Very important

Ranking position


Very important plus important

Not at all important

Not very important

Not at all important plus not very important

The principle of free access for all readers








Table 4:  Authors’ reasons for publishing in Open Access journals




Permitted by publishing agreement

OA authors

Non-OA authors

As a preprint



In final, peer-reviewed and edited form



As a PDF supplied by the publisher



None of these



Table 11:  Permission from publisher of authors’ last article to allow authors to post their own article online


4.8   Article repositories and archiving



Archiving behaviour

Open Access authors

Non-Open Access authors


Preprint form

Final peer-reviewed form

Preprint form

Final peer-reviewed form

Posted an article on my personal web page





Posted an article on my department’s web site





Deposited an article in an electronic institutional repository





Deposited an article in an electronic subject repository





Table 16:  Self-archiving behaviour of respondents


Note that this is not really an estimate of the population proportion of self-archiving; even less is it an estimate of the population proportion of OA publishing!


It is also unclear whether these figures are additive or subsumptive, for different forms of self-archiving or for preprints and postprints.


4.8.2   Familiarity with electronic archives

It is possible that the number of people who have archived their work electronically so far is low because most are unaware of the opportunities to do so. Q35 in the OA survey, and Q27 in the non-OA survey, were designed to find out how familiar open Access authors are with the present range of electronic article repositories in existence.  It gave respondents a list of types and asked them to say which they were familiar with.  Because the terminology may be unfamiliar here we gave the respondents some examples of each type of repository to help them recognise the categories.


Most authors in both groups (71% of the OA authors and 77% of the non-OA authors) were familiar with none of them.  The type of repository that the greatest number were familiar with was subject repositories or archives.  Examples of these are arXiv or SPIRES.  15% of OA respondents and 9% of non-OA respondents are familiar with this type of repository.  8% in each group are familiar with institutional repositories, such as D-Space, and 8% are familiar with superarchives such as FirstGov for Science.  Finally, 6% of OA authors and 3% of non-OA authors were familiar with networked repositories such as DARE or RLN.   The results for this question suggest that to date there is very little awareness in general of repositories archiving scholarly articles even amongst authors who have published in Open Access journals.


Correct, and very important!  The implication is that if institutions and research-funders want to promote OA, they should not only, or even primarily, promote and support OA journal publishing, but they should promote and support OA self-archiving.


4.8.3   Willingness to self-archive

The next question asked authors to say how they would feel if their employer or funding body required them to deposit copies of their published articles in one or more of these repositories.


The vast majority, even of the non-OA author group, said they would do so willingly.


Correct, and very important!  The implication is that if institutions and research-funders want to promote OA, they should not only, or even primarily, promote and support OA journal publishing, but they should promote and support OA self-archiving. And a unified OA provision policy should mandate OA provision (by whichever of the two means is suitable for a given article:


A small proportion would do so but unwillingly and 3% of each group would not be prepared to do so.  The full results for both groups are shown in Table 17 below.  Figures are percentages of respondents.



OA authors

Non-OA authors

I would do so willingly



I would do so, but unwillingly



I would not be prepared to do so



Don’t know



Table 17: Willingness of authors to deposit articles in an open repository if required to do so by their funder or employer



  More studies are required on the impact of open access publishing so that the hypothesis that it increases the impact of a piece of work can be fully tested.  There would be no more persuasive argument in the eyes of authors than this.


Hear Hear!



Eprint archives


Whatever the form, apart from the long-established subject archives such as arXiv and CogPrints, they are largely languishing in the doldrums in comparison to the activity level that is possible. 


Self-archiving is definitely languishing (36%) relative to what is possible (100%), and even relative to what already has the journals’ official green light (55%) but not languishing relative to the other roads to OA (OA journals publishing or paying journals to self-archive for you) (5%)!


The reason for the languishing (of OA in general) is merely uninformedness. This will be remediable by more data on the access/impact relation and on the benefits of mandating institutional OA provision policies.


