All posts by Paolo Mangiafico

Equity in Scholarly Communications – join us for TriangleSCI 2019

Triangle SCI logoReaders of this blog may know that since 2014, Duke University Libraries have been hosting the Triangle Scholarly Communication Institute, in partnership with our colleagues at NC Central University, NC State University, and UNC-Chapel Hill, with financial support from the Andrew W. Mellon Foundation. Each year we invite proposals from teams (rather than individuals) to work on projects they define within their own team and across other teams working on projects around a common theme. This happens across four days in October in Chapel Hill, North Carolina, and expenses for participants whose proposals are selected are covered by a grant from the Mellon Foundation.

We’ve recently announced this year’s theme – Equity in Scholarly Communications – and invite proposals from you! You can find more details in the announcement below and on the trianglesci.org web site, or by following @TriangleSCI on Twitter or seeing what others are saying in the #TriangleSCI hashtag.

Teams from past years have gone on to launch programs such as these, initially developed at TriangleSCI:

Should you #DeleteAcademiaEdu?

[ Note: Many readers of this blog have probably heard by now that Kevin Smith, who has been the primary author here, will soon be leaving Duke to be the Dean of Libraries at the University of Kansas. We do intend to keep the blog going, and to continue to address the same issues you’ve come to expect from the site, though with a greater variety of authors. So do stay tuned. This post is by Paolo Mangiafico.]

Yesterday afternoon a kerfuffle arose on Twitter about Academia.edu, a social networking site for academics, where many academic authors have profiles, share their publications, and connect with other scholars. You can read about the beginning of the controversy in this article the Chronicle of Higher Education posted this morning.

The ensuing tweetstorm followed a fairly typical trajectory – moral outrage, call to action, a hashtag, and then of course the inevitable backlash, with each side calling into question the integrity of each other’s motivations, or at least the consistency of their actions.

The chief concern, or at least the one that appears to have caused the most heated debate initially, was whether paying for promotion of one’s scholarly work was equivalent to “vanity publishing”, but the discussion evolved into the broader issue of whether the fact that Academia.edu is a commercial service meant academics should avoid it, with several people on Twitter calling that out as hypocrisy, given the many other commercial transactions that academic life is entangled with.

My own opinion is that this is a straw man argument, and it misses an opportunity to have a more nuanced discussion about what’s really at stake here. This isn’t a morality play, and it’s not about whether charging for “monetizing” something is in itself a bad thing – for me it’s about choices, and making informed choices about keeping or ceding control to one’s own work. It’s also about being open vs being closed. Despite the impression that #DeleteAcademiaEdu is just railing against capitalism, I’d argue that it’s really about promoting a more competitive marketplace, one where the data is open for any number of potential services (consortial, member-supported, or even commercial) to do interesting and useful things with it – may the best service win, or may many complementary services thrive.

The challenge with sites like Academia.edu is that this is not possible. By most accounts, Academia.edu is a fine service, and clearly it’s meeting a need, as the number of academics who have profiles in it shows. They are doing very well at motivating academics to put their profile data and publications there. But what happens to that information once it’s there? By my read of the site’s terms of service, no other uses can be made of what you’ve put there – it’s up to Academia.edu to decide what you can and can’t do with the information you’ve given them, and they’re not likely to make it easy for alternative methods of access (why would they?). There doesn’t appear to be a public API, and you need to be logged in to do most of the useful things on the site (even as a casual reader). They were among the first to create enough value for academics to encourage them to sign up, and kudos to them for that, but does that mean your profile data and publications should be exclusively available via their platform? This is what’s called “vendor lock-in” – it’s very good for the vendor, not so good for the users.

While it’s understandable that companies will try to recoup their investments through such approaches, it nonetheless goes against the ethos of academia, and of how the Internet functions best. A few years ago at a conference I heard a speaker say

On the Internet the opposite of ‘open’ is not ‘closed’ – the opposite of ‘open’ is ‘broken’

(If I remember correctly, it was John Wilbanks)

So yesterday when I first started reading some tweets about people deleting their Academia.edu accounts, I tweeted

VIVO is an open source, open access, community-based, member-supported profile system for academics. It has been implemented by many universities and research organizations, and makes linked open data available for access and integration across implementations. In some institutions, like my own, it is connected to our open access institutional repository, so Duke researchers can easily make the full text of their publications be linked directly from their profile – open to anyone, no login required, always in the author’s control. And the custodians of the system and the data are the researcher’s home institution, as well as…  well, here I’ll quote from an article Kevin and I wrote a couple of years ago:

