BIROn - Birkbeck Institutional Research Online Open-Access Publishing and Scholarly Communications in Non-Scientific Disciplines

This article presents an overview of the current state of debates surrounding open access in non-STEM disciplines. Design/methodology/approach. This article uses a selective literature review and discussion methodology to give a representative summary of the state of the art. Findings. Non-STEM disciplines persistently lag behind scientific disciplines in their approach to open access, if the teleology towards open dissemination is accepted. This can be attributed to a variety of economic and cultural factors that centre on the problem of resource allocation with respect to quality. Originality/value. This paper will be of value to policymakers, funders, academics and publishers. The original aspect of the paper pertains to the identification of an anxiety of irrelevance in the humanities disciplines and a focus on “quality” in open-access publishing debates.

In the preface to my recent book on open access (OA) and the humanities, Peter Suber succinctly set out, from an advocate's perspective, what has come to be a dominating aspect of the debate surrounding OA in non-scientific disciplines: Open access benefits the sciences and humanities about equally, but has been growing faster in the sciences. […] Certain myths and misunderstandings about OA are more tenacious and widespread in the humanities than in the sciences.
[…] I'd like to think that these myths and misunderstandings are more common in the humanities merely because humanists have had less time than scientists to catch up with the relatively recent advent of OA. But that's not true. They've had exactly as much time. Nor is the explanation that humanists are more careless readers of contracts, policies, statutes, or studies of OA itself. I suspect the true explanation is that humanists have had fewer working examples of OA to prove the concept and prove that the sky does not fall. They've had fewer working examples to dispel misunderstandings, generate enthusiasm, and inspire commitment. If so, then the humanities labor under a vicious circle in which the slower growth of OA causes a slower growth of good understanding, and vice versa (Suber 2014). i Suber, an individual who could claim to be the de facto leader of the Open Access Movement, accurately assesses the situation here. ii Assuming that we follow Suber and believe that all disciplines benefit from and are moving towards open access, the humanities and some of the social sciences lag far behind their scientific counterparts. The problem is, though, that although there are good reasons why such a transformation might be desirable, such a teleology is not universally accepted. In fact, there are a range of resistances and counter-arguments from beyond the sciences, predicated sometimes on misunderstandings but at other times on different disciplinary situations, that have resulted in a range of responses to new forms of dissemination.
Open access, to summarise briefly, refers to the removal of price and permission barriers to 2/17 scholarly research. OA relies upon the economic reconfigurations of dissemination on the internet and world wide web to allow readers to access work without paying (the digital environment eradicates the cost-per-copy and instead concentrates all costs in the cost to first copy). Likewise, OA calls for the modification of standard copyright protections to allow readers to re-use work beyond the statutory provisions of fair dealing. Some methods of achieving open access are symbiotic with the existing subscription ecosystem. So-called "green" open access, for instance, refers to situations in which researchers deposit a copy of their work in an institutional repository, even when it was published in a subscription venue. There is a substantial volume of evidence, covered below, pertaining to unembargoed deposit in the discipline of high-energy physics, coupled with surveys of librarians, that reveals that green open access does not, at present, cause subscription cancellations. If, however, we want to achieve full gold open access, where publishers make the material freely available themselves, at source, then the economics of research publication must be reconfigured. Indeed, if material cannot be sold because it is given away, then publishing becomes a service and must be remunerated from the supply side.
In this article, I will cover the specific challenges and objections that have been mounted from the humanities and social scientific (HSS) disciplines where open access seems slower to gain traction. This discussion will be structured into two separate areas: economics and culture. Inevitably there will be some overlap between these areas. For instance, the monograph is a key component of communication (and assessment) in the HSS disciplines -which is a cultural aspectbut the economics of this particular form are far more complex than journal articles. Open access in the humanities and social sciences is not impossible. It is, in many ways, desirable. However, it is naïve to assume that it will follow the same trajectory as in the sciences. It is not, after all, for no reason that our subject areas have been differentiated in the past through the term "the Two Cultures", regardless of how unhelpful such a distinction may be (Snow 1993).

Open Access, HSS and Economics
The underlying premise of labour and remuneration that makes open access possible for academic research is that researchers are paid a salary by an institution, rather than having to sell their work. Although this is not universally the case (and the increasing degree to which the academy depends upon precarious labour is deeply problematic), it is an ideal system where it does exist because it means that areas of enquiry can be selected for their own esoteric value rather than for reasons of market populism. This is integral to academic freedom as we know it: researchers are free to investigate areas that are of interest, rather than being restricted to topics that will sell. The same is not true, though, for publishers. These entities, in the traditional subscription or sales environment, are beholden to a specific and strangely hermetic market, an aspect that comes with many material knock-on effects for the university and its libraries.

