'Don't close the definition of "open" for books'

To make all academic research findings accessible to everyone, we must explore what, exactly, we mean by 'open access' for books, writes Matthew Day

Over the past few years there has been a movement towards making the results of publicly funded research freely accessible for everyone to read and use.

Now picking up speed dramatically, this open access movement is stimulating major changes in how academic authors, their institutes, and their publishers approach disseminating new knowledge and ideas.

Open access began about 20 years ago – with the formal definition, justifying O and A capitalisations, being largely settled by 2003. Since then this definition of open access has been most successfully applied to journal articles.

Now books are getting focused attention. Some research funders, particularly in Europe but elsewhere too, are determined to radically increase the pace at which open access grows. Research they fund that is published as books may in many cases need to become open access. However, current approaches to open access for journals cannot work for books at a large scale.

If we apply open access to books in the way it is applied to journals, we will fail. If the failure is simply that books do not become more ‘open’, that would be one thing. But it is possible that academic researchers will find themselves required to publish books in ways that will be unsustainable for academic publishers. For Cambridge University Press, where I work, if our books earned only a few percentage points less revenue than they do now, our books programme would become loss-making. Academic books are a vital part of many researchers’ lives and careers. We must not put them at risk.

I’m not arguing for whether open access principles should or shouldn’t be applied to books. The idea, like most interesting and far-reaching ideas, is contentious. But if open access principles are to be applied, then we should maximise the chance of them working.

I’ve no doubt that publishers will explore ways to approach the challenge, especially once the rules are established. But their ability to explore the options, test ideas and home in on the successful approaches will be severely hampered if a tight definition of open access is transferred to books from journals. Unless funders are able and willing to cover the full costs of a books publishing programme (which they probably aren’t), publishers will need to secure additional revenues to support open access books. A definition of ‘open’ for books will therefore need to focus on content being freely readable while relaxing the requirements for allowing re-distribution and re-use. 

Publishers and others have already started exploring options for making content more accessible. For example, at Cambridge University Press we have a service called Cambridge Core Share for sharing journal articles with no paywall or other barrier to any reader. Certainly, Core Share does not meet any formal definition of open access, and therefore it isn’t a way for researchers to comply with their funders’ open access policies. But Core Share does make articles more accessible: We’ve seen articles getting tremendous traffic from public discussion forums that they would never have had without Core Share.

Approaches such as Core Share might or might not work for books. But one thing is clear, if the definition of open access as applied over the last 20 years to journal articles is transplanted directly onto books, then there will be little room to find something that does work. And if the even tighter definition that some funders are now considering for journals is also applied to books, then there will be practically no room for experimentation.

There are many differences between books and journal articles, whether looked at from the perspective of an author, the publisher, a librarian, or a reader. The definition of open access will need to be different too.

Matthew Day is head of open research policy and partnerships at Cambridge University Press