evolved among its authors, should not now be publicly pressured into making further formal changes as if it were the villain, while so many other publishers still aren't even Green. Evolving practice should precede formal precept in the online age, when the research community is still discovering and exploring the potential of the new medium. The APS would no more try to prevent its authors from posting derivative works than it did in the case of the posting of their preprints and postprints. The practice (and mandating) of Green OA self-archiving, once it has generated universal OA, will effectively moot everything that stands in the way of CC licensing too.

---

Re: "Physicists slam publishers over Wikipedia ban" and "Traditional journals and copyright transfer"

Following an exchange of correspondence with Jonathan Oppenheim and Bill Unruh about the above posting, I want to stress that I agree completely with Jonathan Oppenheim's and Bill Unruh's ends:

(1) Derivative Works. Authors should be able to publish new articles which "differ in some reasonable way from the original work, even while possibly retaining much of the original."

I also think APS authors can already do this, and that APS would no more try to prosecute its authors for this practice than it tried to prosecute them for practicing self-archiving (before APS went on to adapt to evolving practice by formally adopting its Green OA policy, the first Green OA publisher policy, and a model for them all).

With derivative works too, formal APS policy will eventually adapt to evolving practice that is to the benefit of research progress in physics. Let practice again precede and guide precept.

(Note that published postprints are in fact "derivative works" relative to unpublished preprints.)

(2) Creative Commons Licensing. I am also fully in favor of CC licensing -- but not as a precondition for OA self-archiving today. All authors should adopt the CC license of their choice whenever they can. And where they cannot, they should just go ahead and self-archive under the Green publisher's current copyright agreement.

(If the publisher is not Green, authors should immediately deposit anyway; and if they wish to set access to their deposit as Closed Access instead of OA during an embargo period, they should rely on their repository's semi-automatic "email eprint request" Button to provide almost-immediate, almost-OA for all would-be users during any publisher embargo.)

(I do believe, though, that CC licensing will prevail as a matter of natural course, after universal OA has prevailed.)

So whereas I agree with Jonathan's and Bill's ends, I do not agree with their means.

Rather than trying to force an immediate formal policy change (if APS feels it needs more time to think it through), I think Jonathan and Bill should just go ahead and practice what they seek to practice: publish new articles which differ in some reasonable way from the original work, even while possibly retaining much of the original, or post them to wikis like Quantiki if they wish. APS formal precept will again follow evolving practice in due course, as it did with author self-archiving.

(By the way, the meaning of the enigmatic title "The American Physical Society is Not the Culprit: We Are" was of course that the reason we don't yet have universal OA [and all that follows from it] is that we are not yet universally self-archiving: I have dubbed this "Zeno's Paralysis.")

Stevan Harnad
American Scientist Open Access Forum

   [See also follow-up message: Part II.]

---

SUMMARY: A journal's copyright transfer agreement is too restrictive only if it tries to disallow author self-archiving of the accepted, refereed final draft (the "postprint"), free for all on the Web, where any user webwide can access, read, download, print-out, store, and data-mine that full-text for any research purpose whatsoever. The American Physical Society (APS) was always the most progressive of the established subscription-based publishers, and the very first to adopt a Green policy on author OA self-archiving. Today, 62% of journals are Green, but only about 15% of articles are being self-archived. Hence the first and foremost priority today is to get all authors self-archiving and all journals Green. Institutional and funder OA self-archiving mandates can and will ensure that both these things come to pass. This is not the time to be pursuing still more rights from Green publishers, particularly the most progressive one of all, APS. It's the time to self-archive and mandate self-archiving. The rest will take care of itself, but not if we keep chasing after what we don't need instead of grasping what is already within our reach.

---

"Physicists slam publishers over Wikipedia ban"
New Scientist 16 March 2008

I have some doubts about the accuracy of this New Scientist piece. What exactly is it that the American Physical Society (APS) is being alleged to be refusing to do?

The APS is the first publisher that endorsed OA self-archiving. It is the greenest of green publishers. APS authors are encouraged to post their unrefereed preprints as well as their refereed postprints, free for all, on the web.

So what exactly is the fuss about?

"Scientists who want to describe their work on Wikipedia should not be forced to give up the kudos of a respected journal."

"Describe" their work on Wikipedia? What does that mean? Of course they can describe their work (published or unpublished) on Wikipedia, or anywhere else.

And what has that to do with giving up the kudos of a respected journal?

Does this passage really mean to say "post the author's version to Wikipedia verbatim?" APS does not mind, but Wikipedia minds, because Wikipedia does not allow the posting of copyrighted work to Wikipedia.

Solution: Revise the text, so it's no longer the verbatim original but a new work the author has written, based on his original work. That can be posted to Wikipedia (but may soon be unrecognizably transformed -- for better or worse -- by legions of Wikimeddlers, some informed, some not). It's a good idea to cite the original canonical APS publication, though, just for the record.

Still nothing to do with APS.

"So says a group of physicists who are going head-to-head with a publisher because it will not allow them to post parts of their work to the online encyclopaedia, blogs and other forums."

In a (free) online encyclopedia that would provide the author's original final draft, verbatim, and unalterable by users, there would be no problem (if the encyclopedia does not insist on copyright transfer) as long as there is a link to the original publication at the APS site.

