Peter Suber on digital preservation of Open Access Journals

Doubts about preservation

“Digital documents last forever –or five years, whichever comes first.”
–Jeff Rothenberg, 2001

The more important the work you publish, the more important it is to preserve it.

Fortunately, serious preservation options are neither expensive nor difficult, at least for the journals being preserved, as opposed to the services preserving them. Several of the best are free of charge, both for the journals and for users who want to access the content.

If you held off on making a preservation plan, fearing the expense and complexity, it’s not too late to adopt one now. I almost said “it’s never too late”, but that is exactly what we cannot say.

E-only TA journals face the same challenge. The difference is that TA journals have more options because they needn’t insist on preservation methods that provide OA to the preserved content.

I recommend LOCKSS (OA at least for the hosting institutions) and the DOAJ preservation system (OA for all). Both are free of charge for the journals.

Another free and stable option is to deposit your digital content with a trusted, financially secure, and well-equipped meatspace library. For example, BioMed Central has been depositing its digital output since 2003 in the National Library of the Netherlands, just as the DOAJ preservation program does today.

Many other publishers deposit their digital contents in the British Library.

National libraries and university libraries with digital preservation programs could render a great service by opening their vaults to peer-reviewed ejournals (OA and TA).

BioMed Central and the Public Library of Science routinely deposit their articles in PubMed Central. This greatly increases the likelihood that users will be able to find them again, and find them OA, if the publishers went out of business. This is smart and easy. I recommend routine deposits in a suitable repository for all OA journal publishers.

(Similarly, I recommend that authors avoid hybrid OA journals which do not allow deposits in a repository independent of the publisher.)

However, repository deposit should not be a journal’s only preservation plan. Peter Hirtle has shown why.

But repository deposit can still provide an extra layer of security for ejournals, both before they work out long-term solutions and then later alongside long-term solutions. Among other benefits, this can free new OA projects from the delay of making ironclad preservation a precondition of launch.

A decisive way to answer doubts about digital preservation is to make printouts and preserve the paper. As I argued in 2002:

So far, paper is the only commonly used medium that we know can preserve texts for hundreds of years. There are many creative methods emerging for storing digital texts electronically with at least the security of paper….The only problem is that it will take hundreds of years to monitor the outcome of present-day experiments. But we don’t have to choose between insecure storage and retreat from the digital revolution: the shortcut to preservation is to print digital texts on paper….Preservation in the digital era [can] be as good as paper, just as it was before the digital era.
Microtome is one company that offers paper-based preservation to OA journals.

Finally, let me repeat the point from #2 that libre OA facilitates preservation by granting permission in advance for migrating digital files to new formats and media to keep them readable as technology changes, and for copying files for multiple deposits. In fact, the built-in reasons why repositories (as opposed to journals) find it difficult for to offer libre OA lie at the basis of Hirtle’s argument that repositories are inadequate for preservation.

In some countries, like the US, copying for preservation is allowed by a special provision of copyright law. But this sort of copying is not allowed in all countries, and not part of fair use. Without this permission from the statute or an open license, we could only preserve a mass of content by hunting down each individual copyright holders to ask permission. The expense and delay can be deal-breakers. If you’re concerned enough about the future of your content to *want* to preserve it, then you should be concerned enough to facilitate that preservation with an open license.

Some friends of OA object that preservation is not intrinsically part of providing OA. That’s true. But the same could be said of peer review and grammatical sentences. “Separate” doesn’t mean “unimportant” or “incompatible”. Think of preservation as a separate essential, like truth, clarity, and punctuation. You can provide OA without it, but you shouldn’t want to.


Peter Suber on libre Open Access

Excerpt from the SPARC newsletter October 2009

Opening up

How many OA journals listed in the DOAJ use some kind of CC license?

Answer = 637 out of 4,362 = 14.6% (as of October 2, 2009)

The challenge is to get more OA journals to be more open.

