The LAGB intends to submit a response to the current consultation co-ordinated by HEFCE on behalf of the UK higher education funding bodies on Open Access and submissions to the REF post-2014. The consultation is based on a statement of the developing intentions of the four UK higher education funding bodies available at http://www.hefce.ac.uk/whatwedo/rsrch/rinfrastruct/openaccess/.

In essence, the funding bodies are proposing to limit submissions to future REFs to items that have been made available via gold or green open access as soon as possible after publication. In initial discussions, the following points have been raised: 

• given the funding made available for article processing charges, realtively few outputs in linguistics will be via gold open access
• researchers without an academic position (including PhD students seeking to build a career and retired staff) will have no access to gold open access
• choice of journal may become limited by open access options making it more important that place of publication is not interpreted as evidence of quality
• some linguistics journals will never be compliant with open access (particularly small specialist ones and those produced by small publishers in countries with no tradition of open access)
• there is concern about adoption of the Creative Commons CC–BY licence, which permits reproduction of publications and derivative works with only general acknowledgement of the source
• on the other hand, too restrictive a licence may prevent development of linguistic infrastructure (corpora etc.)
• there is currently no business model for open access monographs, and introduction of compulsory green open access for monographs depends on one emerging almost immediately
• given long lead-in times for publications in linguistics (where 2 years from submission to publication of a journal article is normal), any requirements should not apply retrospectively to items submitted before new rules come into place
• the longer-term goal of open data raises a number of issues in linguistics involving third-party copyright on textual material and data gathered from interviews with human informants

Discussion on the LAGB forum (http://www.lagb.org.uk/forum) is welcomed on these or additional points relating to the consultation. The deadline is tight, so please make comments on the forum in the next week or so, by 18 March, so that the committee can collate a response.

It's natural for there to be teething problems with a proposal like this. The big question, still largely unanswered, remains: who is going to pay the APCs?

However, I don't think issues of implementation should prevent the UK from going forward with a scheme that removes the tremendous inequalities that are present in the current system. Taxpayer-funded research should be accessible to the taxpayer, and access to knowledge should not be restricted to those who are either a) affiliated to a university or b) rich. That is the common-sense motive behind open access, and it's worth bearing in mind as we consider the practicalities.

I'd like to add a few factual points to clear up potential misunderstandings:

• 'Gold' OA does not necessarily mean that APCs will be charged. That use of the term is specific to the Finch Report, which wrongly assumes (for the most part) that APCs are the only way to fund OA. Gold OA simply means that articles are immediately made available to all, regardless of business model.
• APCs are not the only business model available. Another possibility is for learned societies and/or institutions to fund the costs of running a journal. This is the case, for instance, with the journals running under the eLanguage programme, including the very successful Semantics & Pragmatics, and my own Journal of Historical Syntax. These operate a no-fees gold OA policy.

If the above factual points are accepted, the first three bullet points from the original post lose their force, as they all depend upon the assumption that APCs are the only realistic business model for Gold OA journals. Also, the continuing existence of journals like the Journal of Historical Syntax demonstrates that it is possible for small specialist journals to be OA-compliant.

On a less factual note: for no-fees gold OA to work, a rethink of the costs involved with journal publication is necessary. For a modern e-journal, these are vanishingly low: virtually the only costs involved are web hosting (which costs peanuts) and typesetting. A move towards OA therefore needs to take into account what the Finch Report terms 'disintermediation'. This involves reconsidering the role of for-profit academic publishers, who demonstrably add very little value to the publishing process.

• CC-BY licenses allow the author to retain copyright for their own work, unlike the current policies operated by e.g. Elsevier and John Benjamins.

This is important, as it is a clear benefit to the individual author over the current model. Furthermore, it is not at all clear that CC-BY will cause attribution problems; attribution is attribution, and if it is not present authors will have legal recourse. See this excellent blog post by Martin Eve for more on this issue.

• Business models for open access monographs in linguistics are in development as we speak. The first such endeavour has already been founded by Stefan Müller in Berlin, and manuscripts have already been submitted.

In sum, I think that linguistics has nothing to gain from standing in the way of this development, and that we should support it wholeheartedly.

My only qualm about the HEFCE document is that it does not make provision for any embargo period whatsoever. Some journals (Language, for instance) are moving to a model in which there is a twelve-month embargo period, which would count as OA under the RCUK's current definition but not (necessarily) under the HEFCE definition given in point 11 here. Since HEFCE have stated that they welcome advice, I suggest the LAGB advises them to treat articles with a one-year embargo period as OA for the purposes of the REF. Embargo periods are not ideal, but they would allow for some flexibility during a transitional period.

