A scholar from late seventeenth-century London teleported from the Royal Society into a modern library of any of the world’s universities would find themselves in a strangely familiar world. Clutching a copy of the first issue of Philosophical Transactions: Giving Some Accompt of the Present Undertakings, Studies, and Labours of the Ingenious in Many Parts of the World (1665) in one hand and a copy of Isaac Newton’s Principia Mathematica (1687) in the other, our bewigged scholar might find that the technology of the modern library might take some getting used to, but they would immediately recognize that the organization, collection, and curation of knowledge in either serials or books was largely unchanged from the world they had left behind.

The domination of scholarly communication by a binary choice of journals and books for over three and half centuries is something that we largely take for granted, and is now hardwired into all the editing, production, cataloguing, indexing, purchasing, and measuring systems that have grown up to support the organization of this accumulated knowledge for the benefit of scholars and administrators. Taking some very crude average lengths for journals and books, a back-of-the-envelope calculation suggests that an improbably high percentage (over 75 percent?) of scholarly writing is either less than 10,000 words or more than 100,000 words. It is difficult to believe that this is an optimum outcome.

It is not surprising therefore that we have seen a proliferation of experiments in recent years. Alternative formats of scholarly communication—ranging from short-format books such as Palgrave Pivots, Stanford Shorts, and Springer Briefs to extended blog posts—have provided opportunities for mid-length scholarly writing. (Note: “Gray literature” has always ranged in length, from the multivolume reports of the Intergovernmental Panel on Climate Change to, for example, a short ten-page World Health Organization report tracking the spread of Ebola in Central Africa.) At the same time, the rise of preprint services and open peer review challenge the traditional gatekeeping and curation role of journals.

Along with colleagues from the books and journals divisions at Cambridge University Press, we asked ourselves the question: In a digital environment, if you could combine the best features of books and journals, what would it look like?

To start with, we listed the sorts of things that underpinned the popularity and success of journals:

Trust

The combination of recognized scholars as editors and rigorous peer review are widely respected as strong quality filters, with publication in top-tier journals seen as a valid proxy for evaluation of an individual’s research ability.

Length

Over time the 5,000-to-10,000-word limit has proven itself an effective length to test and report on discrete arguments, as it requires authors to focus on the essential propositions, helping to make the analysis both more rigorous and transparent.

Speed of publication following acceptance

With online publication now typically occurring within 10 to 12 weeks of acceptance of the final manuscript, journals are able to communicate peer-reviewed research in a fast and timely manner.

Innovation

The size and value of the industry and the centrality of journals in scholarly communication has attracted investment in innovation, with an ever-increasing variety of services being developed to help authors, readers, and librarians (e.g., “Code Ocean” for embedded data and code, “Publons” for integrated peer-review crediting services, automated integration with preprint repositories, and services such as “Overleaf” for use of integrated authoring/collaboration tools).

What journals don’t do so well

On the other hand, we noted that the strict word limits constrain the development of wider arguments and that there was an inherent conservatism in the peer-review and acceptance criteria adopted by many leading journals. In particular, it is possible that the organization of journals is not hospitable to interdisciplinary or mixed-method research. There is also some concern, as the late Nobel laureate Ken Arrow wrote, that journals, at least in his field of economics, “reinforce existing paradigms” as “the publication selection procedure [. . . ] has become methodologically more conservative, more given to preferring small wrinkles in existing analysis to genuinely new ideas.” Finally, although publication of accepted manuscripts is generally rapid, the acceptance decision is frequently a very extended one, with many authors waiting up to a year before knowing if their article has been accepted.

The pros and cons of the long-form book format

We repeated the exercise for books. On the positive side, it was clear that the flexibility over length gives space for mature reflection, based on years of research, teaching, and refining the argument in journal articles and conference presentations. The length allows authors to “join the dots” and develop a wider-ranging argument than is possible within a journal article, and there is likewise more scope to reflect in depth on the existing literature and to accommodate interdisciplinary perspectives. Attention to rigorous peer review and quality control ensures a high degree of trust and respect for the best academic lists and series.

On the negative side, all of the above represents a big investment of time, which, outside of the humanities and certain social science disciplines, is generally not well recognized by tenure and promotion committees. Many scholars feel the cost-benefit of investing time in writing a book does not stack up. I have lost count of the conversations I have had with scholars who said that they really wanted to write content that was longer than a journal article but significantly shorter than a full-length monograph.

As we get to understand usage better, it is clear that very few books are read cover to cover, suggesting that the long-form tradition of book writing may in fact in many instances disguise a product that could be understood as a sequence of modular content that happens to be packaged together as a single book.

