Lisa Janicke Hinchliffe
Lisa Janicke Hinchliffe

Editor’s Note: Today’s post is from Lisa Janicke Hinchliffe. Lisa is the Professor/Coordinator for Information Literacy Services and Instruction in the University Library and an affiliate faculty member in the School of Information Sciences at the University of Illinois at Urbana-Champaign. She served as the 2010-2011 President of the Association of College and Research Libraries.

ResearchGate is under assault. As a scholarly collaboration platform that enables both public and private sharing on a networked scale, ResearchGate is seen as dangerous, not only because it is potentially infringing copyright, but because it is doing so on a massive publisher-independent scale. A group of publishers tried to tame ResearchGate through a proposal that it endorse the STM Voluntary Principles on Article Sharing on Scholarly Collaboration Networks and implement antipiracy measures, but ResearchGate rejected this proposal. Though ResearchGate now faces the threat of thousands of takedown notices and a lawsuit, it is positioned to emerge at least unscathed, if not strengthened, from these assaults.

I’ve documented the timeline and major activities of #ResearchGateGate elsewhere. Now, I’d like to draw out possible inferences and implications for the future of ResearchGate in particular and the state of scholarly sharing more generally. My analysis here is informed by an interview with James Milne, Chair of the Coalition for Responsible Sharing (conducted on October 18) as well as my own experiences as a scholar with a profile on ResearchGate and as a librarian who teaches the “Manage Your Online Scholarly Identity to Maximize the Reach and Impact of Your Work” workshop at my university.

It is hard to imagine any scholar, librarian, or publisher arguing against the foundational claim of the Voluntary Principles – that article sharing is important to the advancement of research. Even so, the scholarly communications community is rife with discussion and debate about how such sharing should take place. Some scholars engage in these conversations — but far, far more just go ahead and share.

porcelain bulls
Charging Bulls, Derby Porcelain Manufactory, ca. 1750, Model attributed to Andrew Planché. Image courtesy of the Metropolitan Museum of Art.

That fact – that scholars just go ahead and share – is the challenge that publishers face with Sci-Hub, ResearchGate, SSRN, Academia.edu, etc. Publishers are understandably hesitant to take action against their own authors. Instead, they are focusing their efforts on developing their own workflow and sharing platforms while seeking to collaborate with or weaken competing ones. These two options are brought into sharp relief by the recent activities related to ResearchGate, the International Association of STM Publishers (STM), and the Coalition for Responsible Sharing (CRS).

As mentioned, publishers could pursue a collaborative approach with a scholarly collaboration network. The Principles contain within them the basis for such an approach and a negotiated agreement would enable publishers to avoid the expense of constant monitoring and takedown notices, and the time delays of litigation.  Fundamental to collaboration, however, is the principle that all parties in the collaboration benefit from the collaboration in ways beyond the benefits they receive without the collaboration. If a platform signs on to the Principles, it is agreeing that “publishers and libraries should be able to measure the amount and type of sharing, using standards such as COUNTER.” But what does the platform gain? If authors are sharing their work in keeping with the Principles, the platform receives the benefit of the Principles without endorsing them. And, if data is the new oil, a platform (such as ResearchGate) would rightly protect its data on usage and sharing as a significant business asset and would not want to trade it away in exchange for access to antipiracy software that is possibly available through a third-party provider (or that ResearchGate may already have). Milne confirmed for me that STM had pursued discussions about collaboration with ResearchGate for two years. The details of those discussions are not public; however, we all know the outcome.

Having failed at the preferred strategy of collaboration, publishers were left with the approach of seeking to weaken ResearchGate. Two strategies are simultaneously being deployed – takedown notices via CRS and a lawsuit filed in Germany by Elsevier and the American Chemical Society (ACS). While they will likely prevail in their takedown efforts and perhaps the lawsuit, their effectiveness at weakening ResearchGate is far from guaranteed.

Though CRS initially indicated that it would be sending 100,000 takedown notices “imminently,” this intention was thwarted by ResearchGate when it removed from public view some of the content that would have been targeted. CRS characterized the amount of removals as “significant” and, though Milne would not provide a specific number citing various reasons, he characterized the remaining notices to be sent as “a modest amount.” As the takedown notices begin to go out, ACS also seems to be trying to assure authors that they are not the object of publisher ire.