The Serials Crisis and non-STEM Disciplines
In the journal sphere, when all disciplines are included, the cost for every institution of subscribing to all published serials has risen by approximately 300% above inflation since 1986. By contrast, library budgets have only risen by 79% (Association of Research Libraries 2014). In absolute terms, 3/17 the humanities and social sciences do not contribute to this so-called "serials crisis" at the same levels as the sciences (although they suffer from the effects of it as library budgets are weighted towards STEM subjects under most current political imperatives). Most HSS journals are far cheaper than their scientific counterparts even though they may have higher rejection rates. In relative terms, though, there is a similar hyperinflationary rise (Bosch & Henderson 2013).
This cost situation, across all disciplines, is not purely constituted by publisher price hikes. Instead, it is directly linked to the rise of research assessment and the economic situation of the academic job market. As competition for academic jobs has intensified, the need for academics to publish work potentially outstrips the desire to read the outputs of others. It is not necessarily wise to use the term "over supply" when clearly a lot of this published work is interesting, correct and worthy, but the proliferation of outputs comes with a rise in material costs.
Concomitantly, as the number of candidates applying to jobs/for tenure increases, selection panels face a labour shortage in the evaluation process. Such panels cannot afford to re-read every piece of work submitted to them. In some cases, the process would entail reading hundreds of books every time a post was advertised. To this end, hiring panels employ "proxy measures" to stand in for quality. This is most frequently the journal or publisher name. In some disciplines, the nebulous "impact factor" is used as a quantitative measure instead, though (for a critique of the impact factor, see Brembs et al. 2013 and also the San Francisco Declaration on Research Assessment [DORA]). This has led the executive editor for the humanities at Harvard University Press to join the opinion of a former editor of the British Medical Journal in stating that universities have "effectively outsourced to journals and publishers the function of assessing academic quality" (Smith 2013; see also Waters 2001).
In turn, this has economic knock-on effects. If journal name, publisher name or impact factor are used as measures for hiring and tenure processes, then researchers will seek to publish in venues that fair well by these proxy measures. If there is a strong desire to publish in these venues, it is likely that these journals/publishers will receive the lion's share of high-quality material. If these venues contain high-quality material that is important for researchers, libraries must subscribe to them. If these venues are owned by corporations whose imperative is to ensure shareholder return (as is now the case for many commercial academic publishers, as opposed to mission-driven university presses), they are likely to raise prices for their captive library clientèle in order to do so. In this way, a form of symbolic capital (prestige) is converted to material capital for various entities (academics who gain jobs and publishers who make profits) but only at the expense of the library budget.
This discrepancy between the library budget and the cost of published research, caused by the factors outlined above, leads to an access gap. This is, obviously, experienced more sharply at institutions with smaller budgets and may not be apparent to those in elite and wealthy universities. That said, even Harvard has cancelled subscriptions based on excessive price (Suber 2015). It also means that this work is frequently unaffordable and therefore inaccessible to the broader public, many of whom may themselves hold degrees in humanities and social scientific subjects and may have a particular interest in the cultural and artistic exegesis provided by the humanities and social 4/17 science disciplines. Furthermore, the economic protections of copyright that are needed under the subscription/sales mode mean that others may be prohibited from re-using the work for teaching (reprographic reproduction), digital humanities work (text/data mining) and inclusion in wider fora (such as Wikipedia).
Some forms of open access, which would address this access gap, have no discernible effects on the current economics. This can be a positive or negative phenomenon depending upon one's perspective. Green open access, where copies are placed in an institutional repository, has been shown in several studies to have little impact upon subscription cancellations, making publisher's revenue secure (Suber 2012, pp.149-161). There are also many reasons above a green OA copy being available that induces librarians to cancel subscriptions, with price and pedagogical relevance often ranking more highly (Association of Learned and Professional Society Publishers 2006). On the other hand, though, this mode relies on the continuation of hyperinflationary price increases, which are patently unsustainable.
On the other hand, gold open access, in which publishing is remunerated from an alternative business model, suggests a reconfiguration of the current economics, with a potential for savings (although rhetoric on this varies; the UK government's university funding agency, HEFCE, for instance, claims that cost savings are not their goal in pushing for OA (Meadows & Sweeney 2014)). In this mode, the functions of publishing (which Michael Bhaskar defines as filtering, framing and amplification (Bhaskar 2013)) are not undertaken in the service of creating and selling a commodity object (a journal or book) but are instead to be viewed as a service to authors. In other words, in this view, the value-adding elements of publishing should not be paid by readers but by clients of publishers: academics, their funders and institutions. Some publishers dispute that this is a good idea. Steve Cohn, the Director of Duke University Press, with whom I sat on a panel at the National Humanities Center in March 2015, claimed at that point that the value of the Press was in forcing authors to make their arguments more intelligible and in selecting high-quality material; services for readers. Cohn stated that inverting the logic, so that publishing becomes a service to authors, would lead to a situation of declining quality (as well as expressing scepticism over the viability of OA business models). I do not agree that this must be the case; OA publishers can still use their reputation for publishing high-quality material to deliver an excellent experience to readers. Under an open access mode, they simply do not have to do so by excluding people from reading based on price.