The book industry is finally playing catch-up with journals in terms of indexing (the Web of Science Book Index indexes 60,000 titles—a relatively small percentage of the research that has been published in book format). In an age in which abstracting and indexing services play such an important role in promoting visibility of content, this is a serious drawback for books.

Imagining the best of both worlds

In our thought experiment, we imagined that any hybrid book/journal format would be primarily a digital product, assuming we would want to include the full range of functionality that is now available in an increasing number of journals, such as the ability to accommodate video and audio files within the text or a better display of data with potential for more interactive graphics. We also felt that there was an opportunity to make the content dynamic, with regular updates, plus use of annotation tools and social media to create forums for discussion.

Armed with these insights and assumptions, we put together a proposal for a new hybrid format, which we called Cambridge Elements, that we hoped would bring together some of the most desirable features of journals and books. We discussed the new concept at length with librarians at our regular librarian panel meetings to test our assumptions and to craft sales models that took due note of their budget constraints and priorities.

Taking a lead from the way in which knowledge is organized in journals, we decided to organize this new content in distinct series edited by senior scholars and with a strong emphasis on peer review and quality control. We decided to target lengths of between 20,000 and 30,000 words, sufficiently long not to be confused with a journal article and sufficiently short not to be confused with a book. We wanted content that would hit a sweet spot between a review or survey article and original research. We imagined that the main market would be institutional libraries purchasing the content in digital format, but we guessed that many readers would also want to read print versions, so we allowed for print-on-demand paperbacks. For some series, particularly in the humanities, we anticipated some interest from “enthusiasts” outside of academia, and for other series in fields, such as public administration, where there is a strong academic/practitioner dialogue, we are also keen to explore ways we can make content visible and accessible to policymakers.

Many scholars feel the cost-benefit of investing time in writing a book does not stack up.

The response from the academic community to this initiative has exceeded our expectations. Within the space of a couple of years we have commissioned over 60 series, each edited by senior scholars from the best research universities in North America, Europe, Asia, and Australia, with several hundred authors already under contract to write individual Elements. The format has appealed to communities across the arts and sciences, and our series duly range from ancient Egypt to flexible large area electronics, with a particularly strong concentration in philosophy, political science, economics, and applied social and behavioral sciences. The complete range can be viewed here, with each new Element available to view for free for the first two weeks after its publication.

While we are excited by the potential of this format for Cambridge University Press, working through this process has taught us some lessons that I think have wider resonance for those involved in scholarly communications, whether as authors, readers, or publishers.

Mid-length content meets a need

There is a clear appetite among the research community to write content at this length. We were worried that because Elements was an untried format with no impact factor and no guarantee of recognition by tenure committees that early-career, pretenure scholars might be reluctant to commit. But the evidence so far suggests that people are more willing than we anticipated to take the risk, despite the often very conservative nature of the academic incentive structure. As publishers we see this as an encouraging sign and maybe an indication of a wider appetite for more experimentation and more recognition for alternative formats of scholarly communication.

Catering for needs of interdisciplinary scholarship

The format has been especially appealing to scholars working in interdisciplinary fields and with mixed methodologies, or ones needing to provide a higher degree of contextualization and narrative than would be possible in the leading journals in their fields.

Print still matters!

For all that we designed Elements as a born-digital product, authors and series editors still value the print product as a tangible, reader-friendly output of their research. We have probably had more discussions with authors about print and the aesthetics of print than we have had about digital functionality.

Conversely, the ability to update content is not universally welcomed. We thought this would be one of the big selling points to attract authors, but reaction has been mixed; for every author who has welcomed the opportunity, there has been another who has shuddered at the thought of having a life-long relationship with their writing! Furthermore, we are not seeing the appetite to include embedded audio and video content that we had anticipated.

We still live in a binary world

Combining best of books and journals sounds easy, but we have found it to be surprisingly difficult in ways that, perhaps naively, we did not anticipate at the time. Internally, we have very different workflows and data feeds for books and journals. Externally, cataloguing systems still want to know if it is a book or a journal.

Do we continue to be well served by this binary choice? For better or worse, we wanted to signal the possibility of a third option and list Elements as a distinct content type on our website alongside books and journals. I recently enjoyed reading a piece on the challenges the digital humanities have had in defining themselves in which the author reminds us that Boolean logic allow us both to expand as well as narrow our searches. In the same spirit, we hope that Elements will be recognized as a distinctive format offering the advantages of both books and journals. Time will tell if we were right to reject the binary choice.