Like every other content platform, ResearchGate undoubtedly regularly receives and acts on takedown notices as required by Online Copyright Infringement Liability Act in the United States and equivalent laws globally. Unlike many other platforms, however, ResearchGate also has a well-developed mechanism for private sharing. Faculty regularly tell me that this is one of the features of ResearchGate that they value most – that they now have an online system that stores copies of their publications and facilitates responding to requests. This kind of sharing is in keeping with what publishers typically allow in publishing agreements with authors. (It is disappointing that more institutional repositories don’t offer users this feature.)

Publishers and others would no doubt find it valuable to know the analytics on this sort of sharing and that seems to create an additional business opportunity for ResearchGate. One could imagine a ResearchGate service offering licensed access to this data graph to publishers (that want to know how their publications are performing) and to metrics platforms (that would be able to provide additional value to their customers through the inclusion of analytics calculated on non-public data).

What is most important to notice about ResearchGate’s actions in light of CRS’ stated intention to send 100,000 takedown notices isn’t, however, that it crippled CRS’ ability to take immediate action or that it will now be responding to a “modest amount” of them. What’s important is to note that ResearchGate seems to have the ability to identify copyright infringing material. In other words, it seems that ResearchGate already has some of the technical capacity that STM was offering. Interestingly, ResearchGate has not deployed this capacity to communicate proactively with scholars about legal sharing options and instead relies on authors asserting that they have the right to share their work publicly. It is unclear how deploying this capability will affect ResearchGate’s ability to mount a “safe harbor” defensive claim that it is unaware of the identity of the content on its site.

Concurrently with the announcement of the forthcoming takedown notices came the news that Elsevier and the ACS are suing ResearchGate in Germany.  With no public information (at least not any in English), it has been unclear what the focus of this lawsuit is. Milne explained that the purpose is to pursue a legal finding that ResearchGate has an obligation to deploy antipiracy detection software proactively, analogously to how Facebook or YouTube might prevent upload of a Disney film or music recording. Does availing itself of safe harbor laws for Online Service Providers entail that ResearchGate not only remove infringing material upon notification or coming to knowledge of the infringement in some other, which is a well-established legal responsibility, but proactively reviewing materials to determine whether they are infringing? This proactive review was what STM had sought as a negotiated agreement and now the lawsuit pursues this through the courts instead.

I’m not an attorney but as a lay person it seems to me quite possible that ResearchGate could indeed be found to have this obligation. On the surface, this might appear to be a blow to ResearchGate. Certainly at least some authors will be annoyed by the intrusion of a copyright review step into their sharing workflow, particularly if it is time-consuming, and angry if the system prevents upload due to erroneous calculations. Might it be possible, however, that ResearchGate will be able to present this to users as another value of using ResearchGate as their sharing platform?

Remembering the private sharing feature that scholars already value: one can imagine a user workflow that does not reject an upload because it can’t be shared publicly but instead takes a cue from the message that TurboTax sends when the IRS rejects a tax filing: “This happens from time to time, but we’re here to help.” Wouldn’t most authors prefer a decision support system that walked them through the process rather than trying to figure out SHERPA/ROMEO? If you have a DOI, the DOI-based checking system on HowCanIShareIt can be useful if the publisher’s policies have been included but that’s not a workflow integrated solution. (Libraries should also consider offering this kind of support integrated into institutional repositories. Institutional repositories will still struggle to achieve the effect of a single globally networked platform but enhancing their service profile could make them more competitive with commercial alternatives.)

With a proactive check integrated into the ResearchGate sharing workflow, an author would get a positive outcome – public or private sharing – regardless of the determination. Such a system could also capitalize on the results of Crossmark, the STM Article Tagging Project, and other implementations of the NISO Journal Article Tag Suite to identify whether a file represents the version of record, the accepted manuscript, or the author’s original version/preprint. One could imagine ResearchGate encouraging the upload of all of these versions to its platform and assisting the author in offering them for public or private sharing as appropriate. By doing this, ResearchGate could help authors provide public versions of their work for ease of reading, while facilitating requesting of the version of record when another scholar would like to cite the work.