Blogs and Forums (again on condition that the text itself cannot be bowdlerized by users, or re-sold) will be treated by APS as just another e-print server.

"The physicists were upset after the American Physical Society withdrew its offer to publish two studies in Physical Review Letters because the authors had asked for a rights agreement compatible with Wikipedia."

This is now no longer about the right to post and re-post one's own published APS papers on the Web, it is about satisfying Wikipedia's copyright policy by going against APS's extremely liberal copyright policy. I side with APS. Let Wikipedia bend on this one, and let the text be posted (and then gang-rewritten as everyone sees fit). I see no reason why APS should have to alter its already sufficiently liberal policy to suit Wikipedia.

"The APS asks scientists to transfer their copyright to the society before they can publish in an APS journal. This prevents scientists contributing illustrations or other "derivative works" of their papers to many websites without explicit permission."

APS already says authors can post their entire work just about anywhere on the web without explicit permission. And they can rewrite and republish their work too. This fuss is about formality, not substance.

"The authors of the rescinded papers and 38 other physicists are calling for the APS to change its policy. 'It is unreasonable and completely at odds with the practice in the field. Scientists want as broad an audience for their papers as possible,' says Bill Unruh at the University of British Columbia in Vancouver, Canada, who has been lobbying separately against strict copyright rules."

What possible online audience is broader than the entire Web, which is what you get when you make your article OA?

If we're going to lobby against strict copyright rules, let's pick any of the Gray, or even the Pale-Green publishers in Romeo. But lets leave the Green ones like APS alone until all publishers are at least as green as they are.

And if you are going to lobby against copyright rules, make sure it is about a matter of substance and not just form.

"'To tell us what we can do with our paper is completely at odds with practice in the field' Gene Sprouse, editor-in-chief of the APS journals, says the society plans to review its copyright policy at a meeting in May. 'A group of excellent scientists has asked us to consider revising our copyright, and we take them seriously,' he says."

I am certain that the APS will accommodate all requests that are to the benefit of science, as they always have. I'm not always sure those who are lobbying for copyright reform really know what they want (or need) either. I trust them more if they say that they have made all their papers OA by self-archiving them. If they have not, yet they are still fussing, then they might be thinking of Disney re-mixes rather than science.

"Some publishers, such as the UK's Royal Society, have already adopted copyright policies that allow online reproduction."

The APS has long had a far more liberal OA policy than the Royal Society, a reluctant late-comer to OA.

There is something being misreported or misunderstood here.

A journal's copyright transfer agreement is only too restrictive if it tries to disallow author self-archiving of the accepted, refereed final draft (the "postprint"), free for all on the Web, where any user webwide can access, read, download, print-out, store, and data-mine the full-text for any research purpose whatsoever. That is what is called "Open Access."

Journals that have a policy that formally endorses immediate and permanent author self-archiving of the postprint are called "Green" journals. There is a directory of the policies of the 10,000 principal journals regarding OA self-archiving: 62% of them are Green; 28% are "Pale-Green" (endorsing the self-archiving of pre-refereeing preprints, but not refereed postprints) and 9% are Gray (disallowing the self-archiving of either preprints or postprints).

The American Physical Society (APS) is fully Green; it is the first Green publisher and helped set the example for the rest of the Green publishers.

If anything needs changing today it's the policy of the 9% of journals that are Gray and the 28% that are Pale-Green, not the 62% that are Green.

Once all publishers are Green, and all authors are making their papers OA by self-archiving them, copyright agreements will come into phase with the new OA reality, and everything that comes with that territory. For that to happen, authors doing -- and publishers endorsing -- OA self-archiving is all that is necessary. There is no need to over-reach and insist on reforming copyright agreements.

NB: Whenever and wherever an author does succeed in retaining copyright, or a publisher does agree to just requesting a non-exclusive license rather than a total copyright transfer, that is always very welcome and valuable. But it is not necessary at this time, and over-reaching for it merely makes the task of securing the real necessity -- which is a Green self-archiving policy -- all the more difficult.

In particular, pillorying the APS, which was the earliest and most progressive of Green publishers, is not only unjust, but weakens the case against Gray publishers, who will triumphantly point out that they are justified in not going Green, because the demands of authors are excessive, unnecessary, and unreasonable, as they are not even satisfied with Green OA (and most don't even bother to self-archive)!

The problem for the worldwide research community is not the minority (about 15%, mostly concentrated in computer science and physics) who are already spontaneously making their articles OA by self-archiving them, but the vast majority (85%, across all disciplines, including even some areas of physics) who are not.

Moreover, there is something special about the longstanding practice in some parts of physics of posting and sharing unrefereed preprints: That practice is definitely not for all fields. Hence the universally generalizable component of the physicists' practice is the OA self-archiving of the refereed postprint. Posting one's unrefereed preprints will always be a contingent matter, depending on subject matter and author temperament. (Personally, I'm all for it for my own papers!)