Let me review some terminology to help us talk about this issue. Gratis OA removes price barriers but not permission barriers. It makes content free of charge but not free of copyright or licensing restrictions. It gives users no-fee access for reading but no more reuse rights than they already had through fair use or the local equivalent. It’s free as in beer, not also free as in speech. By contrast, libre OA removes price barriers and at least some permission barriers. It lifts at least some copyright and licensing restrictions and permits at least some uses beyond fair use. It’s free in beer and free as in speech.

In most countries in the world today, new writings fall under all-rights-reserved copyrights from the moment of birth. As a result, authors can only provide libre OA to their work by affirmatively waiving some of their rights. If you’re not sure how to waive some rights without losing your ability to enforce others that you retain, don’t worry. Creative Commons licenses are designed for just this job. There’s more than one CC license because there’s more right in the copyright bundle that you might want to waive. For the same reason, libre OA is a range of things, and the range of CC licenses correspond to the most commonly adopted positions within that range. Apart from assignment to the public domain (waiving all rights), the most open CC license is CC-BY, which waives all rights except the right of attribution. The CC-BY-NC, by contrast, waives all rights except the right of attribution and the right to block or control commercial reuse.

OASPA recommends the CC-BY license or equivalent for all its members. It accepts some less open licenses, such as the CC-BY-NC, but it does not recommend them. In any case, both provide libre OA. OASPA also *requires* that members who want to impose any restrictions on reuse must make the restrictions explicit, for example through a license. (See the OASPA bylaws, Appendix II.)

The SPARC Europe Seal program requires the CC-BY license and libre OA.

SURF recommends the “the most liberal Creative Commons licence” for articles, which is CC-BY. For data it recommends the more liberal assignment to the public domain, as required by the Science Commons Protocol for Implementing Open Access Data. On both fronts it recommends libre OA.

The Budapest, Bethesda, and Berlin statements all call for libre OA, without naming specific licenses.

The challenge is that more than 85% OA journals now in the DOAJ don’t use any kind of CC license. Some of these might use equivalent non-CC licenses and some may use homegrown language with a similar legal effect. But these exceptions are rare. No matter how you slice it, most OA journals are not using open licenses. Most are operating under all-rights-reserved copyrights and leaving their users with no more than what they already had under fair use. Most are not offering libre OA.

(Footnote: Just to clarify some points often misunderstood: All CC licenses provide libre OA, although some are more open than others. But no CC licenses are required for libre OA. You can provide libre OA through an equivalent non-CC license or homegrown language instead, if you like. But libre OA requires some kind of open license. When a copyrighted text doesn’t use any sort of open license, then it falls under an all-rights-reserved copyright, which shrinks user rights down to fair use and rules out libre OA.)

If 14.6% of the journals in the DOAJ use some kind of CC license, how many use the CC-BY license recommended by OASPA, SPARC Europe, and SURF?

Answer = 416 out of 4,362 = 9.5% (as of October 2, 2009)

How many use CC licenses other than CC-BY? Answer = 221 out of 4,362 = 5.1%

(Footnote: The DOAJ doesn’t actually count journals with CC-BY licenses. It counts journals with the SPARC Europe Seal, which requires CC-BY licenses. But the Seal also requires journals to share metadata in a certain way. Hence, it’s possible for many journals to use CC-BY and fail to earn the Seal because they don’t share their metadata appropriately. In that case the SPARC Seal tally would undercount the journals using CC-BY. But in fact, many more DOAJ journals share their metadata than use CC-BY, making the Seal tally a good approximation to a CC-BY tally. Thanks to Lars Björnshauge for this detail.)

I know that many OA journals want to restrict commercial reuse and resist the recommendations to use CC-BY. I don’t want to enter that debate here. If you want to restrict commercial reuse, then use a CC-BY-NC license or equivalent rather than no open license at all. Join the 5% of DOAJ journals which have adopted that solution.