Authors and readers alike would no doubt prefer immediate access to the version of record without having to request a private share. But, given publishers will likely continue to send takedown notices for any versions of record that do get posted publicly, this public/private system still provides greater value than having to find author manuscripts in distributed institutional repositories that are not connected with a request system for versions of record.

As a platform that enables both public and private sharing on a networked scale that is publisher-independent, ResearchGate already rejected a negotiated solution with STM and seems positioned to emerge comparatively unscathed, if not strengthened, from the takedown and lawsuit processes. Any newcomer seeking to compete with ResearchGate will be faced with the same requirement for proactive checking upon upload if the lawsuit prevails, including any workflow platforms that are in development by companies historically focused on publishing and database provision. And, even if another scholarly collaboration network (e.g., Academia.edu) reaches an agreement with STM, the competitive advantage that might give the other platform over ResearchGate seems minimal given that the user experience of a copyright check during uploading would be similar regardless of how it is brought about. Personally, I wouldn’t count ResearchGate down and out.

Discussion

30 Thoughts on "Guest Post — Does ResearchGate Emerge Unscathed, or Even Strengthened?"

As a member of the STM SCN group who is also active commercially in this space, I would be remiss if I didn’t point out that there is an alternative to things like ResearchGate and Academia.edu and other centralized scholarly collaboration networks that recruit users and content away from publishers — that is, a decentralized scholarly collaboration network, like the one my company has built with Remarq. Remarq goes on the publisher’s sites, and enables legal article-sharing (which can be free to the end-user if the publisher wishes), while bringing all the features of the centralized SCNs to the Version of Record. Remarq allows the actual editors and authors to engage with readers, and gives users the ability to follow articles to receive updates from the actual authors and editors. Publishers keep their traffic and COUNTER statistics whole throughout the article-sharing flow. Users can also take notes on articles, connect with one another, and work in groups.

This approach of flipping the script to a decentralized SCN seems to solve many of the problems publishers have with ResearchGate — loss of usage statistics, loss of readers and authors to another environment and brand, and a long-term strategic disadvantage from both of these factors.

ResearchGate and the others require the kinds of legal responses you’ve outlined above, yet a positive response that is more obviously beneficial to users would help offset the perception that publishers are only trying to return things to the status quo by haranguing users and limiting their options while adding nothing new. Offering a solution that supports authors, editors, publishers, and readers with the kinds of things they’ve learned to like at ResearchGate is definitely part of making this disruption work in everyone’s favor, I think. We need a positive response, not just lawyers, takedown notices, and unflattering news stories.

Kent, this is definitely a “watch this space” developing arena. I’m intrigued by the notion of a virtual SCN but I’m not confident that a decentralized, distributed system can achieve the network/platform effects of a service that is publisher-independent and … I’ll add … not solely focused on articles. More than half of my publications are in books/are books and ResearchGate can accommodate that easily. I’ve also published in print by publishers no longer in business. There is no publisher to support access to my work on their server in a decentralized system. Of course, it’s a fast-developing space so … we’ll see!

Our feeling is that the traffic already traversing publisher sites is significant, and that engagement with authors and editors can be a differentiator. We’ve already seen Remarq behave more “socially” than we expected — more connections, more outbound invitations. I think connecting within trusted spaces is highly desired right now. Good information sources are important, and a network of trust is something people want.

How is remarq different from, say, readcube?
Some publishers (e.g. Springer) offer readcube links to allow access to documents via the authors homepage.

I think there are some big questions in terms of how this impacts ResearchGate from a business standpoint. With venture capital, investors are looking for a return on their money after a relatively short time period. What happens to a company that is facing years of expensive lawsuits that have the possibility of undermining the site’s business activities? Does this harm their perceived value for a potential sale or public offering? Does this result in problems raising further investment funds?

“a positive response that is more obviously beneficial to users would help offset the perception that publishers are only trying to return things to the status quo by haranguing users and limiting their options while adding nothing new. Offering a solution that supports authors, editors, publishers, and readers with the kinds of things they’ve learned to like at ResearchGate is definitely part of making this disruption work in everyone’s favor, I think.”