That is not to say interest and the deposition of articles is not growing: OAIster, now harvesting from some 80 eprint repositories plus other sources, reports the current numbers of text items linked to from its records as over 1.5 million, and has seen a 23% increase in the last 5 months. There is a caveat here in that not all of these items will be eprints (i.e. articles from, or destined for, peer-reviewed scholarly journals), though many will.  It is to be hoped that future figures from OAIster will help to clarify how eprint numbers are accumulating.  Other studies are also measuring archive growth rates, and it will be interesting to monitor this over time. 




We also need total annual article counts for OA journals, so the growth rates of the two roads to OA can be monitored and compared. 


Growing they may be, but eprint archives are not garnering the volumes of articles they should be if the majority of scholars were depositing copies of their works in such repositories.   In this present study, fewer than ten percent of authors in either group have deposited an article in an institutional eprint archive and only a few more have used subject repositories either.  Almost certainly this is due largely to ignorance or inertia.


There is a misunderstanding here. First, both the number and the growth rate of the yearly number of OA articles is far higher for OA self-archiving than OA journal publishing. (The number is at least 3 times as high.) So it is not relative to OA journals that OA self-archiving is underperforming, but relative to its own potential ceiling (100%). And this is underperformance for OA, not for OA self-archiving!


Second, the sampling for this study has been such that it does not provide an estimate of the proportion of OAJ vs NOAJ authors: The 50/50 split was imposed by selection and matching. The only estimates are the ones for self-archiving in the NOAJ sample. So it is incorrect to suggest that OA journals are somehow high and self-archiving is somehow low: The opposite is the case.


And the remedy,  as noted below, is to inform the research and research funding community, with hard data on the OA/TA impact enhancement effect, the correlation between downloads and citations, and the means of implementing institutional OA provision and self-archiving policies.


 We cite ignorance as one factor to blame because so few of our respondents were familiar with the various forms of electronic article archives.  Fewer than ten percent of authors know about institutional archives: only a few more of the open access authors are familiar with subject-specific archives.   Given numbers like this it is hardly surprising that there are not burgeoning numbers of thriving institutional archives: but to turn that on its head, given the number of institutional archives in existence, why do not more authors know about them?  That is one question that must be addressed by the open access movement if it is to move forward.


The major focus of this survey is already part of the answer: Attention and energy are mostly focused on OA journals instead of OA self-archiving, despite the latter’s far greater potential for providing immediate OA! What is needed is a concerted educational campaign directed at institutions and research funders about the benefits of OA and the means of providing it.


The other side of this coin is, even if institutions set up archives and even if authors are made aware of them, will they be used?  Stevan Harnad argues that the greatest enemy of these archives is author inertia65.  Harnad has worked for many years to promote the use of institutional repositories and he contends that authors rationalise their inertia by adopting arguments against open access in general – publication fees, poor impact factor scores and so on – none of which have anything to do with depositing a copy of each completed, refereed and accepted article in an institutional archive. 


These arguments  (at least 31 of them!) are virtually identical to the arguments against OA publishing discussed in this survey! They are not arguments against OA self-archiving, they are arguments against OA. And OA publishing has so far succeeded far less in overcoming them even though all are easily refutable, and indeed stand refuted by the 15% of the literature that has been self-archived (and the 5% that has been published in OA journals).


What is going on here is not just down to author inertia, however. 


Correct, it is also institutional inertia and research-funder inertia. So far they have spent more time on how to pay OA journal fees than in how to encourage (or batter, mandate) OA self-archiving!


Stephen Pinfield, whose institution (Nottingham University) is a lead partner in the SHERPA project,  has also had considerable experience in this field and raises a number of factors that authors view as potential barriers to depositing their work in institutional archives: these include technical issues (their ability to submit articles in an accepted format such as HTML, the submission process itself which authors may not feel technically competent to carry out), various concerns about preprints (as opposed to postprints) being available as well as other concerns about ‘quality’, and worries about intellectual property rights and copyright infringement67. 


The replies to these are mostly in


but would it not be useful to look at the methods of the small number of institutions that have succeeded in generating substantial OA self-archiving (Such as Cal Tech and Southampton)? Nottingham is notably not one of them!


