Profile cover photo
Profile photo
Peter Suber

Post has attachment
A few minutes ago in Brooksville, Maine. Sunrise over the Bagaduce River.

Correcting the record on the Harvard OA license.

Yesterday at the Scholarly Kitchen, Karin Wulf and Simon Newman posted some objections to the UK Scholarly Communications License, which is based on the Harvard OA license.

In the process they characterized the Harvard OA license and OA policies, sometimes correctly and sometimes incorrectly.

I posted a response which is -still undergoing moderation- now online. I'm posting a copy here in case my response is rejected, abridged, or delayed.


Here are a few comments on what the authors have said about the Harvard OA policies. I leave comments on the UK-SCL for another time.

When Wulf and Newman first refer to the Harvard OA license, they link to the Harvard repository terms of use. That's confusing. The Harvard OA license is embodied in the OA policies, not the repository terms of use. These are separate and complementary.

For the language of the Harvard OA license, see (for example) the language of the policy from the Faculty of Arts and Sciences or the Harvard model OA policy.

> "There is no evidence that the so-called Harvard model is widespread or that it may become so...."

This doesn't matter to the merits of the Harvard model and even less to the merits of the UK-SCL. But for the record the Harvard model has been adopted by at least 70 institutions in North America, Europe, Africa, and Asia.

> "First, the Harvard model is more of a statement of institutional preference than a directive, and it is entirely voluntary on the part of academic staff who are not compelled to participate...."

It helps to be more careful here. The authors' statement is true in three senses: (1) faculty were not compelled to vote for the OA policies; (2) once adopted by vote, the waiver option in each policy lets faculty members decide for or against OA for each new article, with the default favoring OA; and (3) although the policies include a commitment to deposit a certain version of each new article in the repository, there is no penalty for non-deposit. But the statement is untrue in another sense: (4) the Harvard OA policies have legal consequences, and are not just statements of institutional preference. By voting up the policies, faculty granted the university a certain set of nonexclusive rights.

> "Some [members of the Harvard History Department] may have requested waivers for all of their articles."

There are two ways to read this, one certainly false and the other probably false. (1) It might mean that one waiver can cover all of an author's future articles, and that some members of the History Department have requested this kind of standing waiver. That is untrue. The Harvard OA policies only allow article-by-article waivers, not standing waivers. Faculty who want waivers for separate articles must obtain them separately. (2) It might mean that some faculty in the Department have made separate waiver requests for each of their articles. That's possible but not likely. The highest number of waivers requested by any member of the Department is very low.

(I'm taking the authors' claim in its strongest form, and disregarding the fact that they say some Harvard historians "may" have done this, not that any actually have done it.)

The authors say that the Harvard repository terms of use are "similar to a CC-BY-NC-ND license." I suppose that's true. But if so, it's equally true that the terms are dissimilar to a CC-BY-NC-ND license. The terms permit some but not all commercial use (differing from NC licenses), and some but not all derivative works (differing from ND licenses). If the authors' point was that these terms differ from the straight CC-BY-NC license used by the UK-SCL, that's true.

Finally, commenter Siloh says, "The Harvard model is 'ignored' by top scientists at Harvard because it restricts choice and the freedom to choose where to publish their work, i.e., prevents them from publishing in the most prestigious journals in their subject area (knowledge gained directly from two Harvard Profs)."

I have no doubt that some faculty ignore the policies. But the policies do not restrict the freedom of faculty to choose where to publish their work. There are two kinds of misunderstanding on this front. (1) Some covered authors believe that the policies require them to publish their new articles in OA journals. That belief is untrue, and even a superficial reading of the policies shows it to be untrue. This misunderstanding reflects the tenacious background assumption that all OA is gold OA, or that the only way to make an article OA is to publish it in an OA journal. This kind of misunderstanding is common to all green policies, not just to Harvard-style green policies. (2) Some covered authors overlook or undervalue the waiver option. This option assures the freedom of faculty to choose where to publish their work, and was incorporated into the policies precisely to assure this freedom. This was well-explained at the time of each faculty vote, which is why faculty voted for the policies (at four Harvard schools by unanimous votes), and it's well-explained in all our published material on the policies.

The Harvard repository has more than 39,000 deposits, the vast majority from scientists, who are no slouches at publishing in the journals of their choice.

