Afoot and light-hearted I take to the open road.
(Walt Whitman (1856). âSong of the Open Roadâ)
In this first part of the book we want to make some introductory remarks about open access. Here in Chapter 1, we outline some of the key aspects of the past and present of OA. We also make some brief points about possible OA futures. In Chapter 2, we present a more systematic mapping of the main components of OA. Between them, these chapters will be a useful foundation on which we build our understanding of the theory in the OA domain and its relationship with practice.
Our aim in this chapter is to identify the main contours of the OA landscape. We will do this only briefly, bearing in mind the voluminous literature now available on the topic, including several significant book-length overviews (Anderson, 2018; Bartling & Friesike, 2014; Eve, 2014; Fyfe et al., 2017; Herb & Schöpfel, 2017; Regazzi, 2015; Suber, 2012; Willinsky, 2006). However, it is important we provide an initial summary to set the scene for the rest our investigation. We start this chapter by looking at the beginnings of OA, at its key characteristics, and potential benefits. We then discuss its growth since the beginning of the 21st century. We go on to discuss the market within which OA exists, and the enablers of and barriers to change in this market. We finish the chapter discussing possibilities for the future.
Open access beginnings
At the outset, it is worth reminding ourselves that open access is an approach (or set of approaches) aimed at improving the communication of research outputs in order to improve the research endeavour as a whole. Its advocates believe OA can enable significant enhancements to scholarly communication â the ways in which researchers exchange information about their findings with their peers and others. Communicating the outcomes of research is a critical part of the research process itself, one without which research cannot deliver its value. Researchers need to be able to access, read, test, augment, refine, and refute each otherâs work â that is the way research moves forward. Other people beyond the research community can also make various uses of the research literature, not least to inform practice of different kinds. Research communication has traditionally been achieved through a variety of channels, including books, conference papers, and notably, journal articles. All of these different types of research outputs are usually quality controlled in a range of ways, the most important of which is peer review. This process, where experts in a field assess the outputs of others in order to ensure and improve the quality of the work, is not without its problems, but it is still commonly seen as foundational to academic publishing. Peer-reviewed journals, sold to readers (or their libraries) through subscriptions, are the mainstay of traditional formal scholarly communication in most disciplines. Open access potentially disrupts key aspects of that status quo but does so in a way which its advocates argue improves the process of communicating research results, and in so doing improves the way research as a whole is conducted.
Open access (under various labels) can be traced back as least as far as the 1980s (Moore, 2019), but it started to gather real momentum at the turn of the 21st century. At its heart was the argument that academic content should be freely and openly available for all users: in a form which is âdigital, online, free of charge, and free of most copyright and licensing restrictionsâ (Suber, 2012, p. 4). Statements, such as the Budapest Open Access Initiative (BOAI), released in early 2002, helped to raise awareness of the possibilities of OA to a wider audience. BOAI is often seen as marking the beginning of OA as a credible mainstream approach to scholarly communication. The BOAI statement (BOAI, 2002) encapsulated much of the then current thinking around OA, and has often been a touchstone in subsequent discussion on openness. The fact that the statement was composed with a certain rhetorical Ă©lan meant that it had resonance at the time of its publication and has captured imaginations since. It is still commonly deployed in advocacy and debate.
The BOAI did at least three things. First, it played an early role in establishing the terminology of âopen accessâ. Second, it defined the main implementation routes of OA. Third, it summarised some the main arguments in favour of wider adoption of OA.
The first of these (defining OA), like Suberâs definition quoted above, built both âfree to readâ and âfree to reuseâ into the concept of OA:
permitting any users to read, download, copy, distribute, print, search, or link to the full texts of these articles, crawl them for indexing, pass them as data to software, or use them for any other lawful purpose, without financial, legal, or technical barriers.
(BOAI, 2002)
This bipartite understanding of what constitutes OA is still crucial and has become more important in the last decade as the possibilities of machine-processing of content have increased, widening as it does the potential of reuse. It constitutes a radically different approach to ownership and control, allowing content to be widely distributed and also repurposed. It has led to a favouring of open licences, such as those offered by the Creative Commons, as the basis of publishing and sharing content, since these licences have access and reuse built into them.