Thinking further about this shift in business models, it becomes clear that there is no single way in which these economics can be reconfigured. Some journals operate on the basis of voluntary labour of editorial staff, meaning that the costs are essentially cross-subsidised by institutional or personal time. Others, such as the Open Library of Humanities platform that I am building, the arXiv project and the Knowledge Unlatched model, solicit funding from an international library consortium so that there is no need to sell material. The most well-known (although not the most common (Directory of Open Access Journals n.d.)) way of remunerating the labour of publishing for gold OA, though, is through a mode called Article Processing Charges (APCs). This mode is one wherein the authors, their institution or their funders must pay a fee to the publisher so that the necessary work can be covered. When properly implemented, this is not a payment to bypass peer 5/17 review and it is in no way incompatible with rigorous quality control. It does, though, cause disquiet for several reasons.
The first is that this mode significantly alters the point at which the cost of publication is borne. In the subscription model, the fact that many libraries all pay a subscription transforms the scenario into a risk/cost pool. By this I mean that a large number of libraries all pay a (relatively) small amount per publication so that, centrally, there is enough money to undertake the labour and/or build a surplus/profit. This mode, while spreading risk/cost, which is sensible, creates the access gap, which is not. By contrast, a model involving APCs does the inverse. Instead of spreading risk/cost, it concentrates risk/cost at a single point of payment, but eradicates the access gap to readers. In many disciplines, such as the humanities and social sciences, where little external funding is available for article processing charges, this presents a problem. This means that research-intensive institutions may end up paying hundreds of times more than they currently do while others pay far less. This is not necessarily iniquitous, but it may make it harder for younger institutions to break into a research-intensive mode if the systems of distribution are fixed on the assumption that they don't need publication funds. Furthermore, as I will return to below, because this mode is perceived to potentially interfere with the ability of academics to publish in the venues that will do the most for their assessment, there has been staunch resistance.
The second challenge for an APC mode that has been prominently voiced in the humanities disciplines is ensuring that quality control is rigorously divorced from economics, an aspect to which I will return at the conclusion of this article (in fact, it is monographs as they currently stand that are assessed on the basis of whether they will sell for a publisher; market populism is, to some degree, already here). In a model where publishers are paid for accepting articles, the obvious question is whether this will encourage them to accept material simply to take payment, regardless of its quality. Of course, in the subscription environment, if a paying subscriber has been promised twenty articles per year, they must be given twenty articles or they will have been short-changed. What happens if the publisher doesn't receive twenty top-quality articles (Suber 2006)? In other words, this isn't a new problem, it is just more acutely exposed in an APC environment. The second thing to note is that this is only possible because of blind peer review. If it were known who was validating the work (the reviewers who supply a genealogy of validation) it would be impossible for unscrupulous "publishers" to claim they had undertaken review when none was actually present. iii Of course, this comes with a wide variety of other problems and challenges, particularly with respect to the conservatism or otherwise of reviewers (Eve 2013). In the meantime, we continue to rely on publishers' reputations to keep them in check, trusting that those who admit patent nonsense will quickly develop notoriety. In turn, though, we also need to continually re-enforce the continued importance of the development of information literacy and critical evaluation skills, probably best housed in the academic library.

Monograph Economics
Understanding these economics of serials -both symbolic and material -is crucial to grasping the publication environment. There is a substantial difference, though, between the predominance of serials in the scientific spheres and the importance placed upon the research monograph in the 6/17 humanities and some social science disciplines. This is not the same as a textbook in any discipline, which has a totally different economic profile. It is, rather, to refer to the production of a lengthy, specialised volume, sometimes with limited market appeal. In HSS disciplines, these undertakings are prized for the space necessary to undertake a longer, detailed study, usually of around 80,000 words. They also form the crux of assessment paradigms with the monograph consistently referred to as the "gold standard" for hiring, tenure and promotion, a fact that is somewhat ironic given that the peer review mechanisms are often very different and, in some ways, potentially less strict than for articles. The economics of these productions are somewhat different, however, and merit their own discussion for an open access environment.
