OA could be implemented badly so that it infringes copyright.1 But so could conventional publishing. Both OA and toll access have long since discovered the same recipe for avoiding copyright problems: For sufficiently old works, rely on the public domain, and for newer works under copyright, rely on copyright-holder consent. This shouldn’t be surprising. Toll-access publishers don’t have a shortcut to copyright compliance just because they charge money for access, and OA publishers don’t face an extra hurdle to copyright compliance just because they don’t charge for access. Copyright protects the revenue streams of those who choose to charge for access but doesn’t compel anyone to charge for access.
When researchers publish in OA journals, the permission problem is easily solved. Either the author retains the key rights and the publisher obtains the author’s permission, or the author transfers the key rights to the publisher and the publisher uses them to authorize OA.
Toll-access journals don’t make their articles OA, of course, but more often than not they give blanket permission for authors to make their peer-reviewed manuscripts green OA. (See section 3.1 on green OA.)
When authors transfer all rights to the publisher, then they also transfer the OA decision to the publisher. When the publisher doesn’t already allow green OA, then authors must ask permission to make their work OA. However, many publishers who don’t give blanket permission for green OA will agree to case-by-case requests. (For example, before Elsevier started giving blanket permission in 2004, its policy was to agree to essentially all case-by-case requests.)
When authors submit work to toll-access journals but retain the right to authorize OA, then the OA decision belongs to them. Publishers may refuse to publish their work, of course, but they seldom do so merely because of rights retention when authors are following a policy of their funder or employer. As noted (in chapter 4 on policies), the NIH has one of the strongest rights-retention OA policies anywhere, and to date not a single surveyed publisher refuses to publish NIH-funded authors on account of its mandatory OA policy.2
Publishers who refuse to publish rights-retaining authors are not asserting copyright. They are asserting an independent, background right to refuse to publish any work for any reason. (I support this right and would never want to see publishers lose it.) Authors who retain rights don’t violate rights belonging to publishers; they merely prevent publishers from acquiring those rights in the first place. When rights-retaining authors make their work OA, publishers can’t complain that OA infringes a right they possess, only that it would infringe a right they wished they possessed. Publishers who face rights retention face hard bargaining, not infringement. Publishers still have a remedy, but it’s the remedy to hard bargaining (just say no), not the remedy to infringement (sue or threaten to sue).
We can see this from another angle. If the NIH policy violated copyright law, publishers would have sued. But instead, their strongest response has been to support a bill amending U.S. copyright law to make NIH-style policies unlawful. That’s a concession that the NIH policy is lawful under current law. In that sense, strong rights-retention policies are not only lawful but battle-tested.3
Of course authors may retain rights on their own, even when not required to do so by a funder or university policy. But when authors stand alone, they have very little bargaining power against publishers who demand the rights as a condition of publication. One of the practical benefits of strong rights-retention policies is that they amplify the author’s bargaining power and tend to elicit publisher accommodation.
When authors retain the right to authorize OA, and use that right to authorize OA, then the resulting OA is authorized by the copyright holder. The fact that the decision is from the author rather than the publisher makes it unconventional, but not unlawful, insufficient, or legally dubious.
Authors who retain the right to authorize OA may still transfer all other rights to publishers, and typically do. In these cases, publishers may not acquire all the rights they want, or all the rights they formerly acquired. But they acquire all the rights they need for publishing, and they have undiminished power to enforce the rights they acquire.
This solution works because funders and universities are upstream from publishers. In the case of funders, grantees sign their funding contracts before they sign their publishing contracts. In the case of universities, faculty members vote to authorize university-hosted OA to their future publications before they sign their future publishing contracts.
OA journals obtain the needed permission through a publishing contract with the author, just as conventional journals do. But because OA journals aren’t trying to protect sales revenue, they needn’t prohibit copying and redistribution. On the contrary, OA journals share the author’s interest in maximizing impact by maximizing distribution and reuse rights. Hence, OA journals may request fewer rights from authors and allow more uses than toll-access journals do.4
Conventional wisdom holds that authors need copyright to give them an incentive to write. Others can debate whether this is true for nonacademic authors like novelists and journalists. (L. Ray Patterson liked to point out that it wasn’t true for Chaucer, Shakespeare, or Milton.5) But there are two reasons why it’s simply false for authors of research articles. First, authors of research articles are not paid. When money is even part of an author’s incentive, copyright fortifies the incentive by giving authors a temporary monopoly on their work and the revenue stream arising from it. Without copyright, unauthorized copies might kill the market for authorized copies and reduce sales. But all this is irrelevant to authors who write for impact, not for money, and who voluntarily forgo royalties.
Second, authors of research articles traditionally transferred copyright to publishers. Hence, copyrights on research articles traditionally protected publishers, not authors. If the conventional wisdom about incentives were true for research articles, then transferring the rights to publishers would have diminished author productivity. But that did not happen. On the contrary, scholars have always had independent incentives to write journal articles, such as knowledge sharing, reputation building, and creating a portfolio for promotion and tenure. They never expected revenue from their articles, never needed a temporary monopoly on that revenue, rarely even knew what the revenue was, and never wrote for the purpose of generating revenue for the publishers who actually owned the copyrights in their work.
Because scholars don’t earn royalties on their research articles, they would not be hurt by dramatic copyright reforms designed to restore balance between copyright holders and users—not that such reforms are likely any time soon. Publishers who pretend to speak for authors in defending the current imbalance in copyright law speak for authors of royalty-producing literature. Authors of royalty-free literature have very different interests.