My post today comes to you as many others in the past. It all started with a Twitter conversation that I dropped in on. The premise of the conversation was that all journals are harmed by piracy. The counter challenge was that open access (OA) journals are immune.
Even though Sci-Hub is billed as providing access to paywalled content, there appear to be thousands of open access articles in the host database. Sci-Hub provided usage of their services from 2015 to Science news writer John Bohannon with the full data set. Reviewing just the data from December 2015, I found that over 200 users accessed PLOS ONE content, over 450 users accessed Hindawi content, and a whopping 2,145 users accessed BioMed Central content.
So yes, even when people have free access to content under OA models, they are still using Sci-Hub to access OA content. I really can’t explain that one but if you have a good theory, share it in the comments.
Today I am offering some thoughts on how Sci-Hub harms OA journals. Many of these reasons are the same for subscription journals and the amount of harm may be more, less, or the same depending on the business model of the journals affected.
Survey after survey shows that when authors are asked why they choose to pay to publish in OA journals, the two most common answers are increased visibility/citations and funder requirements. Funder mandates are obvious but there is actually a lot going on with the visibility/citations answer.
Considering that an author pays model is a transactional experience, it is normal for an author to want to quantify whether they got what they paid for. My editorial office gets calls all the time from authors wanting to know how many times their paper has been downloaded. I think it’s safe to assume that download counts are being included in tenure and promotion packages.
One way to quantify success is to use article level metrics, which will include downloads. In fact, if memory serves, PLOS was the leader in developing article level metrics to show authors exactly what they were paying for. If you don’t think visibility is important to paying authors, you likely weren’t on Twitter much last week when posts such as this were popping up all over the place:
When papers are downloaded from Sci-Hub and the associated LibGen database, the publisher site loses the download counts. Now, the same can be said for all the papers in ResearchGate, Academia.edu, and institutional repositories. That said, initiatives like the STM Sharing Principles and CHORUS, are leveraging the existing infrastructure to connect download histories from various platforms. Sci-Hub usage would never be included giving paying authors and their universities that subsidize the OA activity a less than realistic way to quantify usage.
Every Journal Wants “Sticky” Pages
No matter what the business model is, all journals want to keep users on their pages as long as possible. Some OA journals sell advertising to supplement income. Others may try to “upsell” services and would use notifications on the site to tell users about those services. All of this is lost to those that use Sci-Hub to access OA content.
Big OA publishers such as PLOS and BioMed Central sell advertising on their sites and depressed usage due to piracy will hinder those efforts as well. Advertising income could certainly be used to help keep article processing charges lower or support waivers for APCs.
Do You Really Want Flat PDFs?
If I had a dime for every time someone lashed out at publishers for posting PDFs, I’d retire. And yet, that’s all you get from Sci-Hub and as John Bohannon’s piece for Science declared…everyone is using it.
Many of the larger OA journals have really reimagined the user experience. Many journal platforms are now modernized with a version of the eLife Lens. In fact, eLife recently announced a new redesign which boasts enhancement of figure display and the ability to cache the HTML locally. Notably, eLife has also created a “Magazine” to highlight all the editorial content that makes the pages “sticky” (see above section). It make sense that if you get to start from scratch, sites will be fresh and innovative. Of course, none of that development is necessary if users prefer to get PDFs from Sci-Hub.
If we can learn anything from Sci-Hub it’s that in exchange for easier access, users would be perfectly fine with peer reviewed, lightly formatted, author provided camera-ready copy in the form of a PDF. This would be a nightmare for the text and data mining folks but it’s an option for really inexpensive journals.
Free Is Not the Same as OA
The Internet is pretty evenly split with articles hailing Sci-Hub as the heroine of OA and articles reminding everyone that piracy is not the same as OA. But for the masses of people using Sci-Hub, do they care?
A year ago, I wrote about harm being done to the OA movement when some of the loudest advocates started shilling for Sci-Hub. For researchers still skeptical of OA (and there are still a lot of those), aligning the OA movement with piracy is a big mistake.
There are many reasons Sci-Hub is not OA, not the least of which includes:
- Reuse licenses are maintained from the original publication (most of it under copyright and not Creative Commons)
- Sci-Hub leaches off library subscriptions so it’s all paid for anyway
- Sci-Hub is not sustainable. Stealing stuff and giving it away to others is not a sustainable business plan.
- Due to various legal actions, Sci-Hub jumps from domain to domain and search functionality is often blocked.
