With the almost universal availability of the internet, the old publication model for science becomes less and less efficient. Before the internet, there was no good alternative to the model where a fixed number of copies are printed, sold and shipped to potential users, mostly scientific libraries, where they are available indefinitely. But with the internet, distribution no longer requires printing and shipping, nor does one need an estimate of how many readers a publication will have. This should make publication much cheaper than in the past, and it should be possible to move to open-access publication, so that the scientific works are freely available to any internet user.
What science authors still need from science publishers is the following:
(1) selection on the basis of peer evaluation
(2) indefinite online availability (= user-friendly website + archiving)
(3) creation and maintenance of a prestigious brand
What would be desirable but is not absolutely necessary is the following:
(4) proofreading and typesetting, for a pleasant visual impression
(5) connection of the website to a print-on-demand service
Print-on-demand is necessary only for online book publication, because for journal papers, it is fully acceptable if they are made available only electronically and printed ad hoc. However, (1)-(4) apply to journal publication as well as to book publication.
In the following, I say a few more things about each of (1)-(5), before concluding on what scientists need to create an open-access book series.
For journals, there is no change between the old model and the open-access model, because the selection is done by the journal editor, and the peer evaluation is done by colleagues as a free service to the profession.
For book publication, peer evaluation is often remunerated in the old model. Reading and commenting on a book manuscript is more demanding than evaluating a paper manuscript, and this contributes to the cost of publication.
However, the remuneration is not the primary motivation for scholars, so peer evaluation will also work without this component if there is a committed book series editor with good connections to other scholars in his/her field.
Selection of books for publication is often done by the publisher rather than a scientist in the old model, but this aspect of the old publishers’ activities is dispensable: Publishers are not really qualified to make the selection, and often they select on the basis of commercial viability rather than scientific quality, which is irrelevant to science.
In open-access book publication, the publisher has no role in selection. The driving force is the scientist, and the publisher is merely providing technical services to them.
Thus, in open-access book publication, every book will appear in a book series that is the responsibility of a scientist, just as every paper appears in a journal edited by a scientist.
3. Online availability
Open-access books and journals must be readily available online, and authors must have trust that they will be available indefinitely.
For indefinite availability of books, libraries have long provided a well-known model, whereas there is no well-established model for long-term availability of online content. The Deutsche Nationalbibliothek does not reliably store every open-access book, or at least it does not make stored books available the way it makes printed books available of which it has physical copies.
Nevertheless, it seems clear that the long-term availability of open-access online content should eventually be the responsibility of scientific libraries. So although this aspect of open-access publication is not fully clear at the moment, we can probably trust that libraries will take on this responsibility in the near future. Keeping copies of physical books is not cheaper than keeping copies of open-access books, so this should be possible without an addition burden on the budget of libraries.
In the meantime, any other organization can play the role of “publisher”. Note that for legal purposes, the creation of a web site counts as publication, so open-access books are “published” even if they just appear on the author’s website and there is no institution involved. However, due to the requirement of branding (see next section), the (name of the) publisher will probably be relevant in open-access book publication as well.
One difference between books and journals is that with books, the brand prestige typically comes from the publisher, whereas with journals, it comes from the journal title. This has to do with the way in which we cite journal papers and books: With just the journal title, or with publisher and place of publication.
Assuming that the citation model is not affected by the shift from the old model to open access, what one needs to do for book publication is to create a prestigious brand for the open-access publisher. This should not be difficult, because the prestige comes from the works published, and is transferred from them to the publication place, i.e. the publisher (and secondarily the book series).
So what is needed for a new book series and a new publisher to become established and prestigious is five or six initial books that set the stage. This is perhaps one of the biggest challenges of the enterprise: Come up with five or six authors who are willing to publish their books in a new series with a new publisher, which is not yet fully established.
In addition, it is important to think carefully about the name of the publisher and of the book series, but this presents no particular difficulties.
Publishers often claim that their marketing activities are important for the success of a book, but this is no longer the case. Marketing may still be relevant to persuade people who do not need the book to buy it, but open-access books in science will find their readers if they are good books. There are now many cheap ways of disseminating information (e-mail lists, blogs, Facebook, Twitter), so science e-books do not need special marketing.
5. Copy-editing and typesetting
Traditional book publication by high-profile publishers invests a lot of resources into copy-editing and typesetting. Especially in the humanities, this aspect of their work has been valued highly, but as powerful word processors have become widely available, it has become easier and easier for authors themselves to produce quasi-typeset versions of their work, or if they use LaTeX, to produce works whose typesetting quality is difficult to distinguish from professionally typeset works.
Copy-editing, by contrast, is much less important. Traditional copy-editors are generally not specialists of the scientific field, and they may actually introduce errors. But of course, the more people read a work and spot typos or other minor errors or style inconsistencies before publication, the better.
In addition to typesetting, in modern e-book production the use of widely usable formats such as Epub will become more and more important. PDF files can be read on desktop computers and standard tablets, but for e-ink readers and mini-tablets (and other smaller devices), special e-book formats are required. While this is not yet standard for science publications, it will become important in the near future and should therefore be thought about now.
6. Print on demand
For e-books, unlike electronic journals, a printed and bound copy will probably be used by many readers, at least for the next few years, as long as e-book-reading devices are not universally available yet.
Ideally, the e-books should be linked directly to several print-on-demand services, to allow users a choice between different options.
The print-on-demand service should not be thought of as the primary publisher, only as the service that makes a hard copy for traditional paper reading.
To create a high-profile open-access book series for scientific books, scientists need
(i) a network of colleagues supporting the enterprise and willing to serve as reviewers without remuneration (§2)
(ii) a technical service (ideally provided by an academic institution such as a university library) that takes care of web hosting and ideally of website creation and maintenance (§3)
(iii) student assistants, or secretaries, or junior collaborators, who help with proofreading and typesetting (§5) and the coordination between the scientists and the technical service
Assuming that scientists already have their network, the main additional needs are (ii) and (iii).