I find the anonymous 1242488 reply very helpful. Just two points from that post:

• 1.     “access to knowledge should not be restricted to those who are either a) affiliated to a university or b) rich.” Exactly, and this includes access for colleagues in parts of the world where research conditions are dismal. Making research results freely available is something that wealthy countries and (relatively) wealthy institutions can offer fairly easily.
• 2.     “Since HEFCE have stated that they welcome advice, I suggest the LAGB advises them to treat articles with a one-year embargo period as OA for the purposes of the REF.” This is the type of small and sensible suggestion that might just get taken notice of.

Greville Corbett

To respond to David (D.): I think it is useful to mention concerns surrounding Open Data in linguistics at this stage even though nothing specific is being proposed yet. It is likely that Open Data requirements will be considered further and probably introduced formally at some stage. As David points out, the two areas of concern for us are fieldwork notes and recordings, and corpus data. Much of the debate in this whole area has been driven by the sciences and there is a sense that humanities should just follow the procedures that have been proposed in the sciences. I think it is important to point out areas where we are different and can't simply follow. Copyright of third-party material prevents release of data in linguistics and some other subjects. Anonymity of data collected in fieldwork is much harder for us to achieve and is not analogous to, for instance, patient anonymity in biomedical research. For these reasons, I think it's important to say from the start that there may be practical limits on the extent to which we can support Open Data.

(1) comment on open-access monographs:

Stefan Müller and I have been working intensively on a business model for no-fees Gold open-access monographs (see http://hpsg.fu-berlin.de/OALI/). The basic idea is to take book publication back into the hands of academics themselves, and to show that the huge costs charged by for-profit companies are not warranted. Over 200 linguists have joined our initiative, so this seems to be the way of the future. Nonprofit publishing is far more efficient than publication by huge shareholder-owned companies (see Stuart Shieber's detailed discussion of this at http://blogs.law.harvard.edu/pamphlet/2013/01/29/why-open-access-is-better-for-scholarly-societies/). In the longer term, scientific publication should be seen as very similar to doing basic science itself: Both are best supported by public funding, because there is no functioning market and they cannot lead to justified profits. But both lead to prestige, so public (or charitable) funders have a strong incentive to invest in them (see http://www.frank-m-richter.de/freescienceblog/2013/02/14/science-publication-should-be-seen-as-a-public-service-just-like-science-itself/).

(2) comment on the passage "some linguistics journals will never be compliant with open access (particularly small specialist ones and those produced by small publishers in countries with no tradition of open access)":

Actually, the OPPOSITE is true: The major journals in countries like the UK, the US and Germany are not open access, but the linguistics journals in countries like Croatia, Poland, Estonia, Finland, Taiwan ARE open-access:

http://www.linguistics.fi/skyjol-en.shtml
http://hrcak.srce.hr/suvremena-lingvistika
http://versita.com/lp/
http://www.kirj.ee/lu/
http://www.ling.sinica.edu.tw/publ_j_en.asp.htm

This is a general trend, found in all fields. The Brazilian journal platform SciELO (http://www.scielo.org/php/level.php?lang=en&component=42&item=1) has many hundreds of open-access journals from developing countries.

This is easy to explain: Open-access publication on the internet is much cheaper than traditional print publication, and in the less rich countries, the big shareholder-owned companies like Informa (Taylor & Francis), Wiley and Elsevier have not bought up everything yet. So the public funders have decided to make the journals open-access, because otherwise they would have no readers. In the rich Western countries, where we have big library budgets, scholars can just ignore the problem and continue to publish with the usual companies, even though their business model is extremely inefficient (see Stuart Shieber's post mentioned above).

So we should emulate the Estonians and Brazilians: We should submit our papers to no-fees ("Platinum") open-access journals, and we should start new ones, such as the Journal of Historical Linguistics mentioned in the 1242488 reply. The support can come from institutions (such as universities or research institutes), charities, or scholarly associations such as LAGB. In the future, open-access journal platforms such as Revues.org can make it even easier for decentralized journals to operate. This would mean venturing into uncharted territory, but other organizations such as SLE and DGfS are also discussing similar steps.

Martin Haspelmath, Leipzig