Promoting a “No One Pays” Economy Is Very Bad for OA Journals
This may be the most damaging factor. Open access journals have been very careful about not calling things free because there are costs to publishing and nothing is free. This all came to a head a few years ago when PLOS ONE increased their APC and the collective world went nuts. Digging out tax documents showed that they pay vendors (just like every other publisher) and some of their supporters were aghast. The angst was misguided by the belief that publishing is cheap. Again, not a week goes by that someone tweets out that they could easily produce journal papers for pennies. Unless they are simply posting author created PDFs (see above), that’s not going to happen.
Most of these are issues for subscription journals as well but it is naïve to think that OA journals are immune to the damage being done.
What is “free” to many authors and users of scholarly content is the traditional model by which the access is purchased by their universities and publishing with a journal comes at zero cost for the author. Further, the library budgets aren’t exactly getting squeezed so the university can sink more money into research centers. On the contrary, these funds go to administrative overhead, new high tech student centers, and football stadiums. This does not engender sympathy pangs from academics in the actual “free-to-them” market. Even overlay journals built on arXiv are being supported with library funds.
OA was supposed to decrease the money, but 15 years in we haven’t seen that happen at all. The money is the same if not more, it’s just being paid from a slightly different bucket.
Versioning and Editorially Related Content Links Are Lost
Removing the published PDF from the journal page that provides links to related content is a problem for all journals. If Sci-Hub is the only source for a user, they will have no way of knowing if a paper has been corrected or retracted. They will not see any reviewer comments or associated editorials that may exist. The same is true for self-archived versions in repositories, which I remain concerned about in the long run.
Undermines the Need for OA Infrastructure
Why bother with the hassle and expense of OA when you can get it all free on Sci-Hub? This does not lead to the further development of sustainable open access models.
That’s my list. As stated in the beginning, most of these are issues for subscription journals as well but it is naïve to think that OA journals are immune to the damage being done. There is only one publishing industry with all different kinds of business models in it. Piracy is a threat to all of them.
56 Thoughts on "Are Open Access Journals Immune from Piracy?"
Great post, and many important points. Your observation that many OA journals are losing traffic to Sci-Hub despite their content being free reminded me of what I saw years ago with PubMed Central taking traffic away from OA journals, as well. (https://scholarlykitchen.sspnet.org/2012/11/12/what-pubmed-centrals-drag-on-publisher-traffic-could-mean-financially/) In that analysis, it looked like PMC was taking 22% of PLOS’ traffic.
PLOS includes PMC usage in their article-level metrics (ALMs), which is good, but ultimately PLOS and other OA publishers are being measured increasingly by the traffic going through their sites — funders, institutions, and authors are using these measures as main touchstones of value.
Aggregation itself is a convenience, so I’m sure why OA articles are being swept up in PMC and Sci-Hub usage is an aggregation effect. It’s also interesting to note that new commercial aggregations of OA content exist now, proving there is benefit and value to a “one-stop shop” — even if some of the items there could be had for free elsewhere, too.
The title of this article is misleading. The ‘loss’ suffered when an article in PDF is distributed away from the journal that this post describes has nothing to do with piracy. By definition, open access articles are openly licensed and unless the byline and citation information is removed, Sci-Hub or anyone else is legally allowed to reshare them. It is true that once PDFs are distributed elsewhere the journal loses the download counts, but this is also true when the author or anyone else shares or self archives a PDF beyond the journal where it was originally published. The benefits of open access is that articles (and this includes the PDF versions) can be legally distributed by anyone anywhere as long as the license is respected. Even if an article is licensed with the restrictive CC BY NC ND Sci-Hub or anyone else is entitled to distribute for free if the PDF is distributed in its original form as published by the journal. So yes, open access articles and journals are immune from ‘piracy’, because open access makes the need for piracy unnecessary by default.
Yes, this is a valid point. Content under liberal re-use licenses cannot be pirated. Sci-Hub and any other site or person are perfectly within their rights to redistribute this content. So really, some of these points have more to do with content that wanders from the source journal. Again, this may not be of a concern depending on the business model employed but I think the points above are still valid.
I’m going to argue that the headline is accurate. Yes, given the liberal reuse of OA article licenses, they generally aren’t being pirated here (although one could argue that Sci-Hub is financially benefiting from the availability of the articles, hence breaking the NC part of some licenses). But piracy is clearly happening for other papers under other licenses, and that piracy is having the impacts described above on OA papers. Take away piracy and Sci-Hub is no longer an attractive site for users to use. If there were no non-OA papers on Sci-Hub, would it be used any more than any other repository? If the answer is no, then piracy is still playing some role here.
How does Sci-Hub benefit financially?
Relatedly, the word “shilling” is used incorrectly in the article, to defame (I hasten to add, probably not meeting the legal definition of defamation) open access advocates.
According to this article, they are collecting significant amounts of money via bitcoin in return for their services:
Merriam Webster defines the verb “shill” as follows:
: to act as a spokesperson or promoter
> So yes, even when people have free access to content under OA models, they are still using Sci-Hub to access OA content. I really can’t explain that one but if you have a good theory, share it in the comments.