There has been a big technical change since the first days of Arxiv. The online archives or repositories have been made interoperable by the OAI metadata harvesting protocol. Hence it is no longer necessary or even desirable to try to create an Arxiv-like central archive for each field, subfield, and interfield: Each researcher has an institution. Free software creates an OAI-compliant Institutional Repository (IR) where the authors in all fields at that institution can deposit all their papers. The OAI-compliant IRs are all interoperable (including Arxiv), and can then be searched and accessed through harvesters such as OAIster, Citebase, Citeseer or Google Scholar.

Institutional IRs also have the advantage that institutions (like Harvard) can mandate self-archiving for all their disciplines, thereby raising the 15% spontaneous (postprint) self-archiving rate to 100%. Research funders (like NIH) can reinforce institutional OA self-archiving mandates too.

The objective fact today is that all physicists, self-archiving or not, are still submitting their papers for refereeing and publication in peer-reviewed journals. Nothing whatsoever has changed in that regard. The only objective difference is that today (1) 15% of all authors self archive their postprints, and among some physics communities, (2) most are also self-archiving their preprints.

The OA movement is dedicated to generalizing the former practice (1) (self-archiving peer-reviewed postprints, so they can be accessed and used by all potential users, not just those whose institutions happen to be able to afford to subscribe to the journal in which they were published) to 100% of researchers, across all disciplines, worldwide.

Radical publishing reform -- like radical copyright reform -- are another matter, and may or may not eventually follow after we first have 100% OA. But for now, it is a matter of speculation, whereas postprint self-archiving is a reachable matter of urgency.

The physicists, from the very outset, had the good sense not to give it a second thought whether they were self-archiving with or without their journal's blessing. They just went ahead and did it!

But most researchers in other fields did not; and still don't, even today, when 62% of journals have given it their official blessing.

That's why self-archiving mandates are needed. (Author surveys have shown that over 90% of authors, in all fields, would comply, over 80% of them willingly -- but without a mandate they are simply too busy to bother -- just as in the case of the "publish or perish" mandate: if their promotion committees didn't require and reward publishing, many wouldn't bother to do that either!)

What is needed now is not to make a campaign of trying to force APS to change its copyright policy. What is needed now is to generalize APS's Green OA self-archiving policy to all publishers.

And to generalize the existing 39 university and funder Green OA (postprint) self-archiving mandates to all universities and funders.

The rest (copyright reform and publishing reform) will then take care of itself.

Stevan Harnad
American Scientist Open Access Forum

In Newsweek, Brian Braiker discusses some interesting and important copyright, licensing and re-use issues -- about books re-using wikipedia material without attribution, about blogs posting published journal figures, about the re-use of free-culture text, photographs and music -- but it is important to understand that these are not Open Access (OA) issues.

OA primarily concerns the very special case of research articles published in peer-reviewed journals (about 2.5 million articles a year, in about 25,000 journals). All those articles, without exception, are (and always have been) author give-aways, written exclusively for maximal research usage and impact, not for royalty revenue. OA means making them freely accessible online to all would-be users webwide (for reading, downloading, storing, printing-off, data-crunching, using the findings in further research, building upon them, citing them -- but not necessarily for verbatim re-publication or re-posting of the text or figures -- beyond quoting limited verbatim text excerpts, with attribution, under "Fair Use": more than that still requires the author's permission).

OA does not require adopting a special CC attribution/re-use license (although if desired by the author and accepted by the publisher, such a license is always welcome). Nor does OA require the kinds of blanket mix-and-match re-use rights that teenagers might like to have for making and posting their own creative re-mixes out of commercial music or movies. That is a problem, but not an OA problem.

It is not even clear whether a blanket right to mix-and-match scientific content (even with attribution) would be a good thing. (Figures need not be re-posted, for example; the OA version can be linked; HTML even allows pinpoint linking to the specific figure rather than to the document as a whole.)

What sets OA's primary target apart is that it is an exception-free give-away corpus, wanting only to be read and used (but not necessarily re-published). It should not be conflated with or constrained by the needs of the much larger and more complicated and exception-ridden body of creative digital work of which it is merely a small, special subset.