Indeed, the first thing to note is a simple scaling of the artefact in question. If it is true that APCs at the emergent market rate are problematic for HSS disciplines, then the equivalent BPC (Book Processing Charge) is even more so. These charges range from $2,450/chapter from de Gruyter; €640/chapter from InTech; £5,900 from Manchester University Press for books of up to 80,000 words; £11,000 from Palgrave; and approximately €15,000 from Springer, as just a selection (Palgrave Macmillan 2013). In many humanities and social science disciplines, it will simply be impossible to obtain funds at this scale, although international variances in funding situations may render this a more acute problem in certain environments.
There are several reasons for these escalated costs for books. The first is the commensurately higher degree of editorial labour that is invested in such works. Copyediting and proofreading are very different matters at a scale of 80,000 words, although it is notable that most publishers outsource such activities now. Typesetting is also, in many instances, a more laborious affair at this length, particularly if manual reference tagging in commercial software such as eXtyles is required. Furthermore, the efforts of peer review coordination are more substantial than for journals (and reviewers are often paid in this sphere, adding additional financial overhead). There are, in addition, areas of marketing, aggregation, royalty payments and others that are simply absent in the journal sphere.
However, scale is not all for monographs. The channels of reception are often, also, very different. For instance, the fact that some books have a potential to find a popular audience -who will expect to find such items for sale in bookshops, rather than solely available for free download online -changes the scope for a transition to a pure open access environment. Furthermore, print looks here to stay, at least for the time being. Because academics use books in very specific ways (making full use of the scholarly apparatus, such as endnotes, an index and a bibliography), the reading experience is seldom the same as the usual linear approach to conventional fiction, for which e-readers have been designed. Indeed, many existing studies show the continued desire for print (OAPEN-UK 2012). This, though, as shall be seen, is actually a positive element for open access monographs.
In any case, with the scale of the challenge ahead for open access books, it is not surprising that several ongoing projects have sought to investigate the economic situation. The most prominent of these, to-date, are the OAPEN-NL study, the OAPEN-UK project, the HEFCE Open Access Monographs project and the just-launched Mellon-funded enterprises (and proposed collaborations 7/17 such as the "first-book subvention") in the USA.
The OAPEN-NL project -run by Eelco Ferwerda, Ronald Snijder and Janneke Ademaunder the broader remit of the OAPEN (Open Access Publishing in European Networks) initiative was established to investigate the economic implications for gold open-access books. As the project describes itself: OAPEN-NL was a project to gain experience with Open Access publication of monographs in the Netherlands. Between June 2011 and November 2012, 50 Open Access monographs in various subject areas were published in [gold] Open Access by 9 participating publishers. For every Open Access title, the publishers provided a similar title that was published in the conventional way.
[…] Data were collected about usage, sales and costs, to study the effect of Open Access on monographs. OAPEN-NL consisted of a quantitative and a qualitative research component, measuring the effects of Open Access publishing and the perceptions and expectations of publishers and authors (Ferwerda et al. 2013, p.3).
For the OAPEN-NL project, the concept of "similar titles" was based upon the number of pages, the price and the subject area. This project was finished in 2013 with participation from a range of publishers: Koninklijke van Gorcum, IOS Press B.V., Springer Science & Business Media, Techne Press, Wageningen Academic Publishers, Koninklijke Brill NV, KITLV Press and Amsterdam University Press. The project investigated 50 books with a total expenditure of €239,615.85 by the project at a maximum of €5000 per book paid to publishers (Ferwerda et al. 2013, p.67).
From its investigation, the OAPEN-NL project concluded that 'no significant effect of Open Access on monograph sales could be found' but that there was significant increase in digital usage (the number of times a book was viewed on Google Books) when it was made open access. In opposition to several studies in the journal sphere, the project found no increase in citations to open access books (Ferwerda et al. 2013, pp.55-57;Swan 2010). Finally, the project also examined the costs of publishing and concluded that an OA edition is approximately 50% cheaper to produce than the total cost of a conventional, print monograph, a finding that has been queried by many publishers (Ferwerda et al. 2013, pp.40-53).
There are various ways in which these results can be interpreted. For instance, the finding that open access does not negatively affect sales can be queried if one considers that discoverability for open access books (through MARC metadata provision and catalogue ingestion, for instance) is less developed than for its more traditional counterparts. The effect of this could be, for example, that if researchers were unaware that a free version was available, they may have bought the book for that reason, thus skewing the results of the study. Although OAPEN-NL noted that discoverability of monographs (measured by usage) was generally increased (the second finding) through one particular route (Google Books), it is not clear whether this is the discoverability route of those who would purchase, or simply use, the works. It is also true that, since this project began, open access has grown exponentially on the international stage, even in the monograph sphere, through funder mandates such as Wellcome. It is arguable that acquisition librarians are now more attuned to lookout for OA editions and the ways in which they might save their precious resources. Finally, the lack of a citation boost was unexpected; 94% of the book authors surveyed in the study 8/17 expected to see a rise in citations (Ferwerda et al. 2013, p.83). This result might be explained, through, when the long publishing cycle in the humanities against the relatively short run-time of the OAPEN-NL study is considered. The citation findings are due to be reviewed in the near future.