1 Remarq sounds more like part of the defence of the status quo than a real alternative to ResearchGate.

2 I tried Remarq. Its ORCID linker wasn’t working on 10/26

The status quo, in my opinion, is that publishers continue to push content without developing ways to interact with readers, make articles “living” through author and editor updates (about talks, new findings, etc.), connect and collaborate on the VoR, follow articles in a useful way, etc. Remarq may be more incremental than disruptive, but it’s definitely introducing a set of changes, most of which are reflective of popular elements of the SCNs. Users have said that Remarq makes them feel “closer to the research,” and we see a lot of useful activity with it that wasn’t possible on the VoR before. Making article-sharing more prominent while keeping it legal is also a change to the status quo. Given all the things publishers do (https://scholarlykitchen.sspnet.org/2016/02/01/guest-post-kent-anderson-updated-96-things-publishers-do-2016-edition/), I see no problem helping them upgrade their services while protecting the major benefits and functions they deliver.

As for your ORCID experience, we’ve checked our logs, and we do have a bug in the new desktop version of Remarq, which rolled out last week. Thanks for pointing it out. We’ll fix it imminently. The ORCID integration works in the widget and other places. You can also use Remarq Lite (https://remarqable.com/app/remarq-lite.html) across the web and access the same profile. You can import ORCID via the settings in this.

Posting a pre-publication version is so easy that I cannot fathom why libraries pay enormous sums for the final version. Even conference paper versions, easily matched because they use the same title in most cases, are easy to locate or request. The elephant in the room, of course, is sci-hub, so ResearchGate is moot.

This is a good question, and there are at least a couple of answers.

One is that libraries don’t, in fact, “pay enormous sums for the final version.” Typically, we pay very little for any individual final version. What we pay very large amounts of money for is access to very large numbers of final versions.

The question remains, though: why do libraries pay anything at all for access to final published versions when it’s so easy for authors to post preprints? There are several reasons. One is that although posting preprints is easy, a great many authors don’t actually do it. Another is that the same authors who do (or don’t) post their own articles as preprints also put great pressure on their libraries to provide access to the final published versions of other authors’ articles. Libraries subscribe to journals because their patrons want access to journals, not just to preprint servers.

Libraries often pay for multiple versions of the final articles, all because user demand dictates this buying behavior. Publishers and libraries both serve users — authors, readers, editors, researchers, students, etc. Much of what publishers and libraries do is driven by factors that are not in the control of the library or the publisher. As Rick rightly points out, the dramatic increase in the volume of published content is one such factor — we’ve all been grappling with an avalanche of articles, which have increased aggregate spending, journal size and number, reviewer availability, and more.

Another reason the VoR is so valuable is that it has been through many rounds of review and fact/data checking. As a copy editor at an academic journal, I find significant text and/or data errors in about 33% of the papers I read, including entire data sets that were used or reported erroneously. All of those mistakes were in the preprint, and none of them (I hope) were in the VoR.

Despite the ease, Rick, authors don’t deposit preprints in IRs because they prefer to deposit at least the peer-reviewed version of their papers, for whatever reason. The lengthy embargo periods (up to two years) imposed by many publishers on postprints has had a negative effect on deposits in IRs. Consequently, RG, Academia, etc., are filling the gap between locked-down access to non-OA scholarship and content that is freely available via OA journals. We won’t get into the SciHub piece of this. In short, unrealistic embargoes are making RG investors a lot of money and throttling green deposits in IRs.

There is no reason to believe RG is making money yet. It may be the proverbial Silicon Valley “unicorn.” It has attracted a lot of investment, but the investors are not making money yet, from what I can tell.

Given its advertising business model and the relatively small (compared to Facebook or Google users) number of researchers in the world, I’d be willing to be they’re not even close.

“What’s important is to note that ResearchGate seems to have the ability to identify copyright infringing material. In other words, it seems that ResearchGate already has some of the technical capacity that STM was offering.”

This has perplexed me. Why remove infringing content prior to receiving take-down notices after refusing to filter for copyright infringing uploads? It’s a major blink in this game of chicken. Either way, they MUST filter anyway because they cannot host content for which they already received a TDN– if they took it down once in response to a TDN, they cannot permit the users to post it again. They must be keeping a list of those papers and filtering.

I don’t think ResearchGate will go away because of this. It is a platform enjoyed by many researchers for reasons I am not clear on. I do think there will be some sort of compromise to remain as useful without copyright infringing works. Or they will be bought by a publisher.