I think this is explained with a simple model:
Assuming I have a DOI or paper title and want the original paper, and am willing to get it from sci-hub.
I can go to the publisher and find it, then download/view it on the publisher page or find that it’s not OA, followed by going to sci-hub if it’s not OA or the publisher has paywalled what should be an OA paper.
Or, I can just go to sci-hub and get access without worrying about the status.
Quite simply, it’s easier to go to one place for everything than use it only as a backup. Due to it not having to worry about view counts/etc. it can be a simpler service, and the *users* don’t care about view tracking features. The customers do, but the users don’t.
I am glad you mentioned that Ian. I have been struggling with this. How do you procure your lists of DOIs for papers you want? I am trying to figure out the “Sci-Hub is easier” argument when it seems like you would have to do a search, go to the abstract page, record the DOI, then go to Sci-Hub and enter the DOI in the search box. Wouldn’t something like the Unpaywall button be a heck of a lot easier?
This is something I don’t get either — if you have to go to the abstract page anyway to get the DOI, that step is required, so you’re not saving any time by then going to Sci-Hub to get the paper that you have access to through the abstract page.
I suspect the time savings is in not having to sign in to your library or institutional VPN. You just go to the abstract page, then to Sci-Hub without signing in, which for many libraries can be an onerous process (see https://scholarlykitchen.sspnet.org/2015/11/13/dismantling-the-stumbling-blocks-that-impede-researcher-access-to-e-resources/ and https://scholarlykitchen.sspnet.org/2016/05/04/accessing-publisher-resources-via-a-mobile-device-a-users-journey/ as examples).
There are two scenarios I can imagine where a user may be in the position Ian described, where the user has the DOI, but chooses to use Sci-Hub.
1. The DOI is listed in a works cited list (or listed somewhere else, maybe social media or a class rubric).
2. The article abstract was found in a library’s database or Google Scholar, but the user’s library doesn’t subscribe to the journal, and the user is too ornery to use interlibrary loan.
In regard to the ease of Unpaywall button, this seems to go back to Ian’s original described scenario, where trying a more legit method may/may not result in success 100% of the time.
Good point about reference lists — where journals are cited by DOI, this would be a source where a lot of people would find papers they want and not an additional step. Thanks!
One doesn’t need a DOI to use Sci-Hub. You can search by full or partial title, URL, or DOI. There’s also a browser extension that supports fulltext search across all metadata and and abstract. In short, discovery using Sci-Hub is much easier than via almost any other mechanism – combined with the point above about near-guaranteed success in gaining access, this explains the popularity of the service I think. Even if one could only use DOIs, google scholar followed by copying the DOI to Sci-Hub is still more efficient than any other system that I know of.
It certainly would be more “efficient” for us all to be able to download all of our music from one site like Napster at no cost. But courts rightly found Napster to be an illegal operation. So should we be supporting the scholarly equivalent of Napster? Where does morality and legality enter the equation? What kind of example for students does faculty use of Sci-Hub and Lib-Gen set? Why should faculty be worried about plagiarism by their students when they are themselves supporting an illegal/immoral practice?
Sandy, I was merely commenting on the question above about why people use sci-hub over other means from a user experience perspective.
But since you raised it, for many people using sci-hub is also about the moral issue of subverting massive-scale rent seeking on what they see as a public good – the scholarly literature. Legality rarely factors into it – law, especially copyright law, doesn’t represent most people’s moral perspective on the subject.
Sure, and many Napster users tried to justify their behavior by citing the greed of the music industry. That works for just about any commercial business, doesn’t it? One might expect faculty to have greater respect for the law than your average Napster user, but I guess not. Do faculty cheat on paying tolls because they think roads should be a public good?
Sandy, I’m not sure why faculty are important here – students will find Sci-Hub on their own and are capable of their own reasoning. I’ve certainly never absorbed the moral principles of any of my supervisors, having already formed my own long before meeting then.
In any case, whether or not you agree with them, many people think digitally copying music and/or papers they would never have paid for is morally acceptable, and many people think the music industry was forced to improve by massive-scale piracy.
// What kind of example for students does faculty use of Sci-Hub and Lib-Gen set?
It tell students that their faculty *probably* don’t support the current copyright regime as applied to scholarly communications – a position that many people have no problem endorsing. Some students might also take away an aversion to all intellectual property, which while this is still a minority and somewhat extremist position, is intellectually defensible as Boldrin & Levine (2008) ably demonstrated.