“this brings us to a discussion of another major player in this ecosystem that we have not yet addressed—a set of organizations that are mission driven, rather than market driven; that are widely distributed and independently operated, and therefore less vulnerable to single points of failure, and that were designed to be stable over long periods of time; that are catholic in their scope, strong supporters of intellectual freedom, and opponents of censorship and other restrictions on access to knowledge; and that are in full alignment with the mission of learning, teaching, and research that constitutes the primary reason why authors write academic articles. We are, of course, talking about libraries.”

This, ultimately, is why I think scholars will be better served by having the core data for their profiles and their research tied to open systems like VIVO, and to their universities and their libraries. Sure, the interfaces might not be as elegant, and we might move more slowly than a commercial service, but we’re in it for the long haul, we share your values, and we’re not going to try to lock in your data.

If someone wants to harvest the data from VIVO and our repository and layer on a better social networking or indexing service, that’s great – the data is available for that, and we have an open API. Do you want to charge for the service? No problem, as long as the people you’re charging know that they’re paying for your service add-ons, and not the data itself, which remains open and free to anyone else to use it outside the paid service. Do you have a service (like Academia.edu) that’s really good at convincing authors to enter their CV and upload their articles? Wonderful – make the data available unencumbered, and we might be willing to pay you to do the collecting for us (especially since institutional repositories haven’t been as successful in doing so).

The key reasons why authors should choose first to work with their scholarly communities rather than purely commercial enterprises isn’t that making money is bad – we all have to earn a living – but that the goals and values aren’t necessarily in alignment. I’ve used a lot of words to say something that Katie Fortney and Justin Gonder said in December (in “A social networking site is not an open access repository”) and Kathleen Fitzpatrick said a few months before that (in “Academia, Not Edu”), but the Twitter discussion sparked yesterday has made many more people aware of this issue, so I wanted to underline these ideas, and say a bit more about it than would fit in my tweets yesterday afternoon.

You have a choice, and the choice I hope you will think more about is whether you feel more comfortable investing your time and efforts with your home institution and your library, whose incentives and values presumably align with your own, and who will contribute to an open ecosystem, or with a service whose incentives and values and life span are unknown, and whose business model relies on being closed. If you’re comfortable with the trade-offs and risks, and willing to exchange those for the service provided, then don’t #DeleteAcademiaEdu. But I hope you will use this opportunity to look into whether alternatives exist that will meet your needs while keeping your options open and your data open, and preserving your ability to keep control of your work and make sure it’s not helping sustain an ecosystem that’s broken.

——

If you’ve read this far, I hope you’ll also tolerate this shameless plug for an upcoming event that will be a forum for addressing many of the issues discussed above – the Scholarly Communication Institute. The theme of SCI 2016, to be held in Chapel Hill, NC, in October, is “Incentives, Economics, and Values: Changing the Political Economy of Scholarly Publishing.” We invite teams to submit proposals of projects they’d like to work on that fit this theme, and to build a dream team of participants they’d like to spend 4 days with working on it. For proposals that are selected, we pay expenses (thanks to a grant from the Andrew W. Mellon Foundation) for the team to come to the North Carolina Research Triangle and work on their project alongside several other similar teams, in an institute that’s part retreat, part seminar, part unconference, and part development sprint. You can find out more about the institute at trianglesci.orgproposals are due March 14, so if you’re interested, start putting together your team soon.

Open Access at the tipping point

Open Access Day bookmark used under CC-BY license from http://www.openaccessweek.org/page/englishhigh-resolution-1[ guest post by Paolo Mangiafico ]

As readers of this blog almost certainly know, this week was Open Access Week, and it’s been heartening to see all of the stories about how open access is creating new opportunities for scholarship, and transforming scholarly communication.

It’s also been interesting to see organizations that one might not think of as being open access proponents proclaiming their OA bona fides this week. On Tuesday this press release from Nature came across my Twitter feed. I shared it with my colleagues Kevin and Haley, joking that our job was done and we could go home, now that even in Nature over 60% of published research articles were open access under Creative Commons licenses.