It does seem really surprising and contradictory to the notion of any sort of “safe harbor” defense. To claim safe harbor, they have to present themselves as a neutral platform, with no information about, or ability to identify what is uploaded to their platform. Their actions here clearly belie this defense — they know exactly what is up on their platform (further proven by their ability to replace references in uploaded papers with links to other papers on ResearchGate).

It is a platform enjoyed by many researchers for reasons I am not clear on.

Many that I have spoken to say they upload papers in order to alleviate the constant spam from ResearchGate. Perhaps there’s a business model in there — pay us to stop emailing you.

The demands of safe harbor are both clear and unclear. Clear re takedown notices. Unclear re proactive scanning. It is clear that if RG knows a particular item is infringing, action on that item is required. It isn’t clear what level of action is required by having general knowledge that people use your platform in infringing ways.

As to the annotation/cover pages of PDFs, that issue is really secondary to the copyright infringement. If the PDF itself doesn’t infringe because the scholar owns the copyright, then the scholar is free to allow RG to annotate, add a cover page, etc. Myself, I have the cover page feature turned on (and yes this is a user control – I tested this last week) because I find it valuable to have the PDF include where it was retrieved from when people download my work.

To clarify, ResearchGate has engaged in a variety of ways of altering content over the last decade. I’m told there are examples (still available on the site) where they have stripped out the metadata, removed journal/publisher branding (and added in their own), and changed the links in references to point not to the original journal but instead to versions of the referenced papers on their own site (example here: https://www.researchgate.net/file.PostFileLoader.html?id=58c92391217e2031224d805c&assetKey=AS%3A472133680799744%401489576849427). From what I understand, they’ve changed their behavior around to currently adding branded cover sheets, but there is nothing stopping them from going back to their old ways, and they are still liable for these actions, particularly because they’re still distributing these derivative works.

I have to say that this is a major reason why I quite ResearchGate some six or seven years ago – I just go fed up with the spam and the constant harassment to post stuff that was copyright. I stayed with Academia because they didn’t nag me all the time and it is useful to have a place to put some things that time has made difficult for people to access – grey reports for agencies that no longer exist or chapters in books where the publisher has disappeared and they are long out of print.

Robert, I don’t know if you’re affiliated with a university but that is what an institutional or subject repository is for–preserving your scholarship and making it accessible to anyone with internet connectivity. I would not rely on any commercial service, such as Academia, RG, or Mendeley to preserve my scholarship in perpetuity. See http://oad.simmons.edu/oadwiki/Disciplinary_repositories for ideas.

There is no question that publishers do not like ResearchGate with their business model and yes the library community including their researchers use the service without any hesitation. Unfortunately the venture community is behind RG. Looking at the current investors, many of the smartest and successful venture firms are backing RG. These VC firms are long term players in the marketplace and must see a significant upside potential and willing to risk publisher lawsuits. Any one of these VC’s could easily stop any practice that RG is engaged. Look what Benchmark has done to Uber and one can see that the VC is a very powerful voice. So far the VC’s are not putting any pressure on RG to change. So follow the money. Don’t expect RG to change unless the real money folks get involved.

VCs get things wrong all the time. All the time. Juicero, anyone? Just because VCs are involved doesn’t mean the company thus funded is destined to win in the market. As you know, VCs manage portfolios of risk, basically, and the $52M in RG is spread among a number of investors, so the $$ any single player has invested are relatively small. RG is a 9-year-old company, relatively ancient in VCland. The last round of investment was in 2015, almost two years ago. At this point, RG may be a distressed asset to them, and basically written off, which would mean no pressure because no attention is being paid to it. Or they be on the verge of another round and a big pivot to AI. Who knows?

