Wikipedia, authority and the free rider problem

I am a selfish Wikipedian. By which I mean, that while I am very happy to use Wikipedia, I have not been very serious about contributing to it. There are a small handful of pages for which I keep the further reading (reasonably) up to date, and correct if a particularly egregious error appears.  But it is sporadic, and one of the first things to be squeezed out if life gets busy.

And I wonder whether there aren’t real gains for historians from helping Wikipedia become truly authoritative, but which are obscured by natural disincentives in the way in which our scholarly ecosystem works.

Firstly, the disincentives. One is a residual wariness of something that can be edited by ‘just anyone’. I myself have dissuaded students from citing Wikipedia as an authority in itself, as part of what I am teaching is the ability to go to the scholarly article that is cited in Wikipedia, and indeed beyond it to the primary source. But my experience is that, in matters of fact, Wikipedia is very reliable unless it concerns a highly charged topic (the significance of Margaret Thatcher, say). And even the making of that judgement is an important part of learning to think critically about what it is we read.

Perhaps more significant is the fact that Wikipedia appears to be edited by no-one in particular. One of the contradictions of modern academic life is that most scholars would, I think, assert the existence of a common good, the pursuit of knowledge, towards which we work in some abstract sense. At the same time, the ways in which we are habituated to achieve that end are fundamentally about competition between scholars for scarce resources: attention, leading to esteem, leading to career advancement.

We write books and articles, which help us get and then keep a job. A smaller but growing number write blogs like this one, and tweet about those blogs. Part of this is about ‘impact’ (that is to say, increasing our share of those scarce quanta of public attention). And all of it depends on being identified as the creator of an item of intellectual property: tweet, blog post, article, book, media interview. Few, even at the wildest edges of the Open Access movement, propose licensing of scholarly outputs without attribution, even if a work may be licensed for the most radical of remixing. All depends on being known.

But Wikipedia doesn’t credit its authors, or at least not in a prominent and easily reportable way. And so the question arises: even though contributing to Wikipedia is to the common good, what is in it for me ?

The answer may depend on a more speculative and more risky model of collaborative work, but one which holds out the prospect of a genuinely authoritative resource, made by authorities. And that in turn should reward the best published work, in the good old-fashioned and citable way, by channelling readers to it. (It would be even better for works available Open Access.)

But it depends on everyone jumping together. As long as some contribute, but others only consume, there remains a classic economist’s ‘free rider’ problem. When people use a resource without ‘paying’ (in the form of their own time, and their own particular expertise) then the cost of production is unevenly spread, and the quality of the product denuded. But if editing Wikipedia became a genuinely widespread enterprise amongst scholars, then even if my contribution is not recognised with each and every edit, my ‘main’ work (if it is any good) will be cited and integrated into the fabric of Wikipedia by others. And we might get a more informed public debate about each and every matter, which looks like impact to me. Perhaps I should get more serious about this now.

Open Access and open licensing

Much of the recent concern about Open Access in the UK, at least for the humanities, has not been about the general principle, but rather about the means.

In my hearing, however, perhaps at least as much consternation was in reaction to the prospect of subsequently licensing those outputs for re-use using one or other of the Creative Commons suite of licences. CC allows various degrees of redistribution, and re-use, without further recourse to the author, but with credit given. Commercial use can be restricted (or not); the making of derivative works can be provided for (or not). You can Meet the Licenses here.

As an advocate of greater Open Access in the humanities, I suspect that Research Councils UK made a tactical error in suggesting that it intended to enforce the most liberal of these licenses. CC-BY ‘lets others distribute, remix, tweak, and build upon your work, even commercially, as long as they credit you for the original creation.’ Here’s why I think the focus on CC-BY has been a mistake, at this point.

Personally, I have never quite been convinced that ‘full’ or ‘real’ OA was dependent on maximally open licensing. I see free availability of the content for reading and citation as quite distinct from the subsequent reuse of that content in other ways. Both are desirable, but can be decoupled without damage. A move to any form of OA represents a major cultural change, albeit one that is necessary. Given this I would rather see an OA article with all rights reserved (as a staging post) than to not see that article at all. And to couple the two too closely risks the first goal by too strong an insistence on the second. Over time, cultures can and do change; but we ought to practice the art of the possible.

More generally, it isn’t yet clear to me what re-use of a traditional history article looks like. Quotation (with a reference) is a mode historians understand; so is citation as an authority in paraphrase. Both are possible from an article with all rights reserved. Compilation of readers and anthologies would be made easier by CC, but doesn’t require CC-BY. It also isn’t clear what ‘remixing’ of traditional historical writing looks like if it doesn’t involve quotation. Historians are also well used to acknowledging a seminal work in a footnote (or even once only in foreword or acknowledgments) without quoting it directly, but is this all that giving ‘credit’ for ‘remixing’ an idea really means ? If so, there is little to fear; but I’m not sure we know, yet.

Over time, there will be possibilities for data-mining in corpora of scholarly articles, but we ought to think on about whether this can be accommodated without full CC-BY. Much turns on the question of what counts as a derivative work in the context of an aggregated database, and what the output to the user is; and whether an insistence on  non-commercial re-use shuts down important future possibilities that we can’t yet foresee.