Great post, Angela, thank you! Agree with the above that “ease” of SciHub can be perplexing, but I think it points out two messy facts: One is that no user is the same and readers bring myriad expectations and priorities into the info landscape. Two is a matter of reliability, where library login / publisher authentication experiences are rarely the same from time to time, link to link. The posts David cites above from Roger Schonfeld demonstrate the brittle nature of these access controls and the slippery nature of authenticated user experiences compounds with users over time and erodes their trust in those brands. Finally, I think it’s important to understand the power of the PDF: It empowers a user to overcome the brittle, decentralized nature of the academic info experience by creating their own personal digital library. Even with sub-standard PDFs, I can at least exercise some control over my information domain and that is critical for many scholars.
Yes, I think the difficulties around signing in do contribute. In my 10 years with access to an institutional login, I almost never visited campus. Access to library resources was often broken (if not blocked) by employer firewalls and browser issues. For the last few years, I searched almost exclusively using Google Scholar, and only accessed the university library when absolutely necessary, and usually from home. I can imagine if Sci-Hub was my go-to resource, I’d use it as a default and only access by other means if and when something broke.
I’m no longer doing research, so typically I don’t actually need to find papers, but DOIs are being found more and more in the initial publications.
For example, if you were reading this open access paper: https://www.nature.com/articles/ncomms14972#references
Then most of the references come along with a DOI straight away. There are even browser extensions & plugins that will replace all DOIs on a page with a direct link to sci-hub. It could be that a lot of people going straight to sci-hub for OA articles have simply set that as something to happen automatically.
>There are even browser extensions & plugins that will replace all DOIs on a page with a direct link to sci-hub.
That would account for it, scary!
Decentralized access to the literature does create a lot of ancillary problems to the integrity and trustworthiness of the content, not to mention the metrics derived from decentralized services.
Crossmark (https://www.crossref.org/services/crossmark/) does help to address some of the currency and integrity problems created when a paper changes (corrected, retracted, etc.) after publication.
Still, I haven’t seen any universally viable ways to track usage of a decentralized document. Today, I’m less confident of the value of download data than I was 10 years ago. If top-downloaded certificates become important to P&T committees, we should anticipate the rise of bot-run downloading services, similar to services where you can purchase Twitter followers, Facebook likes, or favorable Yelp and Tripadvisor reviews.
I agree that aggregating usage is both useful and challenging. For example, usage information (article level) is available from some repositories. That could be aggregated with hosting-platform usage information. But what does it mean to aggregate non-COUNTER-compliant usage information with COUNTER-compliant usage information? If we do such aggregations, they should at least be asterisked (like some sports stats).
Hi Angela –
Thank you very much indeed for raising awareness of some of the implications and impact distribution of content can have, no matter how it is licensed. Your point about the importance of being able to aggregate usage information for authors is an important one, and an element of the STM scholarly sharing principles. Publishers (and other stakeholders such as funders and libraries who pay for APCs) should be able to measure the amount and type of sharing, using standards such as COUNTER, to better understand the habits of their readers and quantify the value of the services they provide.
And in passing… is it in fact correct that OA content licensed under CC-BY-NC-ND can be distributed by anyone and everyone? That’s not my understanding of how the NC part of the license works.
With best wishes,
Hi, Alicia —
I believe it is correct to say that OA content licensed under CC-BY-NC-ND can be distributed by anyone and everyone. (Whether or not such content is really “OA content” is a separate question, but let’s say for the sake of argument that it is.) The CC-BY-NC-ND license does not restrict distribution in any way, nor does it say who may or may not redistribute. It only restricts commercial use and the creation of derivative versions.
This may be a bit of semantic nitpicking, but I’m not allowed to redistribute NC content if my channel for redistribution requires a commercial transaction (selling access or reprinting the material in a book that’s sold for example). So I suppose it’s right to say that “anyone” can redistribute the material, provided that they follow the rules set by the licensing restrictions. And also that those restrictions aren’t completely forbidden, just that they require the permission of the copyright holder.
To David: the license restricts the _manner_ in which you may redistribute; it doesn’t deny anyone the right to redistribute. So if you run a commercial journal you may not be able to publish the content there (without permission); but the fact that you are a person who runs a commercial journal doesn’t mean you can’t distribute the content freely on, say, a listserv or blog. Anyone and everyone is free to redistribute CC-BY-NC-ND content, as long as they do so in accordance with the license terms.
Hi Rick –
I checked with legal colleagues, and their advice is that the CC BY-NC-ND license does restrict distribution for certain purposes. Its focus is really on the purpose of the use, rather than the nature of the user. So, if a non-profit organisation distributed the relevant content for commercial purposes (for example, sold copies for a fee or incorporated advertising with the content), then that could infringe the license. Conversely of course, if a commercial organisation distributed it for truly non-commercial purposes, then that could be permitted under the license.