One gets the distinct impression that there is real hatred or disdain between the for-profit publishing industry and its most important base, academics. Without academics, and their papers, publishers are absolutely nothing. What ResearchGate, SciHub, Academia.edu and others represent is a state of academic rebellion, in some cases illegal (if one considers the aggressive legal framework protecting copyright). To protect this new “oil” and the profits it brings to the publishing executive, it was evident that at some point the wealthy, powerful and aggressive publishing industry would fight back. And that’s their right. But, as indicated correctly above, rather than fight the source of their profits, the academics and their institutes, publishers prefer to shoot down the platforms that give academics greater freedom of expression for their work. Academics are tired of the hypocrisies of the industry, they can see right through them. But sadly, a vast majority of academics are too selfish, complacent and lazy to take pro-active action to create something for themselves, and prefer to rely on services, which the entire publishing industry and its cottage sub-industries are. So, academics are to blame for being lazy and ignorant, while their institutes are equally foolish for not providing a self-sufficient platform for their own academics, instead opting to dole out millions or more each year in OA APCs, subscription fees, and other “extortionist” fees to the greed-filled publishing-industry. When will academics wake up and understand the concept of “for-profit”?

So, when I read these guest posts and comments above, I feel like it’s just a bunch of industry insiders squabbling with each other about what the best strategy is to secure profits for the interests they are protecting.

Disclaimer: I publish in many journals of the very same for-profit publishers I critique, because they currently hold the strongest databases (a sickening decision academics have to decide, or publish in idiotic preprint servers). I also have a ResearchGate account.

I understand you are the bloke complaining a lot on Retraction Watch. We all know examples of dodgy and extortionate journal practices, quite a few of them in the commercial world (as RW points out almost daily). Once burned, don’t go near them I would say – I am an academic running a journal with no budget, and quite happy winding down my engagements with the ‘big 5’. I don’t need to write so many long complaints – too busy running an alternative. If more of us would do the same, perhaps there would be incremental change towards more ‘socially just’ publishing. This means either doing it ourselves, keeping to journals published by departments or nonprofit professional societies, and in any event being OA. STM disciplines, particularly Maths and led by Tim Gowers, has actually been at the forefront of the Academic Spring movement despite generally less interest in progressive politics and social justice-driven research. If STM researchers forsook their commercial journals en masse, we would have real progress because that is where the commercials make most of their money. So far they seem to be a rather conservative lot.

The fuel for this battle is the exorbitant price publishers charge for access to a single journal article. If single online articles were priced at something closer to the cost of downloading most single musical works–or of going down to a library and copying them–say $.99 to $5.00, fewer scholars would feel compelled to look for workarounds. Price-gouging inevitably produces smuggling. The marginal cost to publishers of each additional download is close to zero, as is the price-per-article that they receive from Big Deal libraries. Pirates who don’t happen to have access through a subscribing library feel their behavior is justifiable when the only honest alternative is to pay an utterly unreasonable price. I understand that publishers have a fixed overhead, but they might at least experiment with a more affordable pricing model since subscriptions often help defray their overhead and they would avoid legal costs and mounting ill-will from their existing and potential authors. They would also regain more control over their product and gather more information about their audiences. And if publishers also paid their authors even a small amount per article viewed on the publisher’s own site, authors would have an incentive not to share their work through other means.

“Though CRS initially indicated that it would be sending 100,000 takedown notices “imminently,” this intention was thwarted by ResearchGate when it removed from public view some of the content that would have been targeted.”

Seems like RG has some effective defensive tactics. Flip the switch on their users’ uploaded articles to “Private,” and thus avoid a takedown notice. That way they seem to evade the TDN, and may only incur minimal irritation from their users. It seems to be a simple matter for users to just flip the switch on their uploaded articles back to public. It only took me <5 minutes to flip mine back once I discovered they can been privatized. Curious, the ones that RG had reached down and set to private seemed haphazard, and maybe affected 1/3 of my uploaded full texts. Even curiouser, I doubt any of my shared versions are illegitimate; they are either my own versions, not within embargo, or copyright free VORs. No rhyme or reason, and no communication or caution about respecting copyright. So if my anecdata are shared by others (n=1, with ~20 pseudoreplicates), RG has a pretty solid dodgeball strategy going. When trouble brews, run its software squeegee across the site to disrupt the threat. Some users might not notice (until re-spammed that many of their pubs are missing full text.), leaving texts private, which would help RG's case. Other users would experience only a minor inconvenience.

I did not care to flip back mine. I doubt they are illegitimate, but why should I spend time checking? What matters is that I have my publications in one place sorted by dates and accessible from any point, isn’t it? Sharing was not a major incentive to upload.

Leave a Comment