Stevan Harnad
American Scientist Open Access Forum

---

SUMMARY: Green OA self-archiving removes not just "most" but all the "permission barriers" pertinent to research use.
   If "naked" (unlicensed) content on the web were really a barrier to use, we would be hearing about an urgent need to license all web content (e.g. advertisements, blogs). But the de facto default assumption since the advent of the web has been that if something is freely accessible online then you can download, print, store and otherwise "re-use" it.
   Far from perceiving themselves as being stuck behind "permission barriers" when they surf the web except when explicitly licensed to do otherwise, web users usually (wrongly) assume that they have even more rights than what comes with the free online territory: They will, for example, not only read, download, store, and print-off copyrighted web images, but also re-post them online, identically or in "derivative works," and sometimes even re-publish them in print publications. (You actually have to have shrill "You may NOT" notices to try to discourage users from going overboard like that!)
   Hence the "problem", if any, is precisely the reverse of what is being imagined by those who think that self-archived OA content risks not being used beyond on-screen reading unless it provides a formal permissive license over and above just being there, free for all on the web.
   The right to re-publish the verbatim text, online or on-paper, is not part of OA and never was, just as plagiarism or re-publishing a corrupted "derivative" version of the text is not and never was a part of OA, online or on-paper. OA's "unprecedented public good" is: free online access to what used to require paid on-paper access. This does not license re-publication or "derivative works" (collected or cut-pasted from the verbatim text), online or on paper (any more than it licenses the re-publication of copyrighted images). But the remarkable property of the new technology is that, for OA, such rights are not needed! For anyone, anywhere, can access (and print off) the online text, any time. Re-publication is not licensed, it is mooted by the properties of the new medium.
   What is left is (1) harvesting, (2) data-mining and (3) "derivative works." Harvesting likewise comes with the territory. (Google did not have to be licensed to harvest the rest of the web, hence needs no special license for OA content.) The same is true for other forms of data-mining by harvesting robots. For the publication of "derivative works," the rule is much the same as for copyrighted web images: Re-publication of verbatim OA texts, online or on paper, is not allowed without permission -- but it is not needed either, because a collection of URLs pointing to the OA URL serves the same function as re-publication.
   Corrupted, improved or otherwise altered versions of the OA texts? Apart from attributed fair-use excerpts, that is not allowed without the author's permission either. But what would ever have made anyone think that the invention of a new technology that would allow free and unlimited access to authors' give-away texts would mean that authors would all want to license that those texts, besides being freely accessible to all users, should be alterable or corruptible, ad lib? Surely we are happier with requiring specific case-by-case permission for such further uses, rather than a blanket license under the guise of "OA"?
   "Fair use" (like "copying") was a paper-based notion. In the case of the online medium, "fair use" quite naturally, indeed unavoidably, expands to include everything else that comes with the freely accessible online territory -- but that only means access and personal use, not re-posting (online), re-publishing (on paper), or perturbing that freely accessible content (online or on paper) without the author's permission -- apart from freely re-distributing its URL.

---

In Open Access News, Peter Suber wrote (by way of reply to my posting):

PS: "Comments. I hope no one minds if I reprint my comments from June 12 2007 in which I responded in detail to a very similar post by Stevan:

"Stevan isn't saying that OA doesn't or shouldn't remove permission barriers. He's saying that removing price barriers (making work accessible online free of charge) already does most or all of the work of removing permission barriers and therefore that no extra steps are needed."

So far this is exactly correct -- except I would definitely say that Green OA self-archiving removes not just "most" but all the "permission barriers" pertinent to research use, which is what OA is all about. Remember that there is also a "permission barrier" to re-publishing in print, but OA is not and never was intended to address that!

PS: "The chief problem with this view is the law. If a work is online without a special license or permission statement, then either it stands or appears to stand under an all-rights-reserved copyright. The only assured rights for users are those collected under fair use or fair dealing. These rights are far fewer and less adequate than OA contemplates, and in any case the boundaries of fair use and fair dealing are vague and contestable."

If "naked" (unlicensed) content on the web is really a barrier to use, how come we are not hearing about the need to license all web content (e.g. advertisements, blogs) because people are otherwise afraid to download, print, store and otherwise "re-use" them? (Answer: Because people are doing all those things, without hesitation.)

I think the truth is the exact opposite! That the default option, if something is freely accessible on the web, is that it's fine to do all those other things that come with it, and then some. (You can't even view web content without downloading and "storing" at least for long enough to read, yet that downloading and storing are not explicitly licensed!)

Far from perceiving themselves as being stuck behind "permission barriers" when they surf the web except when explicitly license to do otherwise, web users usually (wrongly) assume that they have even more rights than what comes with the territory: They will, for example, not only read, download, store, and print copyrighted web images, but also re-post them online, identically or in "derivative" form, and sometimes even re-publish them in print publications, identically or in "derivative" form. You actually have to have shrill "you may not" notices to try to discourage them from going overboard like that!

Now I have no particular interest in these excesses one way or the other (either to cultivate them or to curb them). But I think that they clearly illustrate that the "problem", if any, is precisely the reverse of what is being imagined by those who think that self-archived OA content needs a formal permissive license over and above just being there, free for all on the web, otherwise it risks not being used beyond on-screen reading! All seven individual uses I described in my earlier posting (as well as the two robotic ones) can be and are being fully exercised for all Green OA content today -- and there just aren't any further uses to which OA can justifiably lay claim (as OA).

PS: "This legal problem leads to a practical problem: conscientious users will feel obliged to err on the side of asking permission and sometimes even paying permission fees (hurdles that OA is designed to remove) or to err on the side of non-use (further damaging research and scholarship). Either that, or conscientious users will feel pressure to become less conscientious. This may be happening, but it cannot be a strategy for a movement which claims that its central practices are lawful."

Paying permission fees for Green OA content? Paying whom?

I honestly cannot imagine who or what you have in mind here, Peter! I am conscientious about not re-using web-accessible images in re-postings or publications unless I know they are public domain or I have permission. But I (and every other researcher on the planet) don't give a second thought as to whether I may read, download, store, print-off, and re-use the contents -- but not re-post or re-publish the verbatim text (which is like the image) -- of journal articles we can access freely on the web.