#oa #openaccess

Post has attachment
A few minutes ago in Brooksville, Maine. Sunrise over the Bagaduce River.

A new journal in fluid mechanics is putting together its inaugural issue, and asked "some illustrious people like [me]" to submit an article. My fields are philosophy, law, and open access. I find it hard to think up a greater mismatch between a journal's topic and my areas of strength.

I've gotten these emails in the past, but this morning I have a minute to name names.

The journal is the Fluid Mechanics Research International Journal, and the publisher is MedCrave.

Here's a longer passage from the solicitation: "I am pleased to inform you that Fluid Mechanics Research International Journal (FMRIJ) is planning to release Inaugural issue by the end of this month and we need only two articles to accomplish this issue. In fact I am afraid as I am hardly having few days to release the issue. Hence I have chosen some illustrious people like you to support us for releasing the upcoming issue. So will you please help us by submitting a Research/Reviews for publication towards FMRIJ. Your prompt submission sustains us a lot and impacts my ranking in end of this month...."

Why would a journal solicit an article from an author who doesn't work in the field? Why reveal that it doesn't spend a second checking on the authors from whom it solicits work, and is willing to call them illustrious when they are not even remotely qualified? Why ruin its reputation before the release of its inaugural issue by demonstrating that it uses spam to solicit articles from random strangers? Why try to put together its inaugural issue in just a "few days"? Why even launch if it doesn't have reason to think there is a niche to fill and good work to fill it?

If we name names when we get these random solicitations, will we stop the practice? I don't know, but it's an easy experiment to try.

Some evidence of the demand for research literature by nonprofessionals

A recent tweet from @openscience...

...asked me to support my claim in a 2012 interview that "more than 40 percent of the visitors to PubMed Central come from domains."

I'm glad to. But the answer is too long for Twitter, or at least I'd like to say more about it than I can fit in a tweet. So I'm giving it here and linking to it from Twitter.

(For more on this practice, see my page on escaping from the confines of Twitter.)


The NIH itself released this number in two places:

1. Congressional testimony by David J. Lipman, M.D., Director, National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, April 19, 2011. "Based on the type of Internet domain from which they access PubMed Central (e.g., .com, .edu, .net, .gov), we estimate that approximately 25% of our users are from universities, 40% are private citizens or those using personal Internet accounts, and 17% are from companies (the remainder consists of government users or others)."

2. The NIH Public Access Policy, April 2012. "Based on internet addresses, an estimated 25% of users [of PubMed Central] are from universities, 17% are from companies, and 40% from the general public."


For more on the demand for research literature by nonprofessionals, see section 5.5.1 ("Access for Lay Readers") in my 2012 book (Open Access, MIT Press, ).

Here's a deep link to an OA edition of that section of the book.

For more evidence, see endnote 17 from that section. (The note call is at p. 117, and the note text at p. 205.) Here's a deep link to that note.

Here's the full note:

17. See Richard K. Johnson, "Will Research Sharing Keep Pace with the Internet?" The Journal of Neuroscience 26 (37) (September 13, 2006), pp. 9349-9351.

"The large audience for freely accessible scientific knowledge maybe surprising to many, but the hunger for it is apparent from experience of the National Library of Medicine (NLM). A few years ago, NLM transformed its fee-based index and abstracts of biomedical journal articles to free availability on the Web as PubMed. Use of the database increased 100-fold once it became freely available. The potential scope of this usage could never have been anticipated by looking solely at use of the controlled-access version. Who are these new readers? They surely include scientists around the globe at institutions that may not be able to afford needed journals. They also may be researchers in unexpected fields, search engine users who didn't realize previously they could use work in a seemingly unrelated field. They may be students, patients or their families, physicians, community health workers, or others from the general public: taxpayers who finance so much biomedical research."

As early as 2004, Donald Lindberg, then-director of the National Library of Medicine, reported that the NLM's OA web site had more than one million visitors per day and "close to a billion a year. ... A good, heavy part of that are
consumers." Quoted in Gene Koprowski, "The Web: Patients heal themselves online," United Press International, August 14, 2004.


For newer evidence since my book appeared, see my updates and supplements for p. 117.


For still more, see the tag library for "oa.lay" at the Open Access Tracking Project (OATP).

All OATP tag libraries are crowd-sourced and updated in real time. You can make them more complete by taking part as an OATP tagger. See the OATP home page for more detail.