With best wishes,
Hi, Alicia —
You’re absolutely correct: a CC-BY-NC-ND license does indeed restrict certain types of redistribution. But your question was whether it is “in fact correct that OA content licensed under CC-BY-NC-ND can be distributed by anyone and everyone,” and the answer to that question is yes. Anyone and everyone can distribute content under that license — but they can’t do it in absolutely any way they want. The restrictions that license imposes are restrictions on how redistribution (and other kinds of reuse) may be done, not on who may do it.
One possibility is that Sci-Hub is intelligent enough to maintain a list of OA publishers, OA journals, or OA articles, and simply redirect the DOI query to the article on the journal site.
A simple experiment (n=7) has shown that is sometimes the case — and sometimes not! I won’t speculate on the logic, since several alternative models are possible.
Thus in some cases, the OA journal may retain usage and other information. Not sure what that is worth if it is inconsistent, though.
We tend to use “workflow” in the singular, when it really should be plural**: workflows. When we want to provide better user experiences than Sci-Hub, we need to help the user workflowS. There are multiple points to help users before they feel the need to turn to Sci-Hub:
– in literature search
– on an abstract page
– on a paywall page
In a recent talk at CSE I commented that the scholarly communications enterprise is working on multiple, complementary solutions, one for each of the above points. These would be CASA, Unpaywall, and RA21 (respectively). Fortunately, they all work just fine together!
**This reminds me of a comment a well-seasoned Stanford administrator told me when I was a young staffer at Stanford: “always remember that the word ‘faculty’ is plural — they don’t have a goal of speaking with one voice, even it that would be convenient for you”.
Just curious as to why those who are hurt by Sci-Hub do not launch a denial-of-service attack on the site. Can it be illegal to use such a method to shut down a site that is itself illegal?
You are wrongly presupposing that publishing should be done by external companies – but we don’t buy research output from external companies either. If research organizations do the publishing themselves (as they used to do, e.g. proccedings of academies), and external companies just provide the infrastructure, then we can choose the cheapest bidder, and all problems of piracy disappear (see http://bjoern.brembs.net/2016/12/should-public-institutions-not-be-choosing-the-lowest-responsible-bidder/, as well as my own post: https://www.frank-m-richter.de/freescienceblog/2015/10/28/how-to-switch-quickly-to-diamond-open-access-the-best-journals-are-free-for-authors-and-readers/
Can anyone provide evidence of an OA publisher complaining in public about Sci-Hub’s activities?
This is a great opportunity to tech students about copyright obligation and their need to identify sources of information they access and provide to others. I find it amazing that so many students do not see the need to cite sources and see copy and paste as such an acceptable practice. Not only does this prevent the original author from receiving credit for his or her work and thoughts, but it also avoids allowing for the original author to be accountable for what they wrote to begin with.
Student do need to understand that sharing information requires accountability and responsibility. In order to be a responsible distributor of information, you must also be responsible in citations and credit and accountability.
Many authors know the excitement of finding the perfectly worded passage that provides an excellent explanation to the issue at hand. I have found that providing a citation and placing the verbiage into my own vernacular is challenging but also useful to allowing to understand the writer’s train of thought while developing and expanding my own.
Agreed, Richard! This would be a fine starting point to instruct students on the different, but similar, issues pertaining to copyright infringement and plagiarism.
Thank you AJ for the affirmation.
// So yes, even when people have free access to content under OA models, they are still using Sci-Hub to access OA content. I really can’t explain that one but if you have a good theory, share it in the comments.
The only real data on this comes from the Travis (2016) survey in Science. I cross-tabbed it here: http://eprints.rclis.org/30147/ (see slides 25-38). While most use is because users perceive a lack of access, a non-trivial percentage is people who use it purely for convenience. They know they have access but prefer to use Sci-Hub (slide 38). Others in this thread have pointed out how much easier it is than legal access.
Regarding Angela’s and David’s remarks on the potential Sci-Hub user workflow, let me remind you of the Sci-Hub bookmarklet. [link redacted by editor] The workflow goes like this:
1) find any article
2) click a button (the bookmark)
Yes, it really is that simple. No one even needs to visit the actual Sci-Hub homepage anymore; not even cutting and pasting is required if they have a link to follow. After trying the Sci-Hub bookmark, why would anyone use their library?
Again, this requires navigating to the article page in the journal. If the article is OA, you can download it from there rather than Sci-Hub (which if I recall correctly, requires an extra step of a Captcha). Further, if your library uses a good authentication system (or moves to a better one in the near future), then you have access to subscription article PDFs in that same one click, which seems at least as easy.