It may be that CC-BY is the right default option; my feeling is that it probably will be. But I think we should probably take more time to document some of these use cases, in order to plan a movement towards licensing for historical writing that is neither more restrictive nor more liberal than it need be, and allows scholars to dip in their toes without plunging in up to the neck. For now, there are horses we should avoid scaring, lest they bolt.

Humanities publishing and the Finch report

[The text below appeared in the Annual Review of my former employer, the School of Advanced Study, just before Christmas. Since I finished writing it, the debate about Gold open access in the humanities had continued, with no little sound and fury concerning the statement from the editors of some twenty prominent historical journals, most interestingly from Cameron Neylon. Re-reading my piece now, it strikes a more conservative note than I intended, since I spent some three years preaching the benefits of OA, green and gold, in a HSS institution, and have been delighted to see what was a rather marginal issue move to centre stage. There are issues to be addressed, but HSS scholars and journal editors do need to join the debate, robustly but openly and constructively, since if heads become buried in sand we shall have a model suited to the natural sciences imposed on us whether we like it or not. The goal of maximal open access is (I think) clear; let's make it happen.]

It is now ten years since the seminal Budapest declaration on Open Access, and eight years since parliamentarians first endorsed the general principle that publicly funded research ought to be available free at the point of use. And whilst the natural sciences have embraced Open Access very fully, the situation in the arts and humanities is very different. As I argued in Research Fortnight this summer (25th July), for all the talk of Open Access coming of age, the humanities are in danger of being left behind.

However, since the publication of the Finch report in the summer, the issue has moved to centre stage. The UK government has strongly supported the report, and so after a decade of debate, the general thrust of its proposals seem set actually to be implemented. Yet grave reservations have been expressed, not least in the two recent statements from the American Historical Association and, in the UK, from the Royal Historical Society.

One main source of concern (which matches my own) is its support for the ‘Gold’ route to open access, based on the ‘author pays’ principle. Instead of the publisher’s costs being covered by payment from the reader (or their library), the publisher charges a fee to the author, but access to the work is free at the point of use. The model has an appealing simplicity, and in theory should make a work available to anyone who might be interested in it, rather than simply to those with access to a research library. It is already well established in areas of the natural sciences, and in small pockets of the humanities, notably in the history of medicine. However, there are significant issues in its implementation, the most significant of which is the impact on those who cannot pay.

The Gold model works best when research is funded by direct grant, with a small additional sum to cover publication fees. But a vanishingly small proportion of humanities research is funded on this basis, and so those fees must be met by some other means. The government has pledged extra funds to cover this, but only to a number of research-intensive universities, which sends a clear and unwelcome signal about the prospects for research produced in other HEIs; to say nothing of early career researchers in (and out of) short-term positions and the army of independent scholars producing first class work outside the universities. Looking back at my own publications, I cannot imagine how any of them could have been funded in this way; and so they would not now exist.

There is still room, however, for dissenting voices to be heard; and there is an opportunity for the School and its Institutes to take the lead in creating the spaces in which those conversations may take place. Through SAS-Space, the establishment of SAS Open Journals, and associated events, the School has taken part in these debates over the last few years; may it continue to do so.

Humanities left behind in the dash for Open Access ?

In the last few days I’ve been very gratified at the reception (insofar as I’m aware of it) to my article in last week’s Research Fortnight. I’m particularly grateful to the following Twitterers for their kind comments:  @Ghaylam, @Emmanuel_clerc, @rmathematicus, @beckyfh, @j_w_baker and also @ukcorr . It happens also to have appeared at the same time as other significant blog posts on similar themes, by Sara Dorman on the DeadDogBlog  and by Mark Carrigan (@Mark_Carrigan), both drawing attention to the implications of a shift to Gold OA on those without the means to pay, as I do.

The article notes the disparity between the adoption of OA, both green and gold, between different parts of academia, and just how far the humanities are behind. This is based on some sample research I’ve carried out in the last few months, the detail and methodology of which I’d be happy to share. It then goes on to examine some of the reasons, one of which is the speed with which research passes out of date:

It is rare to find competing research groups racing to find the historical equivalent of a cure for cancer or the Higgs boson. Humanities research often retains its currency for a good deal longer than work in the natural sciences, and so there is not the same need for speed; a lag of a year or two between submission and publication is not felt so keenly. The most downloaded of my own papers in 2012 is also the oldest, published in 2006 and largely written in 2004.

The second reason is the small proportion of humanities work which is funded directly by research grant, and again there are some impressionistic numbers for this, for what they’re worth:

…. but there is clearly a gulf between the amount of research being published and the amount that is directly funded. If this is to be bridged, universities will need to find funds to cover the upfront charges for gold open access for their staff…..

which will need to be found from somewhere; wiser heads than mine will need to figure that out. Of my ten articles to date, not a word has been directly funded by any organisation, despite my having worked in UK universities for all that time.

The final point relates to  independent scholars, of whom there are a great many publishing top-drawer work in the humanities, not to mention post-doctorate scholars looking for a job, and everyone else to whom university publication funds won’t be open.

By and large, humanities scholars do not need large capital equipment and facilities, beyond a good library. As such, scholars outside universities—in museums, libraries, archives, across the professions and not least among the retired—regularly publish world-leading research. Universal gold open access funded by the author would wipe much of this work out.

It ends:

All the disciplines stand to gain from a successful move to open access. However, much of the discussion about open access has been driven by the needs of the sciences. Let’s not allow the humanities to be collateral damage along the way.

Read it all here.