Even though Nature neglects to mention in this release that they are bringing in a lot of money from open access through high article processing charges (they aren’t doing this just to be nice) I still think it’s an important milestone because it shows that open access is becoming the norm, even in mainstream, high visibility journals. I’m optimistic that this is another indicator that we’re on our way to some kind of tipping point for open access, where other effects will come into play.

One of the statistics given in the press release is that the percentage of authors choosing CC-BY licenses in Nature Publishing Group’s open access journals rose from 26% in 2014 to 96% in September 2015. Just last year, a study by Taylor & Francis indicated that, when asked (or at least when asked with the leading questions in the T&F study), authors were more likely to choose other CC variants, yet in Nature open access journals the choice of CC-BY is now nearly unanimous. Maybe “choice” is too strong a word – they appear to have achieved this primarily by setting CC-BY as the default. Just as in the past when signing over all your rights to a publisher was the default (and, unfortunately, in many journals still is), it seems that few authors realize they can make a change, or see a strong reason to do so. What this signals is the power of setting a default.

When we were working toward an open access policy for Duke University faculty in 2010, we talked about setting the default to open. As we discussed the proposed open access policy with Duke faculty, we never called it a mandate, and we haven’t treated it as a mandate, in that the policy doesn’t force anyone to do something they are disinclined to do. But absent any expressed desire to the contrary (via an opt out) the policy enabled the faculty and the University to make as much scholarship produced at Duke be as widely available as possible. We approached the policy as a default position, and built services to make it easy for Duke authors to make their work open access via an institutional repository and have it appear on their University and departmental profile pages, so there are few reasons now not to do it. It will still take time, but I think this “green” open access option is something authors will increasingly be aware of and see as a natural and easy step in their publishing process. They’ll see open access links showing up on their colleagues’ profiles, being included in syllabi and getting cited by new audiences around the world, and linked from news stories, for example, and word of mouth will tell them that it’s really easy to get that for themselves too.

What makes me optimistic about the figures in the Nature press release is that they point to an environment where even in high visibility journals open access is no longer that thing only your activist colleague does, but is something that many people are doing as a matter of course. And as the percentage of authors making their work open access grows, suddenly various decision-making heuristics and biases start to tip in the other direction. Pretty soon the outlier will be the scholar whose work is not openly available, either via “green” repositories or “gold” open access journals, and I think momentum toward almost universal OA will increase.

Our work isn’t done, of course. Even with open access as a default, the next challenge will be to manage the costs. So far the shift to OA has mostly been an additional cost, and the big publishers who made big profits before are continuing to make big profits now via these new models. Even as OA becomes prevalent, and scholars see it as the norm, we’ll still have to work hard to find ways to exert downward pressure on author processing charges and other publishing costs, so that open access doesn’t just become another profit center that exploits scholarly authors and their funders and institutions. We need to do better to surface these costs, and to put in place mechanisms and perhaps shift to supporting other publishers and other models that will keep costs down.

But for now let’s call this a victory. Recognizing there’s still a lot to do, let’s pop the champagne bottle, celebrate open access week, and then get back to work on the next round of creating a better scholarly communication ecosystem.

 

Who pays, and what are we paying for?

[ guest post by Paolo Mangiafico ]

I wasn’t at the Society for Scholarly Publishing’s annual meeting in Virginia last week, but was able to follow some of the presentations and discussions via the #SSP2015 hashtag on Twitter and some followup blog posts. Something that caught my eye yesterday was a post on Medium by @CollabraOA titled “What exactly am I paying for?” that summarized a panel discussion at SSP on the topic of “How Much Does it Cost?” versus “What are you Getting for/doing with the Money?” An Overview and Discussion of the Open Access Journal Business Model, (lack of) Transparency, and What is Important for the Various Stakeholders.

The post has summaries (and links to slides) of the presentations by panelists Dan Morgan (University of California Press), Rebecca Kennison (K|N Consultants), Peter Binfield (PeerJ), and Robert Kiley (The Wellcome Trust), as well as links to other readings on the topic, such as this article from a couple of years ago titled “Open access: The true cost of science publishing” by Richard Van Noorden in Nature.