We showed here http://eprints.rclis.org/30981/ that a fair portion of all USA Sci-Hub usage was automated. This is almost certainly not the case anymore since they have now instituted the Captcha. It would be interesting to see if the open access use patterns are the same post-Captcha. My expectation is that they’d decline, i.e. that the OA usage was driven by the bots. We could use the existing data to test for that but I haven’t.
Still not a single piece of evidence provided by OA journal publishers that their OA journals are damaged by Sci-Hub and its like. We have discussed the efficiency of Sci-Hub’s search. This is interesting and provides a possible explanation for its popularity, but we still have nothing to support Angela’s original contention.
As the cliche goes, absence of evidence is not evidence of absence. You’re assuming that OA publishers read this blog and read the comments, and follow the comments for days after the original post appears. Further, it’s not clear to me how one would present clear evidence of “harm”, other than the established data showing traffic to OA articles on Sci-Hub, clearly traffic that could have been used to drive impressions for ad sales on OA journal sites, providing a further revenue stream (and either increased profits or a means to lower APC prices). That’s about as clear as it gets. What evidence would you expect in terms of authors seeing lower download counts?
Further, OA advocates have regularly spoken out against Sci-Hub as many see it as harmful to the creation of a truly OA landscape (one that addresses reuse rights rather than just availability), and that the presence of the unsustainable Sci-Hub slows progress toward OA because if illegal access is readily available, the drive toward legal access has less impetus. Examples include the Winnower and Peter Suber:
Sci-Hub is not “pirating” OA journals as their content is free to re-use. OA journal editors are not complaining about Sci-Hub. I strongly object to what Sci-Hub gets up to in relation to subscription journals, but all claims that its piracy is damaging OA journals should be ignored.
Angela attempted to clarify in the comment above:
But let’s be clear, Sci-Hub is profiting from the use of articles, so they are, at the very least, breaking the terms of any OA articles with an NC license (not sure if one would call that “piracy”).
Further, it seems a logical argument to state that OA journals are affected by piracy, because piracy is at the root of what Sci-Hub is doing, and the source of the vast majority of their traffic. If some OA journals are harmed by Sci-Hub’s presence and prominence as a source for articles (as noted above the loss in advertising impressions as but one example), and Sci-Hub is based on piracy, then it would follow that piracy is having an effect on OA journals, whether their own articles are being pirated or not.
OA journal editors are not complaining about Sci-Hub.
I can think of thousands of subscription journal editors that have not publicly complained about Sci-Hub as well. Does this mean that their business is unaffected?
David, Sci-hub are not profiting from the use of articles. They accept bitcoin donations to support the infrastructure, but do not take money for themselves (even to cover their time).
Because they are not a commercial enterprise, and (even if they were) there is no required transaction to obtain an article, they are compliant with CC-BY-NC.
There has been no evidence of harm to OA journals presented, and in the absence of even a coherent reason for potential harm there is no reason to assume OA journals are harmed.
Further, there is no evidence that non-OA publishers are harmed. They might claim harm, but unless Sci-Hub is depriving them of revenue they would otherwise have received there is no direct monetary harm. Perhaps sci-hub distorts download metrics, but they also provide download count data, so the metrics can be corrected if publishers wish.
David, Sci-hub are not profiting from the use of articles. They accept bitcoin donations to support the infrastructure, but do not take money for themselves (even to cover their time).
Really? How do you know? Can you provide financial reporting that shows this is the case? What do the people behind Sci-Hub do for a living? Who are these people and how do you know this?
Can you explain how “accepting bitcoin donations to support infrastructure” is not a commercial transaction? If I take something that belongs to you and give it to someone else and then “accept a donation” from that person for my infrastructure, are the two actions wholly unconnected?
Because they are not a commercial enterprise, and (even if they were) there is no required transaction to obtain an article, they are compliant with CC-BY-NC.
Does being a commercial enterprise factor into the NC license? Can a not-for-profit sell NC articles if they use the resulting funds for “infrastructure”?
What is the definition of “commercial reuse”? Can you point to the legal case where that was established? As far as I know, it’s not so clear. If I sell advertising on my website where I’m giving away your article, is that a commercial activity for example?
There has been no evidence of harm to OA journals presented
Simplest case: there are OA publishers that sell advertisements on their journal sites. These ads are sold by impressions. Any traffic that is diverted away from the journal reduced impressions, thus reducing ad revenues. Those are important revenues because they do not come from the research community and can allow a journal to reduce their APC price levels. Traffic can be monetized, so losing traffic to someone else reduces one’s abilities to do so.
Further, there is no evidence that non-OA publishers are harmed. They might claim harm, but unless Sci-Hub is depriving them of revenue they would otherwise have received there is no direct monetary harm.