The work of the OAPEN-NL project is continued and supplemented, in some sense, by the OAPEN-UK experiment. By contrast to OAPEN-NL, though, OAPEN-UK is structured around the concept of 'matched pairs', in which profiled books are designed to be compared with one another. The OAPEN-UK project also contains more titles than OAPEN-NL, with 90 books, 45 of which make up the experimental gold open access group, while the other 45 function as a traditional, purchased control group (OAPEN-UK 2013). The publishers participating in OAPEN-UK are Routledge, University of Wales Press, Liverpool University Press, Palgrave Macmillan, Berg Publishers and Oxford University Press, who joined the project at a later point. The disciplinary range of the monographs included spans international criminal law, classics, literature and history, through to marketing, among many others. As the OAPEN-UK project is still running, no definitive results from the quantitative portion of the experiment/control study are yet available. That said, the project has already released some extremely valuable findings from various case studies and focus groups as part of a structured qualitative research programme. This includes problematic findings that many in HSS disciplines still link OA to lower quality outputs and interviews with learned societies that demonstrate economic entanglements with the subscription system that will prove difficult to reverse.
Another monograph exploration has also been undertaken by the UK's HEFCE (the Higher Education Funding Council for England), a quango (quasi-autonomous non-governmental organisation) that translates the government's higher education budget allocation into usable funds. In April 2014, HEFCE announced that eligibility to receive future funding through its "qualityrelated" (QR) stream would depend exclusively upon the assessment of green open access material: a mandate. Specifically, authors must deposit the accepted version of their articles at the time of acceptance (Higher Education Funding Council for England 2014, para.17-19). However, monographs 'and other-long form publications', edited collections, non-text outputs and data are all excluded from this mandate (Higher Education Funding Council for England 2014, para.14).
From the rhetoric deployed by HEFCE and the UK's Research Councils, some academics have surmised that these bodies would like to mandate monographs for a future exercise; after all, why should one form be deemed different to others in their eyes when both are supported by QR funding (Evans 2014)? However, in recognition of the additional barriers (and researcher sensitivities) surrounding open access monographs, HEFCE has instead opted for now to mount an investigation into the subject, the first national-level funding council investigation of its type. The investigation was led by Professor Geoffrey Crossick, an ex-Vice Chancellor of the University of London and a Distinguished Professor of History and supported by an expert reference group.
Crossick's report consists of a detailed analysis of the importance and role of the monograph in non-STEM disciplines, along with an assessment of the feasibility of open access for books and an independently commissioned study into business models. While more than strictly economic, for the purposes of this article, a select range of Crossick's fundamental conclusions may be 9/17 paraphrased thus:

 Monographs remain important for humanities scholars for communication and accreditation.
Their viability shouldn't be jeopardised by OA policies.
 Surprisingly, the monograph does not appear to be in crisis.
 Peer review is still key to academic publishing.
 Publisher brand is a dominating factor in the assessment of publication quality. This means that if OA is to develop for monographs, it will require the participation of established presses.
 Print books are still viewed favourably over their digital counterparts for academic books and any OA for monographs should be symbiotic with print editions.
 Policymakers must clearly articulate the opportunities and benefits of OA for monographs if they are to win the hearts and minds of researchers. Academics must be brought on board for OA monographs to succeed.
 A small, select group of academics will resist OA for monographs on the grounds that they earn substantial royalties through the sale of their books. Crossick recommends that policymakers have an exemption policy for this eventuality.
 A requirement for the most liberal of open licenses may hinder uptake as these are contentious.
 The inclusion of third-party material in disciplines such as art history may be problematic for OA books.
 First books are usually so different from the Ph.D. theses from which they derive that mandates for OA theses are immaterial for policymakers considering OA books.
 There is a global movement towards OA, including for books.
 There is no single, established economic model for gold or green OA books.
 Business models for OA books are at an early, experimental phase. It is unlikely that one model could be imposed through policy decisions.
Many of Crossick's findings are uncontentious. That books are important, that print remains valued for its combination of sequential and random access, and that the current set of business models are experimental hardly cause a stir. By contrast, some of these points could be queried. That the monograph appears to be financially healthy, for instance, runs contra to the prevailing discourse that claims that books are in crisis. Indeed, several papers given at a conference back in 1997 questioned whether the rhetoric of crisis was better framed as "chronic illness" given the perpetual nature of this claim (Case 1999). Likewise, the acceptance of publisher brand as a mark of quality goes, here, as an uncontested proxy measure, despite the problematic economic situations that this engenders (see above).