I am not feigning puzzlement: I am truly baffled about why, when the reality is the exact opposite, OA advocates, of all people, would worry that web users might be too coy (or "conscientious") to do with OA texts exactly the same things that we all do with all other free web content -- and too coy or "conscientious" to do so specifically in the case OA texts, of all things, because they lack a formal license to do it (exactly as virtually all other web content lacks such a license!).

[Could it be, Peter, that when you think of "unlicensed OA content" you are thinking of hybrid Gray/Gold publishers, who take an author's money, but don't adopt the full Creative Commons License that that money has surely paid for? I would probably agree with you on such cases, but my paradigmatic case is not paid Gray/Gold OA but Green OA, where it does not matter what copyright transfer agreement an author has signed as long the journal endorses immediate Green OA self-archiving, as 62% already do. All the rest comes with the Green OA territory.]

I think part of the problem is that (some) OA advocates may indeed be over-reaching in what they are taking to be licensed by "OA". I make absolutely no bones about the fact that the right to re-publish the verbatim text, online or on-paper, is not part of OA and never was, just as plagiarism or re-publishing a corrupted "derivative" version of the text is not and never was a part of OA, online or on-paper. OA is a new capability opened up by a new medium. I quote your own stirring words:

"An old tradition and a new technology have converged to make possible an unprecedented public good. The old tradition is the willingness of scientists and scholars to publish the fruits of their research in scholarly journals without payment, for the sake of inquiry and knowledge. The new technology is the internet. The public good they make possible is the world-wide electronic distribution of the peer-reviewed journal literature and completely free and unrestricted access to it by all scientists, scholars, teachers, students, and other curious minds."

The unprecedented public good is free online access to what used to require paid on-paper access. This does not license re-publication or "derivative works" (cut-pasted from the verbatim text), online or on paper, but the remarkable property of the new technology is that it does not need to! For one of the other things that "comes with the [Green OA] territory" is that anyone, anywhere, can access (and print off) the online text, any time. Re-publication is not licensed, it is mooted. There is simply no need for it. (The worriers about licensing content for "course-packs" are still thinking the old way: OA content does not need a license to be put in a course-pack: The text does not need to be put in a course-pack! Only its URL does.)

So, having disposed of the red herring of a special license that is supposedly needed to allow downloading, storing, and printing-off of freely available web content, we now see that there is no need to license re-publishing either, either online or on paper. What is left? Harvesting, data-mining and derivative works. Harvesting, as noted, likewise comes with the territory. If google did not have to be licensed to harvest the rest of the web, why on earth would we imagine that it needed to be licensed to harvest OA content, of all things? Ditto for the data-mining by harvesting robots. (Individual data-mining on your downloaded copy is not even at issue.)

Now what about "derivative works"? Let's be specific: re-publication of verbatim OA texts, online or on paper? That is not allowed without permission -- but nor is it needed, because a collection of URLs does the trick just as well.

Altered or corrupted versions of the OA texts? Apart from attributed fair-use excerpts, that is not allowed without permission either. But what would ever have made anyone think that the invention of a new technology that would allow unlimited access to authors' give-away texts would mean that authors would all want to license that those texts, besides being accessible, should be alterable or corruptible, ad lib? Surely we are happier with requiring specific case-by-case permission for such further uses, rather than a blanket license under the guise of "OA"?

What uses are left that research or researchers could possibly want, as a general rule: Harvestability into a commercial, pay-to-use database? It seems to me that that is no longer an OA matter. Indeed, authors might prefer to license their content to free database providers in preference to commercial ones. But either way, that is not part of OA, which was about "completely free and unrestricted access to it by all scientists, scholars, teachers, students, and other curious minds." Nothing there about commercial interests.

PS: "This doesn't mean that articles in OA repositories without special licenses or permission statements may not be read or used. It means that users have access free of charge (a significant breakthrough) but are limited to fair use."

"Fair use" was a paper-based notion. In the case of the online medium, "fair use" quite naturally, indeed unavoidably, expands to include everything else that comes with the online territory. In the case of freely accessible web documents, that "fair use" simply includes downloadability, storeability, printability, and data-minability, for individuals; and, for harvesters: robotic harvestability, data-minability, and certain derivative services (though I would not venture to specify which, though they certainly include free boolean searchability). With the Green OA territory comes also the accessibility online to everyone everywhere, mooting forever all need for collections, course-packs, re-publications, or other such "derivative works," online or on paper. For individual "derivative works," some form of "fair use" criterion still has to apply to determine how much verbatim content is permissible without the original author's permission.