A few things from the summary of the panel discussion that stood out to me (excerpted or paraphrased here):

  • From Robert Kiley’s discussion of the Wellcome Trust’s experience with paying article processing charges (APCs) on behalf of their funded authors: the average APC levied by hybrid journals (which publish both subscription and OA [open access] articles) is 64% higher than the average APC charged by wholly OA, or “born OA”, journals. Despite these higher prices, some of the problems the Trust have encountered, such as articles not being deposited to Europe PubMed Central, incorrect or contradictory licenses appearing on articles, and confusion as to whether the APC has been paid, were almost exclusively related to articles in hybrid journals. Robert asked: “Are we getting what we pay for?”
  • From Rebecca Kennison’s discussion on transparency of publishing costs, and how the initial APC for PLOS Biology was set when it was launched: it was based on the average price paid by authors publishing in that era’s top science journals, for page and color charges, etc. The thinking was that if biology authors are used to paying around $3000 USD to get published in a subscription journal, they will be able to transfer this to pay the APC for PLOS Biology instead. She noted how much of a role this $3000 price point has played in OA price-setting since the early 2000s. This is fascinating when you consider that it was a “What the Market Will Bear” price point, and not based on publishing costs. / The desire for transparency is not so much to make publishers reveal all costs, or push publishers to offer services “at cost”, but to ensure that librarians and funders, or anyone paying an OA charge, are simply more aware, and sure, of what they are paying for, and whether it is the best use of funds. It is not a matter of caveat emptor, but emptor informari.
  • From Pete Binfield’s discussion of the relationship between cost and prestige: despite the fact that “born OA” publishers can be much more efficient, authors still seem to be willing to pay for things like “prestige” and “the best venue for discoverability,” where more traditional publishers are still perceived to have an advantage because of established “brands.”

This discussion resonated with a different one that has been playing out among anthropologists in the past few weeks, regarding whether and when to transition the long established journals of the American Anthropological Association (AAA) to open access, a process that has already begun with the high profile Cultural Anthropology journal.

In an editorial in the February 2015 issue of American Anthropologist, the editor, Michael Chibnik, argued that while he “cannot disagree with the rhetoric of those advocating open access for American Anthropologist” he also could not see how to make the finances work without continuing to rely on the existing subscription model via a publisher like Wiley Blackwell. While admitting “I do not know all the details of the financial arrangements between AAA and WB” (see discussion about the lack of transparency explored in the panel mentioned above) he briefly outlines why several alternative funding models he has heard about are unlikely to work, concluding “The obstacles to AA becoming open access in the near future may be difficult to overcome.”

This elicited several responses, from Martin Eve, who challenged many of the assertions in the piece, one by one; from the Board of the Society for Cultural Anthropology, who argued in a commentary titled “Open Access: A Collective Ecology for AAA Publishing in the Digital Age” that open access was the right thing to do despite the difficulties; and from Alex Golub, who wrote a blog post titled “Open access: What Cultural Anthropology gets right, and American Anthropologist gets wrong.”

The Society for Cultural Anthropology commentary points out that research libraries are key stakeholders in the emerging OA landscape, and potential partners with scholarly societies for new models of scholarly publishing. Both SCA and Golub reference some new projects like Collabra, Open Library of the Humanities, Knowledge Unlatched, and SciELO, that, in Golub’s words, “blur the distinction between journal, platform, and community the same way Duke Ellington blurred the boundary between composer, performer, and conductor” and are examples of “experiments to move beyond cold war publishing institutions.”

It’s not clear yet what financial models will ultimately prove successful and sustainable for scholarly publishing and scholarly societies going forward, but simply maintaining the status quo with its hidden and inflated costs and frequently vestigial practices is almost certainly not the answer. As Alex Golub concludes in his post:

The AAA wasn’t always structured the way it is today, and it may not be structured this way in the future. The question now is whether the AAA can change quickly enough to be relevant, or whether institutions like the SCA are the true future of our discipline. These are issues tied up with a lot more than just publishing: The shrinking of academe, the growing role of nonacademic stakeholders in academic practices, and much besides. Does Cultural Anthropology face a lot of issues down the road? Absolutely. Is complete and total failure on the menu? Yes. But I reckon that in ten years when I sit down to reblog this post, we will look back on this debate and say: The people who did the right thing and took a leap of faith fared far better than the ones who clung to a broken solution. Cultural Anthropology acted like Netflix, while American Anthropologist acted like Blockbuster. Except, of course, no one will remember what Blockbuster was.