While open access is all about setting research ¡°free¡±, the transition towards it can feel to academics like just another facet of the ever-stricter assessment regimes to which they are becoming subject.

A recent example in the UK is the ³ÉÈËVRÊÓƵ Funding Council for England¡¯s announcement that all articles and conference papers submitted to the next research excellence framework exercise will have to be available on an open-access institutional or subject repository. Books are exempted for 2021, but they won¡¯t be the next time around.

To its enthusiasts, open access is an incontestable good, for books as much as for papers ¨C and those of us who point out the considerable difficulties in implementation are dismissed as reactionary Luddites. Recent heavyweight UK reports on open-access monographs, such as the 2015 and the are less gung-ho, but both urge measured and rigorously evaluated progress towards open access for books.

Everyone writes to be read, so wider access to our books is, on the face of it, an unalloyed good. But we should not ignore the downsides. The Indian literary scholar Sukanta Choudhury, for instance, calls open access a ¡°doubtful panacea¡± as it privileges readers in the developing world with good internet access, and potentially disadvantages authors who cannot pay the article processing fees associated with publisher-provided gold open access. For Choudhury, open access could perpetuate or even exacerbate the differential in academic productivity between global North and global South.

³ÉÈËVRÊÓƵ

One major argument in favour of open access advanced by both policymakers and enthusiasts is that research has typically been paid for by institutions or taxpayers, so neither should be expected to pay publishers to access its outputs. But this is only partly true. Although it is not always acknowledged, publishers add considerable value to academic research. A 2016 blog by the publishing consultant Kent Anderson on the ?website, for example, details 96 things that publishers do.

Many scholars are broadly supportive of the benefits of open access, but they often display a degree of disquiet when all the implications are pointed out to them. One problem is the liberality of some of the Creative Commons licences that open-access authors are required to sign. At their most permissive, these allow almost any reuse of a work, including adaptation and abridgement, as long as the original author is credited. This is particularly hard to swallow when it comes to monographs. In the arts and humanities, scholars do not regard the research embodied in their books as data to be mined, but, rather, as carefully crafted arguments: creative works in their own right. And they view with dismay the prospect of those works¡¯ atomisation and appropriation by others.

³ÉÈËVRÊÓƵ

Then there is the question of where and how open-access monographs are published. Repository-based green open access is unlikely to be the solution because the version of a work that can be made available in a repository is the accepted manuscript. In the case of books, this can differ markedly from the final publication. It can also be difficult for readers to find things in the repositories if they don¡¯t already know they are there.

That leaves you with gold open access. But this immediately raises the issue of academic freedom: will monograph authors be pushed towards publishers that are not their first choice simply because they are open access? The UK funding bodies insist that REF assessment panels are publisher-blind, but academics and their institutions continue to value long-established and reputable houses, whether they offer easy routes to open access or not. Brand is and will remain a powerful incentive in the choice of publisher.

Business models and funding are another problematic area. Those publishers that offer open-access monographs currently charge authors upwards of ?10,000. Some 8,000 monographs were submitted to the 2014 REF. Even open-access enthusiast Martin Eve, co-founder of the open-access humanities platform Open Library of the Humanities, noted in a that making all those books open access would have accounted for the entire purchasing budget of UK research libraries. Eve still thinks that a solution can and must be found, and there are interesting experiments by new, open-access-only university presses in partnership with university libraries and by smaller publishers. But these are comparatively small, with no implication that they can work at the scale required for compliance with the future REF requirements. Gold open access for all books submitted to the REF would require nothing less than a revolutionary change in the publishing industry: something that can be achieved only at enormous cost and risk, and will probably take many decades.

We could, of course, just write publishers out of the picture at a stroke. We could switch to entirely new, open-access platforms controlled by institutions and funders. But academic publishing is an industry that serves us well. It is not Luddism to advocate that we should refrain from destroying it by rushing pell-mell towards openness without a reasoned appreciation of the difficulties as well as the advantages.

³ÉÈËVRÊÓƵ

Marilyn Deegan is professor of digital humanities and honorary research fellow at King¡¯s College London.