Stevan Harnad
American Scientist Open Access Forum

---

[Update: See new definition of "Weak" and "Strong" OA, 29/4/2008]

SUMMARY: Not one, not two, but three of my valued OA comrades-at-arms have so far publicly registered their disagreement with my position on "re-use" rights. Here is my summary of the points at issue: Judicet Lector.
   Individual re-use capabilities: If a document's full-text is freely accessible online (OA), that means any individual can (1) access it, (2) read it, (3) download it, (4) store it (for personal use), (5) print it off (for personal use), (6) "data-mine" it and (7) re-use the results of the data-mining in further research publications (but they may not re-publish or re-sell the full-text itself: "derivative works" must instead link to its URL).
   Robotic harvestability: In addition, (8*) robotic harvesters like Google can harvest and index the freely available Web-based text, making it boolean full-text searchable. (9*) Robotic data-miners can also harvest the full-text, machine-analyse it, and re-use the results for research purposes (but they may not re-publish or re-sell the full-text itself: "derivative works" must instead link to its URL).
   OA is about access and use, not re-publication or re-sale: Online re-publishing or re-sale rights were never part of OA, any more than on-paper re-publishing or re-sale rights were; nor do they need to be, because of all the capabilities that come with the free online territory.
   The Green OA territory: Capabilities (1)-(9*) all come automatically with the Green OA territory. Hence there is no need to pay for Gold OA to have these capabilities, nor any need for further re-use rights beyond those already inherent in Green OA. Sixty-two percent of journals today already endorse immediate Green OA self-archiving.
   Gold OA includes Green OA: If you do elect to pay a publisher for Gold OA, you also get the right to deposit your refereed final draft ["postprint"] in your own OA Institutional Repository. Hence even here there is no need for further "re-use rights." (If you pay for "Gold OA" without also getting this Green OA, you have done something exceedingly foolish.)
   "Harvesting rights"? If authors self-archive their articles on the web, accessible freely (Green OA), then robots like Google can and do harvest and data-mine them, and have been doing so without exception or challenge, for years now.
   What about Gray publishers? With Gray publishers (i.e., neither Green nor Gold) the interim solution today is (i) Immediate Deposit (IDOA) Mandates, (ii) Closed Access deposit for Gray articles, and (iii) reliance on the semi-automatized "Email Eprint Request" ("Fair Use") Button to provide for individual researchers' usage and re-usage needs for these Gray articles during any Closed Access embargo interval (but note that the Fair Use Button cannot provide for robotic harvesting and data-mining of these embargoed full-texts).
   Extra Gold OA rights? For those articles published in the 38% of journals that are still non-Green today, I think that to rely on (i)-(iii) above is a far better interim strategy for attaining 100% OA globally than to pay hybrid Gray/Gold publishers for Gold OA today. But regardless of whether you agree that (i)-(iii) is the better strategy in such cases, what is not at issue either way is whether Gold OA itself requires or provides "re-use" rights over and above those capabilities already inherent in Green OA -- hence whether in paying for Gold OA one is indeed paying for something further that is needed for research, yet not already vouchsafed by Green OA.

---

As not one, not two, but no fewer than three of my valued OA comrades-at-arms have so far publicly registered their disagreement with my position on one (possibly two) points of detail concerning "re-use" rights, it is perhaps worthwhile taking a closer look at these points to see exactly what is and is not at issue:

Individual re-use capabilities. The concern is about "re-use rights," but I prefer to speak of "re-use capabilities." My OA comrades suggest that these consist of more than just the ability to read, and they are certainly right about that: If a document is OA -- i.e., if its full-text is freely accessible online, immediately and permanently, webwide -- then that means that any individual, webwide, can (1) access the document online, (2) read it, (3) download it, (4) store it (for personal use), (5) print it off (for personal use), (6) "data-mine" it and (7) re-use the results of the data-crunching in further research and research publications (however, they may not re-publish or re-sell the full-text itself, in "derivative works," either online or in print, without permission, beyond a reasonable number of quoted/cited excerpts: instead, they may only link to the OA full-text's URL in such derivative works, leaving the user to click to access it).

Robotic harvestability. In addition to the individual re-use capabilities (1-7), there are the following: (8*) Robotic harvesters like Google can harvest the freely available Web-based text (exactly as they harvest all other texts that are freely available on the Web) and inverse-index it, thereby making it searchable by boolean full-text search in their search engines. (9*) Robotic data-miners can also harvest the text, machine-analyse it, and re-use the results of their data-crunching for research purposes in further research and research publications (however, they may not re-publish or re-sell the full-text itself, in "derivative works," either online or in print, without permission, beyond a reasonable number of quoted/cited excerpts: instead, they may only link to the OA full-text's URL in such derivative works, leaving the user to click to access it).

The Green OA territory. All the above -- (1)-(7) plus (8*)-(9*) -- already come automatically with the (Green) OA territory when a full-text is made freely accessible online, immediately and permanently, webwide. It is for this reason that I continue to insist -- and this is the fundamental point of disagreement with my three OA comrades -- that there is no need whatsoever for any further re-use rights beyond what already comes automatically with the Green OA territory. In particular, there is no need to pay extra for Gold OA, in order to "purchase" these "extra" re-use rights. Nor is there any need to add any further re-use rights to Gold OA copyright agreements (although formalizing the rights is always fine, and a good idea).

Gold OA includes Green OA. If you have paid a publisher for Gold OA, you have, among other things, certainly paid for the right to deposit your refereed final draft ["postprint"] in your own OA Institutional Repository (along with any XML tagging you may wish to add to facilitate usage, search, harvesting and data-mining): hence you already have (1)-(9*). Hence what you are paying for, if you elect to pay for Gold OA, is not extra re-use rights, but simply Gold OA, which already includes Green OA, which in turn already provides all the requisite re-use capabilities.