#oa #openaccess

Here's my paraphrase of one thesis in a fascinating new paper by Alessandro Iaria, Carlo Schwarz, and Fabian Waldinger:

WWI reduced the flow of new scientific knowledge into Germany, Austria-Hungary, Ottoman Empire, and Bulgaria. This was not just from violent disruption, which also affected Allied countries, but also from a deliberate scientific boycott of the Central Powers by the Allies. Compared to pre-War years, researchers in the Central Powers had less access to research from other countries, and roughly unchanged access to research from their own countries. This resulted in a decline in their productivity, as reflected in several measures including a drop in international prizes, a drop in patent applications, and a drop in the similarity of journal-article titles (measured by algorithm), suggesting a drop in cooperation. This is a new kind of evidence that science is international. It's also indirect support for one of the common arguments for OA, otherwise difficult to document empirically, namely, that enhancing access to research enhances researcher productivity.

Regret: One of the authors, Waldinger, was at my institution today to present these results, and I couldn't attend because of a conflict. I hope this post is partial compensation!

I like the new Clarivate-Impactstory partnership for several reasons.

I like the Impactstory oaDOI initiative, and the Unpaywall tool built on it. I like the idea that Clarivate data will now help users click through to much more OA content than before. I also like the fact that Clarivate is funding this initiative and not just supporting it with data. Congratulations to all.

However, the Clarivate PR team couldn't just describe the benefits of the new partnership. It inserted this passage into the press release:

Researchers conducting online searches for scholarly articles frequently get unreliable results that can compromise their work. This is typically because the results omit journal articles behind paid-subscription paywalls or because "web-scraping" utilities return versions of articles that are not peer-reviewed or are in violation of copyright laws....

It's true that search results can be unreliable because they omit paywalled articles. But there are a few problems with the rest of the passage. I'm sorry to qualify my applause for the announcement by highlighting these problems, but I must say a few words about them. Note that they're about a careless passage in the announcement, not about the partnership itself.

* The sentence on web-scraping utilities is obscure. Because it mentions articles that are not peer-reviewed, it seems to be an oblique criticism of preprint repositories. But preprint repositories depend on voluntary author deposits, not web scraping. Moreover, finding preprints in a search is a feature for people who know how to use them, not a bug. It doesn't make the search less reliable. The criticism misses the target.

* Perhaps the reference to web scraping is an oblique criticism of Sci-Hub. But Sci-Hub focuses on refereed postprints, indeed versions of record, not unrefereed preprints. Moreover, it depends on downloads, even if illicit, not web scraping. The criticism misses the target.

* The final part implies that finding illegal copies of peer-reviewed articles in a search makes the search unreliable. This is false. The writer probably meant to criticize these copies for infringement, but instead criticizes them for unreliability. The criticism misses the target.

#oa #openaccess #preprints #sci-hub #copyright #clarivate #impactstory

Post has attachment
Thanks to Helen Wong for updating the Open Access Directory list of #openaccess publication funds.

See whether your institution has a fund to pay APCs at fee-based OA journals, or BPCs at fee-based OA book publishers.

If your institution has a fund and we don't yet include it, or if our information isn't up to date, please let us know or add the information yourself. The OAD is a wiki and depends on the community to keep it accurate, current, and comprehensive. It's crowd-sourced and distributed under a CC-BY license. To limit spam, editing is limited to registered users, but registration is free and easy. Reading and reuse are free for all.

#oa #openaccess #oad

Post has attachment
Dog food bowl of ice water outside Toscano's restaurant on a hot day in Cambridge, MA.


I wanted to visit your Open Access Curricula for Researchers and Library Schools at this URL:

But my security software warned me of a risk:

Threat found....
Threat: HTML/SEOSpam.A trojan

I don't have the time or expertise to figure out whether this is a false alarm, and so far I've had no reason to distrust this security software (Endpoint from ESET, picked by my employer). So I'm erring on the side of caution, and not clicking through.

UNESCO: Please fix this. The last thing you want to do is exclude conscientious users.

Readers: If you want to see a copy of the UNESCO Open Access Curricula for Researchers and Library Schools at a safe site, here's one that doesn't trigger a security alert.

#oa #openaccess #trojan #malware #unesco
Wait while more posts are being loaded