10/17
Finally, the statements on book royalties present difficulties. Very few academic authors sell enough copies to thrive off their book royalties simply by dint of the fact that the average print run of a monograph is estimated to be 200-250. Presumably, authors will not definitively know, in advance, whether their book is going to sell well. If an exemption is allowed for royalties, then authors may pre-emptively (and over-optimistically) veto open access on the grounds that they might make money off their books in future. To reiterate, open access is only really designed for authors who don't need to sell their works for a living (because, in the case of academics, they already have a salary or too small a market to usually make this feasible). If books span that divide and it is culturally accepted that authors will earn revenue from them, then OA is harder to achieve in this realm. Conversely, though, if books are the outcome of funded research work, funders may consider such royalty payments too much of a private, as opposed to public, benefit from their financial contribution.
Along with various emergent, early-stage Mellon-funded projects in the US, these represent the main investigations into the economics of open-access monographs. Clearly, the economics of publishing in HSS disciplines is linked to a holistic environment; it does not make sense to speak of the higher-education research publication landscape as discrete to each discipline. The economics of one sphere, such as biomedical journal publication, are felt in HSS monograph purchasing. That said, there is a greater distance left to travel in the HSS disciplines if the economics of open access publishing are to stack up for the diverse types of output valued in these fields.

Cultural Differences and Resistances to OA
For a variety of reasons, open access is not universally accepted as the future path for scholarly communications. This does not just extend to non-STEM disciplines; chemistry and medicine have poor open-access uptake, comparable to some humanities disciplines (Gargouri et al. 2012). That said, non-STEM disciplines persistently lag behind their scientific counterparts (assuming that Crossick's identification of a trajectory towards OA holds). In this section of this article, I will outline some of the cultural differences that exacerbate resistance to open access in these fields.
The first and most pressing social difference stems from the economics and is best phrased as an "anxiety of irrelevance". The humanities disciplines and some social sciences have become accustomed to thinking in terms of crisis. Frequently, such a rhetoric of crisis and anxiety comes from the fact that HSS practices are evaluated by comparison to the natural sciences. In the prevailing discourses, it is hard to justify funding research into the production and reception of literature, say, when the money could be spent on the noble, but also far-more politically expedient, goal of "curing cancer" (or similar).
Wherever one stands on this matter, however, the strategies that non-STEM disciplines tend to adopt are clear. The first is to devalue the measuring strategy. This consists of decrying a scientism of evaluation by claiming that it is inappropriate to measure these fields by the same yardstick as the sciences. For some, like Sarah Churchwell, this takes the form of defining the humanities as the safeguard of "any hope of beatitude in a secular age", a point that seems to protest a little too much given not only the astonishing beauty and elegance of some science but also that 11/17 the equation of beauty and truth has been contested for many years (Churchwell 2014). Such an argument does, nonetheless, continue to make the claim that a single evaluative framework for the sciences and the humanities is impossible. This first stratagem serves to undermine the premise that "the humanities can be measured in the same way as the sciences". Combined with a second strand of argument that decries all measurement as systematically against the purpose of the humanities, these approaches are designed to undermine the inductive reasoning of quantified measurement. Of course, the social sciences are on weaker ground here; they cannot wholly differentiate themselves from the natural sciences.
The next tactic, however, attempts to deal with this weakness. While the former arguments are designed to differentiate the disciplinary purposes from the sciences, another is to use this foundation to suggest a different set of practices, which is where open access comes in. If, it is argued, the purpose of these disciplines is different, then why should they be subjected to the same expectations in terms of communications as their scientific counterparts? This can even be extended to the underlying economic principles. Robin Osborne, for instance, has taken this line previously. Osborne wrote, in a piece for the British Academy, that "Academic research is not something to which free access is possible. Academic research is a process -a process which universities teach (at a fee). It is neither a database, nor the ways and techniques by which the database is manipulated" and if it is "only trivially a result of the research-funding" it should be exempt from OA mandates (Osborne 2013, p.104). This contrast with a "database" (which, in fact, could actually be the output of "academic research" in the sciences or even in the digital humanities) implies that there is something exceptional about the humanities that should exempt these disciplines from OA mandates.
This difficulty is amplified, then, when OA mandates are tied to research assessment exercises. The UK's Research Excellence Framework (REF), for instance, is widely disliked by academics, particularly in the humanities, who accuse it of being a wasteful and expensive exercise in inappropriate measurement. However, its centralised nature also makes it a powerful tool through which green open access (and other behaviours that funders deem desirable) can be encouraged; a (soft-)power instrument. At the same time, though, the negative association of open access with such paradigms may be problematic, particularly if, as I have argued above, the humanities often deploy rhetorical strategies that attempt to place themselves outside of such bounds.