Gold OA without Green OA? If any author (or funder) were ever to pay for "Gold OA" without thereby also getting the publisher's blessing to deposit the refereed final draft (postprint) in the author's own Institutional OA Repository (Green OA), then that author (or funder) would be doing something exceedingly foolish. (I know of no "Gold OA" today that does not automatically include Green OA.) But, apart from that, paying for Gold OA is still an unnecessary expenditure today for all except those to whom money is no object and who consider paid Gold OA to be worth the cost because it helps promote Gold OA, reinforcing the fact that it is a potentially viable cost-recovery model. Gold OA itself is certainly not necessary for any re-use needs that are purportedly not fulfillable through Green OA alone.

Pay for Green OA rights? The second possible point of disagreement with my three OA comrades, a more minor one, would be about whether it is worth paying for Gold OA to a hybrid Gray/Gold publisher who does not endorse Green OA self-archiving except if paid for Gold OA: I'm inclined to say that Closed Access self-archiving in your Institutional Repository (IR), along with the IR's "Email Eprint Request" Button, is a much better strategy than paying such a hybrid Gray/Gold publisher for Gold OA in such cases, because it facilitates exception-free IDOA Deposit Mandates. But this is a less important point of disagreement than the logical, practical point about whether paid Gold OA is indeed needed for certain re-use rights.

"Harvesting rights"? I will close on the sole potentially substantive matter on which my three OA comrades do have at least a theoretical point -- but, I will argue, a point that has no practical import: The reason I put an asterisk after 8* and 9* is that it can be argued that whereas the individual uses (1) - (7) do indeed come with the territory if one makes a document freely accessible on the web, this does not necessarily cover robotic uses such as harvesting.

"Could?" is trumped by "Does." I will give a very simple and pragmatic answer: "Can," "could," "cannot" and "could not" are all trumped here by "does." My OA comrades are needlessly reasoning hypothetically in this case, when the objective evidence is already in: "If authors were to self-archive their articles on the web, freely accessibly (Green OA), as described above, could robots like Google harvest and data-mine them?" The answer is a resounding "yes": they could, and can, as demonstrated by that fact that they already do, without exception or challenge, and have been doing so for years now!

Articles vs. books. We are not talking here about the full-texts of books, ambivalently provided to Google by their publishers (and authors), or scanned directly by Google, with certain conditions imposed by their publishers and authors on their re-use. We are talking about authors' final drafts (postprints) of their peer-reviewed journal articles, self-archived free for all by their authors in order to maximize their accessibility, usage and impact. In the case of books, there can be and have been contentious harvesting issues. But in the case of self-archiving, not a single article's harvestability has been contested, and we already have a decade and a half of precedent and practice behind us in this. So those who are worrying about the need to formally guarantee Google's (and other harvesters') "right" to do what they are already doing, without exception or challenge, since the advent of the Web, are worrying about a notional obstacle, not a real one.

OA is not about or for re-publication or re-sale, online or in print; OA is about access and use. Before replying to insist that I am wrong about about "re-use" being a nonproblem for self-archived postprints, may I ask my readers please to recall (i) the parentheticals I carefully inserted earlier, concerning both individual users and harvesters: "(though they may not re-publish or re-sell the full-text itself, in "derivative works," either online or in print, without permission, beyond a reasonable number of quoted/cited excerpts: instead, they may only link to the OA full-text's URL in such derivative works, leaving the user to click to access it)". None of that is part of OA, nor has it ever been ("BBB" Declarations to the contrary notwithstanding). OA is a brand new possibility, opened up by a brand new medium: the Web. "Online re-publishing or re-sale rights" were never part of OA, any more than on-paper re-publishing or re-sale rights were -- nor do they need to be, because of everything that comes with the OA territory (i.e., with being freely accessible to one and all online).

What about Gray publishers? Recall also that (ii) Gold OA already includes Green OA (as part of what you are paying for) and that (iii) with Gray publishers (i.e., those that are neither Green nor Gold) the interim solution for now is Immediate Deposit mandates plus the semi-automatized "Email Eprint Request" (or "Fair Use") Button for any Closed Access deposits. That does provide for individual researchers' uses and re-uses even for this "Gray" literature (meaning non-Green, non-Gold journal articles) -- although it does not provide for robotic harvesting and data-mining of the (Closed Access) full-texts, just their metadata.

IDOA and the Button -- or Paid Gold OA? Here, as I said, my colleagues and I may agree to disagree on the second, minor point, as to whether (a) it is a better strategy to rely, for now, on mandated IDOA and the Button for articles published in non-Green journals (38%), trusting that that will eventually force those journals to go Green (62%)? or, rather, (b) it is a better strategy to pay for Gold OA right now? But note that what is not at issue either way is whether Gold OA itself requires or provides "re-use" rights over and above those capabilities already provided by Green OA -- hence whether in paying for Gold OA one is indeed paying for something further that is needed for research, but not already vouchsafed by Green OA.

Stevan Harnad
American Scientist Open Access Forum

On Tue, 9 Oct 2007 Andrew Albanese, Associate Editor, Library Journal, wrote:

"[J]ust writing to see if you have any thoughts on the UKPMC [UK PubMed Central] statement on re-use...seems a little unnecessary to me. Stating the obvious? Rather than say "copyright still applies," would it not have been more useful to issue guidelines on, say, how to craft a copyright clause that facilitates open access? Do these broad statements help anyone?"

