The Bookseller reproduced this address by the CEO of Cambridge University Press on 14 May, under the headline “Different strokes”.

Far too much discussion of scholarly communication is bedevilled by generalised assertions about “what students want”, “what researchers need”, “what librarians require’”, “what governments determine”, when in practice the world is pretty diverse.

For librarians, the latest Ithaka Library Survey has rammed home that what they want is not at all uniform: what drives young and committed life scientists studying at Stanford or Cambridge isn’t necessarily what drives all researchers under all circumstances.

One obvious area of divergence is take-up of digital. Print has remained important for longer than many had expected, and not just in territories such as Japan, where both institutional and cultural frameworks have slowed digital adoption. It’s striking that most academics find journal articles digitally but tend to read hard copies of those longer than half a dozen pages. It is commonplace that the more influence a university’s faculty has over acquisition policy, the slower the take-up of digital books in the university’s library.

I’m sure generational shift has started to change that, and that lack of digital availability means invisibility to the consumer. But at Cambridge University Press we aren’t at all unusual in still having the majority of our global academic book sales in hard-copy form, despite the rapid growth of digital. There was a big feature in the Times Higher Education Supplement in March on “Publishing Your First Book”, in which various senior academics were canvassed for their dos and don’ts: what struck me was the resolute assumption by everyone that this “first book” would be a printed artefact first and foremost. Digital availability wasn’t mentioned by a single respondent. That’s changing, but there’s some way to go.

That flags another divergence between humanities and sciences and the differing fortunes of long-form research books. For many arts and social science scholars, it is the research article or book itself where the value lies, as opposed to the applications or replications to which that research might be put (which lies behind the dominant science perspective).

Schizophrenic views

Incidentally, no fewer than half of the members of the History faculty of Cambridge University enjoy the services of a literary agent, primarily engaged in selling the fruits of their research and their content (the two being one and the same). This divide has been at the heart of much of the recent debate between historians and life scientists. It also informs the frankly schizophrenic views of many universities about Patent Law (of which universities are generally warm and protective supporters) and Copyright Law (against which some universities rail, as restrictive, expensive and protectionist).

What’s clearly endangered is the long-form scientific research book, whether in print or online. The incentives for scientists to write long-form content are clearly in decline. Pressure on scientists to produce more frequent outputs is one cause. Another is the need to get work out rapidly because of global competition. The English language is now pretty dominant in advanced scientific communication (though not at all in arts and social sciences). That may also discourage some non-native Anglophone scientists from writing 300-page monographs. It is quite possible that within the next five to 10 years, long-form science research publication—outside the arenas of textbooks and popular science—may be a tiny fraction of what it is today.

There are more general questions about the future of academic books. We’re taking part in several Open Access (OA) experiments, but I don’t yet see signs of culture and practice changing much.

The final area of divergence to highlight is around OA journals. Rick Anderson, a librarian at the University of Utah, gave a recent thoughtful address to the Smithsonian on OA. He was trying to bring some cool and dispassionate sanity to what remains all too often an area of messianic claim and counter-claim. One thing Rick is surely right to emphasise is that OA publication is at present a minority activity. One abiding reason is that for many tenured faculty in established institutions, the current system isn’t broken sufficiently to provoke them to the personal inconvenience of changing their own publication practices. It is very striking—and entirely understandable—how much of the advocacy pressure for OA has come from those who feel themselves disenfranchised by the current system, and from those who feel that the current system only speaks to, is only accessible to, those within the ranks of the resource-privileged. The latter perception matters a lot more, it’s self-evident, to some faculty members than to others.

There is also marked variation in the take up of OA by subject. Within science journals, life sciences are clearly leading the way, driven by funders’ requirements, but certain areas have barely moved—chemistry, for example. And in subjects driven by individual endeavour rather than large grant-funded teams—such as humanities, history and my own field of mathematics—peer-reviewed OA remains a very small proportion of published articles.

Finally, it’s striking who OA payments are going to. One of the first major analyses of OA funding was published by the Wellcome Trust a few weeks ago. Four publishers accounted for more than half of Wellcome’s total funding for article processing charges in 2012/13. It’s possible gold OA will lead more to market consolidation rather than disruption. Nonetheless, the tide of governmental and funder sentiment—particularly in the UK—is flowing in one direction only. The challenges this poses are considerable, especially in book-based disciplines in the arts and social sciences.

Peter Phillips is chief executive of CUP. This is an edited extract of his address to the 16th Fiesole Collection Development Retreat