On the second point (the routes to OA), the BOAI outlined what are still seen as the two main âroadsâ to OA: OA publishing in academic journals (now often called âGoldâ OA), and depositing copies of outputs in OA repositories (âGreenâ OA). The relationship between Green and Gold is still at the heart of debates around OA today. The BOAI called them âcomplementary strategiesâ, but it was true even then that there were tensions between solutions proposed by Green advocates and Gold advocates. Those tensions have not gone away â quite the opposite, in fact. They continue to underlie much of the OA discourse, and have arguably weakened the leverage of the OA movement, and blurred the focus of OA policy, over a lengthy period.
On the third issue (the benefits of OA), the BOAI sets out the potential of openness. OA, it states, will,
accelerate research, enrich education, share the learning of the rich with the poor and the poor with the rich, make this literature as useful as it can be, and lay the foundation for uniting humanity in a common intellectual conversation and quest for knowledge.
(BOAI, 2002)
These are lofty claims, which identify benefits both within and beyond the academy. Most of the arguments in favour of OA have concentrated on the former, with OA likely to widen access to research outputs, for the broadest possible range of academic institutions and their members globally. The levelling of the playing field (as it is seen), between richer and poorer institutions, within and between countries (including those in the Global South), is often emphasised in the OA discourse (Barbour, Jones, Jones, Norton, & Veitch, 2011). The impact potential of publications is thus maximised, meaning the research communication can become most effective and research itself is improved by, for example, avoiding unnecessary duplication and even speeding up research. Most studies show that open research is used and cited more than non-OA research (SPARC, 2016). Moreover, for the academy as a whole, OA it is argued will fix systemic problems in the journal publishing market, where large global suppliers exercise oligopolistic power in the market, charging high prices and maintaining unusually large profit margins, syphoning money out of the research system.
Benefits are also often claimed to extend beyond the academy. OA has the potential to benefit a whole range of groups: clinicians, commercial research scientists, lawyers, teachers, journalists, policymakers, citizen scientists, amongst others (ElSabry, 2017). The argument is frequently made that academic research is a common good which is publicly funded and therefore ought to be in the public domain. This argument to been developed as a âmoralâ case for OA (Bacevic & Muellerleile, 2018; Peters & Roberts, 2012; Willinsky, 2006). Scherlen and Robinson (2008) make the case for OA based on social justice philosophy. Others have argued the case for the economic benefits of OA (Houghton & Sheehan, 2009), although less work has been done actually demonstrating benefits to those beyond the academy than might be expected.
The growth of OA
In the period since the BOAI was launched, and as the arguments in favour of OA have gained wider acceptance, the growth of OA has become apparent. The first major area of growth has been in the number of OA options available to authors. With regard to Gold OA, the number of journals offering authors OA publication has grown markedly. New OA publishers, such as Public Library of Science (PLOS) and Frontiers, have entered the market offering new fully OA journals. Existing publishers have also launched numerous new OA titles. Many OA titles have been based on a new business model, charging fees for publication (so-called, article-processing charges or APCs), rather than subscriptions. A large number of publishers have also introduced a hybrid approach, where particular articles in established subscription journals can be made OA on payment of an APC for that paper. Still other Gold OA journals do not charge an APC at all, but rather obtain funding from other sources, such as sponsorship from institutions and research funders. All of these options, and others, have been developed as publishers have begun to embrace OA (albeit, with varying degrees of enthusiasm). In the period between 2000 and 2009, the number of OA journals listed in the international Directory of Open Access Journals (DOAJ) rose from 741 to 4,767 (Laakso et al., 2011). By 2019, the number of journals on DOAJ was 13,168. For just those journals indexed in Scopus database, which focuses on established predominately English-language titles, the proportions of journals offering immediate Gold OA rose over the short period between 2012 and 2016 from 49% to 60% (Jubb et al., 2017).