Other disciplinary specificities and claims for exceptionalism are also often invoked. A common line of argument taken in informal discussion is that it is the sciences that have created the budgetary crisis and that the humanities and/or social sciences can (and should) play no role in reworking the economics of scholarly communication. There is a grain of truth to this. It is notable that many of the large, commercial, scientific publishers (usually framed as the enemies of the Open Access movement) have simply managed to make OA an additional revenue stream on top of their already-hefty subscription income, a strategy known as "double dipping" and decried by prominent library figures (Prosser 2015). In this light, a critique from the humanities and social sciences of an author-pays gold model sometimes centres on this disciplinary difference in expenditure: we didn't create the problem, our budgets are smaller, it's not incumbent upon us to fix it.

12/17
There are also, though, some social and technological challenges that pose challenges for open access and that meet with resistance that I will cover in the remainder of this section. These can be structured into technologies of reading; author rights and licensing; and the role of peer review.
With respect to technologies of reading, it is important to note, as per Crossick's findings above, that no better technology exists for long-form reading than the codex. The material book object is adept at sequential and random access. It has "annotation functionality" in the form of margins, which even become "social annotation" features when one lends a copy that contains annotations. The Kindle and other e-reading devices do not currently successfully act as substitutes for the codex, particularly in a teaching environment. In this respect, the most frequent (incorrect) assumption among humanists is that open access calls for the immediate and systematic eradication of the material, print, book object. Certainly, most believe, as Jerome McGann has recently made clear, "that scholarly communications will soon be largely organized in digital venues" (McGann 2014, pp.1, 4, 14, 20, 132). For now, though, as McGann also notes, we exist in a "half-world", a space where print exists alongside the digital forms and, in fact, it may be that open access helps publishers to generate revenue from print sales, so long as no comparable digital technology to the codex emerges.
The penultimate area of social contention that I will broach here pertains to author rights and licensing. Open licensing, the second component of all of the original declarations on open access, is not well accepted in non-STEM disciplines. Some fears here are unfounded: that open licensing will lead to plagiarism, for instance, despite every suggested Creative Commons license containing an attribution clause. Plagiarism will happen in any case because the people who plagiarise already do not care for institutional rules. Indeed, the more interesting critique here pertains to moral rights of authors.
Copyright is generally considered to consist of two components: economic rights and moral rights. The former is designed to give a time-limited monopoly to content producers so that they can materially benefit from the mental labour "invested" in the creation of "intellectual property". As has already been mentioned, under the theoretical model that underpins academic remuneration (which doesn't take into account precarity or royalties/advances), academics do not use the economic rights of copyright; publishers do.
The other side of copyright consists of the "moral rights" of the author. These are often broken down into three separate components: the right to attribution, the right to a pseudonym if desired and the right to the integrity of the work (to object to derogatory treatment of the work). Some critics of open licensing for open access work feel that open licensing, and particularly the more liberal Creative Commons attribution licenses, do grave damage to the moral rights of the academic author. While noting, again, that the licenses in question do require attribution, it is also worth delving into a more theoretical stance on this by asking: what is the basis for these moral rights? What, exactly, is moral about them? And is the law the best way to enforce this in order to advance the espoused goals of university research?
Firstly, like the economics of the assessment of scholarly communication, moral rights are 13/17 based on the idea of a reputational, or symbolic, currency that rewards labour and that treats ideas as though they were physical property. Those familiar with the work of Pierre Bourdieu will appreciate that that this seems to resonate exactly as per the symbolic economies that he famously described (Bourdieu 1977, p.180). Moral rights have an economic function in that they are designed to allow an author to accumulate a form of capital -be it social, symbolic or cultural -but that the forms are all also interchangeable with one another and, also, with material capital (money) in some ways. For instance, in the case of academics, accumulating a name in a field, via citations, can lead to a post or promotion, a real material return from an otherwise symbolic, reputational form. Likewise, this thinking presumes that the first person to publish an idea has a claim to own it, as though an idea were a geographical space, ready to be colonised and occupied by the first settler, despite the fact that many similar ideas can occur discretely to different people. As with all historical forms of colonialism, the settlers were never purely discovering. In this sense, the objection to derogatory treatment of one's work is brand preservation. In this first way of thinking about moral rights, they are, actually, simply another form of economic rights.
Secondly, and more provocatively, it can be argued that moral rights are based on an egocentric individualism. It is true that academics (myself included) like to be cited. This reward structure is an incentive to write but it is nonetheless ego driven.