If there is significant activity on Sci-Hub from areas or IP addresses where a journal has no subscribers, then that journal has been deprived of revenues for that transaction. As has been noted in many places, there are areas where universities are canceling their subscription packages with publishers and researchers have publicly stated that they are doing fine because they have access to Sci-Hub. That would seem to me pretty direct evidence of financial harm.
David, you seem to have mistaken me for google. You’ve asked a lot of questions that can be answered by googling, or to which you clearly know the answers already, having demonstrated the knowledge in other comments. I’m not interested in using my time to save you time, so I suggest you google the questions yourself. To briefly address the key issues so the conversation is not derailed:
1. Sci-Hub’s use of donations is discussed on the Sci-Hub forums (you can use google translate to read them). You can any blockhain forensics tool to inspect the bitcoin transaction records. Details of the operators and what they do is public.
2. Non-commercial use is defined in the latest NC licenses as use “not primarily intended for or directed towards commercial advantage or monetary compensation.” (previous versions differed slightly). There are exceptions, as in at least one jurisdiction (Germany) the license cannot override the legal meaning of ‘non-commercial’ (see test case: https://www.heise.de/newsticker/meldung/Verstoss-gegen-CC-Lizenz-Deutschlandradio-muss-zahlen-2151308.html). However, Sci-Hub is not under German jurisdiction.
3. If you distribute copies of CC-BY-NC licensed materials free, but also have a way for people who want to support you in doing so to donate, then the actions of distribution and accepting donations are not wholly unrelated but they are also not transactional or commercial. If anyone can donate money regardless of whether they accept items from you, and anyone can accept items regardless of whether they donate, then it would be very tenuous to argue violation of the NC clause of the license. Feel free to employ the services of a lawyer if you want other hypothetical scenarios explained.
4. Any organisation making money from ad impressions on pages that distribute CC-BY(-etc.) articles might make less money if they get fewer ad impressions as a result of anyone exercising the rights afforded them by open licenses to redistribute the article. However, it is the publisher who then has the responsibility to provide a more attractive platform than any alternatives if they want to make money this way – distributors are not at fault. If they want to make money from ads alone, they should choose a different license. People exercising their redistribution rights are not harming the publisher, the publisher is harming themselves. I would submit that many OA papers are downloaded by directly clicking on the PDF download link of a result in Google Scholar, which will also not involve an ad impression on the publisher’s site. The core problem in the hypothetical ad revenue example is that ad revenue is a poor choice of revenue stream for content that is intended to be freely redistributed.
5. To demonstrate harm to a commercial publisher by Sci-Hub you have to show that contracts with that publisher were cancelled *because* Sci-Hub provides the service the users need. You have only stated that some contracts have been cancelled, and that some users have said they don’t mind because Sci-Hub exists. In fact, there are many excellent reasons not to do business with the major commercial publishers (I refer you again to google), and these motivate all the movements to avoid renewing contracts that I have spoken to people about. Publishers’ bad behaviour harms them leading to fewer subscriptions, and at the same time Sci-Hub independently provides a massive amount of value to people who would never have paid publishers, in addition to those who have legitimate access but find it a better service.
I appreciate your confidence in my knowledge base, but really, my use of questions is either 1) out of my own ignorance and 2) in some cases a rhetorical tool of argumentative style. And yes, I could spend days translating forums but frankly, I have better things to do with my time.
1) You claim details of the operators of Sci-Hub are public. As far as I know, the only person publicly identified with Sci-Hub is Elbakyan. If there are others publicly known, they would also be targets of lawsuits from Elsevier, would they not? If you know of any other individuals behind the site, I’m sure Elsevier would appreciate that knowledge. Further, you have no idea what is done with the money after it is received, do you? What proof is offered as to how it is spent? As far as I know, Bitcoin forensic records do not tell me if someone is using the funds to pay their rent or buy food but please correct me if I’m wrong and detailed financial records of Bitcoin’s financial operations are publicly available. Again, I’m sure Elsevier would appreciate having them.
2) “not intended for monetary compensation”, and that is exactly what is happening here, so clearly a violation of the NC license is occurring.
3) So it’s just an opinion on your part, not one with legal expertise, that it might be a difficult argument. I’ll argue from my same lack of legal expertise that I could make a reasonable case that since the two activities are directly linked, monetary compensation is happening.
4) “Any organisation making money from ad impressions on pages that distribute CC-BY(-etc.) articles might make less money if they get fewer ad impressions as a result of anyone exercising the rights afforded them by open licenses to redistribute the article.” Thank you for agreeing with me and the ideas expressed in this blog post. You may not like this particular business model or think it’s a very good one, but clear financial harm is being done.