I once talked to an OA journal publisher who feared that open licenses would deter submissions. It turned out he feared that allowing commercial use would deter submissions and assumed that all open licenses allowed commercial use. Moreover, he only had a few anecdotes to support his theory about the effect on submissions. But whether or not his fear was groundless, there’s a simple way to use an open license and restrict commercial use. Just use a CC-BY-NC or equivalent. Let’s agree to move beyond an all-rights-reserved copyright to libre OA, and argue later about whether to move beyond CC-BY-NC to CC-BY.

If a journal is already free of charge, then why use any open license at all? Why move beyond gratis OA to libre OA?

The short answer is to spare users the delay and expense of seeking permission whenever they want to exceed fair use.

Why would users want to exceed fair use? Here are some of the answers I gave to Richard Poynder in October 2007 (p. 37):

[T]here are good reasons to exceed fair use, for example, to quote long excerpts, print full-text copies, email copies to students or colleagues, burn copies on CDs for bandwidth-poor parts of the world, distribute semantically-tagged or otherwise enhanced versions of a text, migrate copies to new formats or media to keep a text readable as technologies change, archive copies for preservation, include the work in a database or mashup, copy the text for indexing [or] text-mining…, make an audio recording of the text, or translate it into another language.
Fortunately it’s easy for OA journals to adopt open licenses and permit these uses in advance. It’s easy free their users. It’s easy to provide libre OA. In fact, it’s easier for OA journals than OA repositories (or easier for gold OA than green OA). OA repositories generally stick to gratis OA because they can’t generate the needed permissions for libre OA on their own. But OA journals can generate the permissions on their own.

Some OA journals know that CC licenses are free of charge and easy to use. Their reservations are less straightforward. I once talked to an OA journal editor-publisher who said: “Our journal is already free of charge. What else would anyone want? If users want to copy and redistribute the files, they are free to do so. We don’t care. We allow that.”

How would users know that the journal allows them to copy and redistribute the files? Fair use does *not* allow that, and conscientious users will limit themselves to fair use. When they don’t have permission to exceed fair use, they will either slow down to ask permission or err on the side of non-use. These are impediments to research that OA was designed to remove.

Some people believe that, for better or worse, conscientiousness about copyright is a scruple fading into quaintness, like sexual prudery. But it remains the formal policy of every university and library in the world. Even when individuals are not conscientious, their institutions are, and they require their affiliated users to be. Even if you can wink at individuals to let them know that they may do what they want with your files, institutions are not allowed to take the hint. They can exercise their rights under libre OA, when they know they have them. But they can’t act on a wink alone. And if you really are willing to provide libre OA, then there’s no reason whatsoever not to say so in an explicit statement or license.

Even when users want to do something allowed by fair use, they have to deal with the fact that fair use itself is vague and contestable. For example, informed people disagree about whether it covers text-mining. Another reason to use an open license, then, is to free users from the fear of liability and from self-imposed restrictions arising from that fear. This will benefit not just the users who need libre OA, but also benefits the users who want to do something lawful, though not widely known to be lawful, under fair use.

Don’t make conscientious users choose between the delay of seeking permission and the risk of proceeding without it. Don’t make them ask permission. Don’t make them pay for permission. Don’t make them err on the side of non-use. Don’t increase the pressure to make them less conscientious. Don’t free your users to exceed fair use without telling them that they are free to exceed fair use. If you think you’re providing libre OA but don’t use an open license or equivalent language, then you’re not providing libre OA. You’re using an all-rights-reserved copyright, perhaps yoked with a private resolve not to enforce it, forcing conscientious to seek permission. Your private resolve not to enforce your legal rights doesn’t free anyone from the need to learn what their own rights are. But your public resolve through an open license will do so simply and elegantly.

ALO alt abgeschaltet

ALO hat seine frühere Oberfläche abgeschaltet mit der unerfreulichen Konsequenz, dass Links auf einzelne Seiten dann nicht mehr funktionieren, wenn man einen Popup-Blocker aktiviert hat.

Wenn man auf eine Metadatenseite (meine “Gmünder Chroniken”)

kommt, erfährt man nicht, dass man, um das Buch im Viewer zu sehen, auf den Titel klicken muss.