So, in what way are these rights "moral"? A brief detour into the purpose of research might help to clarify this. Research work, regardless of discipline, is a process of communicating new truths or interpretations, or of contesting existing truths or interpretations. Certainly there are core differences in how this is achieved between fields (not to mention how such findings are communicated). The central purpose remains the same, though. And it is a noble, ethical purpose in its pure form within the university: contribute to epistemology and hermeneutics for the broader good.
The same cannot be said for moral rights of academic authors. Beyond the incentivizing aspect, it is not the case that symbolic economics or egocentric recognition are fundamental to the moral purpose of research work. How does it help to build truth or interpretation to insist upon credit? Certainly, authors still desire it, but it is needed for the benefits of recognition to the author, which is not really a "moral" stance.
All of this is a long way of addressing the fact that demanding attribution and integrity of work is not necessarily a "moral" act. For others to credit work is a moral act on their part. For an author to demand it as a "right" is an economic and individualist act that may work as an incentive. We might also consider, though, whether copyright is a good mechanism for enforcing attribution and integrity. I'd say not. It is time limited and will expire at some point. At that moment, none of its protections will apply any more. Concurrently, is there a time limit on the truth within the research? Of course not. Copyright doesn't chart the needs and desires of the academy; the map is not the territory. Instead, the academy has developed sophisticated rules within its own walls that mean that people must properly cite others, regardless of whether there are legal protections on the work. For these reasons, among others (such as John Holmwood's claim that private educational providers will profit from openly licensed work (Holmwood 2013a;Holmwood 2013b)), open 14/17 licensing of HSS research material remains a contentious topic.
However, turning now to the final point I will raise in this article: in the last instance, all the "debate" and argument over open access (whether in the sciences or HSS) can, I contend, be boiled down to one, fundamental aspect: quality. How, in an economy of scarcity, can we allocate resources so that the highest quality work is published, disseminated and read? How can we judge what quality means in advance and what labour is involved in that process? What does it mean for "quality" to reconfigure our scholarly communications infrastructure and its economics, in the digital age?
Peer review is usually held up as the gold standard for accrediting and pre-filtering research work for quality. A practice that originated in the sciences as a means of engendering intersubjective verification, other disciplines were quick to adopt some of its principles. The process is far from perfect, however. If deemed "blind", it is, in actuality, rarely fully anonymous (Eve 2013). It is often subjective in the sciences (if judging importance) but is certainly even more so in the humanities. What does "quality" even actually mean? It seems likely that quality is, in one sense, like pornography. In the words of Justice Potter Stewart: we may not be able to define it, only to know it when we see it.
Nonetheless, universities have limited budgets to pay for published research work and researchers have limited time to read. A regulating economy of scarcity that denotes quality is, therefore, desired by many. In bygone eras, this was provided by print. With limited page budgets, it was easy to regulate quality; it could truthfully be claimed that there was only space for the very best work. In the first wave of digital research publications, print scarcity was simply replicated in the digital space. Journals had issues, volumes, pages and print correlatives. Open access can appear as a challenge to this. Indeed, a tectonic shift in the underlying economic plate -triggered by the fact that digital reproduction is instantaneous and non-rivalrous -is bound to cause a tremor as the fault-lines of this artificial economy are exposed. Despite all serious OA advocates insisting on the continuation of peer review, it seems likely that this is the core anxiety at root, deeply felt in the non-STEM disciplines where a nervousness about value and quality is never far away. In the age of unlimited, free, digital reproduction, what challenges are posed to the aura of quality and what difficult questions are asked of our assumptions about our processes to assure this (see Fitzpatrick 2011)?

Conclusion
In this article, I have given a potted summary of some of the additional challenges and disciplinary complications for an implementation of open access in non-scientific disciplines. It is impossible, in a piece of this length, to do full justice to the nuance of argument and the many hundreds of thousands of words that others have previously spilled on the topic. It is also no secret that I am an advocate for open access, although I would like to think that this support comes from a rational consideration and critique of the scholarly communications infrastructure and ecosystem, rather than from dogma. What is certain is that some forms of open access are coming to non-STEM disciplines; it is not possible to roll back the economic changes that are a result of digital 15/17 technologies and these have important and groundbreaking implications for the production of research work. Business as usual seems impossible. At a difficult time of transition, however, OA causes unrest and disquiet. The temptation is to stand at the digital shore and command the tide to recede. What must happen instead is at once to critically appraise what we need from a scholarly communications infrastructure and to simultaneously build pragmatic and non-damaging transition strategies to harness the full power of open, digital dissemination. ii I here use "open access" in lowercase to refer to the phenomenon of removing permission and price barriers to research while the title case "Open Access movement" is used to refer to the array of individuals dedicated to making this a reality. iii I owe the phrase "genealogy of validation" to Martin McQuillan.