I agree that the UKPMC re-use statement is unnecessary and stating the obvious. (Even advice on amending copyright clauses to facilitate Green OA self-archiving is not necessary as a precondition for self-archiving, or for mandating self-archiving, although it is a good idea to try to amend copyright agreements where feasible and desired -- hence good advice is always welcome.)

(1) To begin with, the UKPMC statement is about paid Gold OA, and (for reasons I have adduced many times before) I believe that -- except for those researchers and funders who are so well off that money is no object -- paying for Gold OA at this time is unnecessary and a waste of money (until and unless most or all of the institutional money that is currently being spent on subscriptions is released to pay for Gold OA).

(2) Successfully establishing a credible, high-quality fleet of paid Gold OA journals was definitely useful in order to demonstrate the principle of paid Gold OA as a feasible one (especially under the current financially straitened circumstance in which most of the potential Gold OA funds are still tied up in institutional journal subscriptions); but that does not change the fact that Gold OA is far from being either the fastest or surest way to scale up to 100% OA today.

(3) The fastest and surest way to provide 100% OA today is for authors to self-archive their (published) articles in their own Institutional Repositories [IRs] (not in Central Repositories [CRs] like PubMed Central or UKPMC: CRs should harvest from IRs) -- and for authors' institutions and funders to mandate that they self-archive.

(4) This Green OA self-archiving does not require the description or assertion of any new "re-use rights": All the requisite uses already come with the Green OA territory itself (i.e., with the full text being made freely accessible to all on the web).

So this is a lot of fuss and fanfare about nothing: details peculiar to paid Gold OA and to direct deposit in 3rd-party CRs like UKPMC. Not what the research community urgently needs today (100% OA), nor what will get us there.

Stevan Harnad
American Scientist Open Access Forum

Andrew Adams (2007) has written a powerful, relentless and devastating critique of (the Open Access aspects of) Kevin Taylor's (2007) "Copyright and research: an academic publisher's perspective." Adams cites other archivangelists in support of his position, but this lucid, timely, rigorous and compelling synthesis is entirely his own. It will be seen and cited as a landmark in the research community's delayed but inexorable transition to Open Access.

Taylor, K. (2007) Copyright and research: an academic publisher's perspective. SCRIPT-ed 4(2) 233-236

Adams, Andrew A, (2007) Copyright and research: an archivangelist's perspective. 4(3) SCRIPT-ed 285

(Kevin Taylor is Intellectual Property Director at Cambridge University Press, a publisher that is on the side of the angels insofar as its author self-archiving policy is concerned, which is as green as green can be. However, although Kevin's views on other aspects of copyright and publishing may well be irreproachable, his views on Open Access need substantial rethinking.)

Stevan Harnad
American Scientist Open Access Forum

Klaus Graf writes:

"1. time of free access (the embargo-question): This is the only question Stevan Harnad is interested in. If we can call the OA-FREE journals of DOAJ 'OA' we should also... call [self-archived articles that are] freely accessible articles after an embargo 'OA'."

This is incorrect. OA means immediate, permanent, free, full-text access online to published journal articles, webwide. ("Immediate" means immediately upon acceptance for publication.) Hence embargoed access means embargoed access, not OA.

I am interested in OA but it has become quite evident across the past 13 years that not nearly enough authors make their articles OA spontaneously, of their own accord (only about 15% do), despite its demonstrated benefits. It is also quite evident that the only real barrier to 100% OA is the keystrokes that it takes to deposit the article and its metadata into the author's Institutional Repository.

It is for this reason that my own focus is currently on (1) institutional (and research funder) mandates that ensure that those keystrokes are executed as a matter of institutional/funder policy and on (2) developing the OA metrics that will quantify and reward those benefits. Administrative deposit mandates of course only ensure deposit, not OA. But the benefits of OA themselves will ensure that all those deposits will be made free as worldwide deposits approach 100%, and new deposits will not long thereafter be OA ab ovo.

And during the brief life of embargos, Institutional Repositories will provide "almost OA" via their "Fair Use" Button, allowing would-be users to request -- and authors to provide -- an email version almost instantly, with a click from the requester and then a click from the author.

American Scientist Open Access Forum

We are simply repeating ourselves, at greater and greater length, so my reply to Peter Murray-Rust's latest will just be a brief recap: Green Open Access (OA) (free online access) to journal articles is fully within reach: The only thing authors need to do is self-archive and the only thing their universities and funders need to do is mandate that they self-archive (and some are at long last starting to do it). Peter Murray-Rust wants more, and so he wants to complicate the definition of OA as well as the means of attaining it. He wants us to do something more complicated and not immediately attainable (reform copyright/licensing, reform publishing) when we are not yet doing the simple thing that is immediately attainable (because many already find even that too complicated!). I say we should put all our efforts behind simply attaining the immediately attainable. The rest will come soon enough, with the OA territory. Otherwise we are merely compounding our already unconscionable delay in reaching the optimal and inevitable, at long last.

Stevan Harnad
American Scientist Open Access Forum