All of these issues have simple solutions but there is a cultural problem of author resistance that needs to be tackled.  The case that institutional archiving is supplementary to publishing their articles in journals of their choice needs to be made.  In addition, there is the powerful argument that placing an article in a repository increases its visibility and therefore its potential impact.  Around 55% of publishers surveyed by  Romeo


permit authors to post a copy of their published article in an eprint repository.  Whilst this may have some long-term implications for publishers, in the short-term it is the surest way to increase exposure of an article and maximise its impact.


Author resistance is one thing, but it seems it frequently comes coupled with institutional resistance.  Research institutions need to be convinced of the worth of institutional repositories.  There are two complementary approaches.  Pinfield’s argument rests mainly upon the case for managing institutional information assets – retaining intellectual property rights and copyright within the institution rather than allowing them to seep away as authors sign them over to publishers68. 


This is a library view.  It does not resonate with researchers. The largest group of self-archivers, the high-energy physicists, who have been doing it since 1991 and have long reached 100%, are not doing it so as to “manage institutional information assets.” Nor have they taken any steps to retain intellectual property rights! They simply self-archived – for the sake of the access to and impact of their research.


The problem of encouraging researchers to self-archive must be based on researchers’ needs, not on libraries’ needs. The library community gets the eternal credit for having first alerted us to the problem, but the solution will have to come from addressing the specific needs of the research community.


Harnad argues that the most persuasive reasons for an institution to set up and maintain an eprint archive centre around the increased visibility of its faculty output, maximising citations, impact and thus the overall impact and reputation of the institution.  A new service, Citebase, makes it possible to correlate downloads with citations and thus with the impact of a particular piece of research69, an approach that may prove more persuasive to potential institutional champions than any other. 


Technically, setting up an institutional archive is simple70, 71,


See handbook:


so why are not more institutions going in this direction?  Librarians may be doing their best to lobby for such a thing, but a champion is needed at pro-vice chancellor-for-research (vice provost for research) level.  Growing numbers of individuals at this level are signing up to the Budapest Open Access Initiative.  Given the will, and the deed in the form of a functioning institutional repository, authors can be persuaded to deposit their articles.  It needs some coercion72 – and assistance – on the part of the institution to overcome author inertia, but the vast majority of authors in this study said they would willingly deposit copies of their articles in such as repository if required to by their employer or their funder


(only 3% of authors would refuse to comply), so such requirements need to be implemented.  This is the ‘green’ route that Harnad73 promotes.  From the author’s circumstance, very little changes except that he or she has to make that commitment to ‘self-archive’.


Whither goest?


First, there is the issue of author behaviour.  Academic authors are generally a conservative group of people, intent only upon establishing a claim on their own advances and disseminating their work as widely as possible.  Morris74 argues that author behaviour is far less likely to be driven by altruism than by ‘the normal, human need to make a living through career advancement and research funding’


The benefits of OA provision have nothing to do with altruism: The rewards of enhancing the impact of one’s research are substantial,  as the RAE shows.



It follows, then, that they publish their work in journals with as much impact as possible on their peers – a good reputation, an international audience, a high rank against other journals in the field.  Authors want their work to be noticed, read and built upon.  The most important of these is the first, because the others can only follow from that.  


And they can have all of these – plus the maximisation of the visibility, usage and impact of their articles – by publishing them in whichever journal they wish and self-archiving them!


Why then, when presented with new means of increasing the level of notice taken of their work are authors resistant to adopting it?  Since we already understand from this study that the main reason that authors have not considered publishing in open access journals is that they are not aware of any that they could use one answer is already at hand – increase their awareness.  However, we are not referring to open access journals here so much as eprint archives.  These entities provide an excellent means of increasing the impact of their work, at negligible cost either cash-wise or time-wise, and with no implications (yet) for the authors’ preferred methods of publishing.  They may continue to submit their work to traditional (subscription-based) peer-reviewed, quality-controlled journals, but can increase the impact by placing a copy in an institutional eprint archive.  And yet, they do not do these things.  There is a cultural issue here that open access proponents will need to change.



Hear, hear!