5) Here you’re asking me to read the minds of the negotiators, which is impossible. You’re offering other hypotheticals, just as impossible to prove. I can refer you to statements like the one below, where Sci-Hub has clearly allowed universities more leverage in their negotiations with big publishers (you may consider this a good thing, but it is clearly financially harming those publishers):
In Peru, researchers are also set to lose online access to Elsevier’s Science Direct and Scopus platforms from 2017 because of a lack of government funding. But some scientists there say that it’s not a problem, because they can get the papers they need illegally from the Sci-Hub website. “I’m not worried. Downloading papers is rather easy now with Sci-Hub,” says one plant biologist who doesn’t want to be named.
If you want the information I’ve told you where it is. If you can’t be bothered to access it, you’ll continue to not know and that’s fine by me.
On points 2 and 3, again, I’m happy for you to continue to be wrong on the basis that you’re not interested in finding the information. I have taken legal advice on the NC issue, but I am not a lawyer so please do not interpret my lay explanation as legal advice. And no, I did not agree that Sci-Hub is harming publishers – publishers are harming themselves.
Whether Elsevier have found this information is not of interest to me. If they want it, and are competent, they can find it. Given they’ve been pursuing a completely pointless lawsuit to get an unenforceable injunction in a jurisdiction irrelevant to the operation they are targeting, I am not confident in their competence. While I’m happy for them to waste money on expensive lawyers to promote Sci-Hub, I will not help them do it (or anything else).
Rather than trying to read the minds of negotiating teams, you could read the public statements they have made. Your example reiterates my point.
This seems unlikely to lead anywhere more productive, so I’m unsubscribing from the thread. I wish you all an enjoyable weekend.
David Crotty stated “I can think of thousands of subscription journal editors that have not publicly complained about Sci-Hub as well. Does this mean that their business is unaffected?”. Funny, I could have sworn Elsevier, representing thousands of non-OA journals, was suing Sci-Hub, with many other non-OA journal publishers/editors cheering Elsevier on in this legal action. To sum up: the original article was misleading and inaccurate, and no amount of posturing by the author’s supporters can change that.
Hi Charles, I’m not sure I have much of a response to the various strawmen you’ve set up here. Rather than argue against the actual points made in the post, you’ve decided that harm can only be happening if you’ve heard complaints from OA publishers (and that apparently, Elsevier represents all publishers). The post talks about loss in ad sales, and the groundbreaking efforts that OA publishers are making in experimenting with new technologies, not to mention the need for OA to be financially sustainable over the long term. Just as you don’t seem interested in addressing the actual content of the post, I’m not really interested in addressing the argument that if you haven’t heard any public complaints then Sci-Hub must not be having any impact on OA journals. As linked above, Peter Suber and others disagree with you.
“The post talks about loss in ad sales, and the groundbreaking efforts that OA publishers are making in experimenting with new technologies, not to mention the need for OA to be financially sustainable over the long term.”
There are many OA journals that do not have, nore rely on in any way, on ads income (not even from Google). Financial sustainability of truly open access journals does not rely on payment for access, nor on users viewing content on its ‘native’ platforms. As we all should know and should really stop having to discuss, Open Access means content can be legally distributed (under conditions of license) further, and this means that the sharing of OA content on Sci-Hub or elsewhere cannot be correctly labeled as “piracy”. The strawman here is that OA is affected by Sci-Hub or piracy, and in spite of tiresome extended exchanges, we still have to hear from non-profit, non-hybrid OA publishers stating they are affected by their content fulfillint its mission (i.e., being distributed well beyond their platforms in ways they may or may not have predicted). It’s hybrid publishers, who still have trouble implementing Open Access whilst relying on paywalls, that have expressed discontent with Sci-Hub. That these publishers continue engaged in discrediting Open Access by insisting on inadequate business models is to us researchers and editors upsetting to say the least.
No journal publishing isn’t free (though for an online only journal it can be cheap). However, it is paid for by universities who BENEFIT from greater distribution and would prefer if everyone could read research they produced. The key question is if universities ultimately pay for virtually all academic publication (indeed pay twice…once in salaries and again in subscriptions) why can’t they get what they would like: universal access?
The answer is the stranglehold of traditional publishers. Elsevier and others very cleverly use bundle pricing to ensure that the academic world can’t migrate step by step away from their platforms and to a model where the universities simply support journals directly with the money they would have spent on subscriptions (plus a substantial savings given profit margins of the scientific publishers).
Given these manipulative attempts to retain control over the scientific publishing world I think we are authorized, indeed required, to do what it takes to burn down traditional scientific publishers by destroying their ability to profit from non OA journals. As high impact journals become unprofitable they will either vanish (and the vacuum will summon new, free or OA journals to fill their place), convert to OA or be sold cheaply to someone who will run it as OA or free submission free access.