At the same time, the availability of Green OA options has also grown. The number of OA repositories has increased substantially. OpenDOAR (Directory of Open Access Repositories) listed 691 OA repositories worldwide at the beginning of 2007 and this had risen to 3,820 at the beginning of 2019. The majority of repositories are run by institutions to house the outputs of their own members. There are also a substantial number of other repositories of different types, notably those run by specific disciplines. Institutional repositories (IRs) vary in size but tend to be relatively small compared with subject repositories. One of the largest subject repositories is arXiv, established as early as 1991, covering a number of sub-disciplines in several branches of physics, computer science, and mathematics, which now contains over 1.5 million papers. arXiv was originally set up to share âpreprintsâ (versions of articles prior to peer review), but has now become a widely used venue for sharing both preprints and âpostprintsâ (the versions of articles after acceptance by a journal). Other disciplines have set up their own repositories over a long period but, interestingly, there has been an acceleration in the numbers of preprints servers set up since 2013. This apparent second wave of preprints servers is a potentially important trend (Chiarelli, Johnson, Pinfield, & Richens, 2019).
However, populating repositories, particularly IRs, has not always been easy, with the notable exception of servers like arXiv. Depositing papers has become even more challenging as publishers have introduced policies making the conditions of deposit more restrictive. Since 2010 in particular, in an effort to protect the conventional subscription business model, many publishers have imposed increasingly restrictive conditions in their contracts with authors on how, when, and where published articles in their journals may be shared. Such embargoes often mean that articles cannot be made available on repositories until 12 or even 24 months following publication, thus blunting the usefulness of repositories and dampening enthusiasm to contribute to them (Gadd & Troll Covey, 2016).
Despite this, it is evident that take up of OA options by authors has risen in the last 20 years. This is the second aspect of the growth of OA (complementing the first, in the growth of OA options and venues) which has been clearly shown by a number of studies. Different studies use different data sources and methods, but all consistently show a rise in outputs being made OA. In their large scale analysis of OA in all of its forms, Piwowar et al. (2018) found that the proportion of the scholarly literature that is OA has been âgrowing steadily over the last 20 yearsâ. In the most recent year of their study, 2015, the proportion was as high as 45%. In another study focusing on papers indexed in Scopus, Jubb et al. (2017) showed that in 2014, 25% of articles were available in an OA form within 12 months of publication; by 2016, this had risen to 32% (including Gold and Green OA). Looking at Gold OA articles in particular, Wang, Cui, Xu, and Hu (2018) tracked the growth of OA articles from 7.5% in 1990 to 25.4% in 2015.
Adding to the growth of OA options and the growth of take up, the third aspect of the growth of OA is that of usage. Although evidence is a little more difficult to come by on usage, what is available shows growing use of OA materials, relative to non-OA materials. Jubb et al. (2017), whose report sets out evidence for these three aspects of the growth of OA particularly for the UK (OA options available to authors, the take-up of those options, and use of OA resources), identify several sources of evidence indicating a clear growth in usage of OA resources. Downloads of OA articles from publishersâ websites, for example, are on average between twice and four times higher than non-OA material.
The growth of OA is not evenly spread, however. There are, first, significant differences between disciplines; second, there are differences between countries. Differences across disciplines are marked. A large-scale analysis of articles accessible via Google Scholar in 2014 found 60% in the area of medical and life sciences were available in an OA form, whereas for law, arts, and the humanities it was as low as 32% (MartĂn-MartĂn, Costas, van Leeuwen, & Delgado LĂłpez-CĂłzar, 2018). At a more detailed, disciplinary level, the differences were even wider, with astronomy and astrophysics at 88% OA, and literature only 14%. OA for astronomy and astrophysics is mostly Green OA, delivered through arXiv. Other disciplines, such as the medical and life sciences, have been less ready to share versions of papers on repositories, but have historically favoured OA publication in journals, that is, Gold OA (Science-Metrix, 2018). The cultural differences of different disciplines have been used to explain differences in the adoption of OA (Fry, Spez...
