#icanhazpdf: Civil disobedience?

Some lively Twitter banter has arisen this evening regarding the practice of sharing PDFs of scientific articles when one does not have personal or institutional access.

Specifically, some among my stead have taken to tweeting requests for articles using the #icanhazpdf hashtag.

For non-open-access articles, does this practice violate a publisher’s copyright?


(And I welcome input from my ACS overlords.)


Update 24 December: I have changed the title of this post to reflect a comment below by Michael Eisen that sharing PDFs of journal articles is an act civil disobedience toward the scientific publishing enterprise. I had previously compared the practice to the Underground Railroad or Napster music file sharing. I deeply regret the use of the analogy of PDF file sharing to the Underground Railroad, a network of abolitionists who facilitated the safe escape of enslaved African-Americans in the southern US to freedom in the North and northward to Canada. I, in particular, should be especially sensitive to making such an ill-considered analogy of one of the most degrading episodes in US history to an intellectual discussion of sharing scientific papers. It was wrong, period. I apologize deeply to those offended by my thoughtless mistake.

Author: David Kroll

Share This Post On


  1. That would have to be tested in court homes.

    • Indeed, and that’s where my question comes from. *Who* is going to be made an example of by a scientific publisher?

      Currently, I view the #icanhazpdf movement as a group of science writers, students, and trainees who either have no access to institutional publications or whose institutions may not have access to the subdisciplinary journals whose contents are often required for this multidisciplinary world. For papers not available under open access, most journals charge between $25 and $55 for access to a single paper. If you’re at an institution, you may be fortunate to have interlibrary loan agreements to get the paper for less than $10 but in a timeframe of days.

      So, for example, I want to work on the pharmacology of a novel marine natural product but one of the seminal papers is in a oceanography journal. I contact a marine biology colleague for the paper from their institutional holdings and they send me a PDF. At first glance, that is perfectly legal. But I believe that only the author themselves has the right to make copies for colleagues (ACS journals permit the author to make 50 copies but not to post the PDF on their lab website).

      But it’s now easier to just throw out a link and a hashtag and have a large number of generous souls offer to send you the PDF. Does that global request by itself violate copyright? Does the person deciding to fulfill the request violate copyright?

    • It’s spell Holmes. As in Sherlock.

  2. It’s the wrong question – it seems to unambiguously violate copyright. The question is whether they should hold copyright in order to wield it in this obviously destructive way, and whether people should engage in civil disobedience to undermine this immoral system.

    • Michael, thank you so much for commenting – you’re a pioneer in the open-access community so I really appreciate you coming by. Apologies for your comment getting caught in moderation but you’re now automatically green-lighted.

      In reference to my response to DrugMonkey above, who is the copyright violator in the case of sharing with a colleague a PDF you haven’t authored? I assume that it’s the provider and not the requestor.

      The question is whether they should hold copyright in order to wield it in this obviously destructive way, and whether people should engage in civil disobedience to undermine this immoral system.

      I truly appreciate your sentiment. I couldn’t (and obviously didn’t) phrase it so well.

  3. Right now if my home university or our close partners don’t have access to a paper, I can make a request through the Interlibrary loan system, which is simply a larger network of university and governmental libraries. I’ll normally get the PDF (or more frequently a paper copy) in 1 to 2 weeks depending on the rarity of the paper/journal. I don’t see the difference between this system and getting it from a colleague via the #icanhazpdf system, other than the relative speed of transmission and the cutting of library personnel middlemen/women!

  4. I think it’s OK and expected to share articles with your actual colleagues or professional acquaintances. This facilitates discussion and the advancement of science. I wouldn’t feel so comfortable passing them out to people with whom my only contact has been the request for the paper.

  5. It does violate their copyright, but as other already said, that’s irrelevant.

    I do not earn any money with my publications. I publish them for all interested researchers to see, the more the better. Then, there’s that publisher that I’m basically forced to sign a Copyright contract with, because the institutional research world is so fucked up that I need a private corporation selling my publication so it can be recognized by my peers.

    Publishers are evil here, but honestly, it’s the whole research system that is evil. It’s my University who is obsessed by the name of the journal or conference I publish with. It’s my University and my peers who don’t recognize any work outside a given list of journals, even if that publication has been peer-reviewed by top-notch researchers of my field, but has not gone through the publishing process of a private corporation.

    #icanhazpdf should aim, not only to publishers, but to the whole scientific community.

  6. David makes an excellent point which some folks seem to be unaware of. ILL is not free.

  7. it seems to unambiguously violate copyright.

    Thank you. Exactly the point I was making to the Open Access Acolytes on the Twitts. Yes, it would need to be tested legally but on the face of it, this appears to be a violation of current publishing / journal access agreements.

    I don’t believe that most of those who so enthusiastically use #icanhazpdf on the Twitterscape really understand what they are doing. They should. And, as you suggest, make an informed decision as to whether they are going to engage in civil disobedience *and risk the penalties* on this.

  8. When I commented last night, I did so naively and off the cuff. Having read this discussion and looking into the terms associated with my university’s journal access policies, I can see I was pretty far off target. Clearly I was wrong to associate #Icanhazpdf as an alternative to ILL, and misunderstood how ILL works, and retract my above comment. Thanks for making me think more critically about this subject, and I agree with “drugmonkey”, users should be making informed decisions about their use of the hashtag!

  9. This may seem like a simple questions, but it is really a whole series of questions. Let me try to list a few:

    Is it illegal or is it (morally) wrong?

    Is there a legal or moral difference between a) asking for a copy from the corresponding author, vs. b) asking for copy from someone else?

    Is there a legal or moral difference between a) asking for a copy from a single person, vs. b) asking a group of people?

    Is there a legal or moral difference between a) asking for a copy “privately” from a mailing list with thousands of members of a scientific society, vs. b) asking for a copy “publicly” somewhere online (e.g., on Twitter, blog, G+, Facebook or the FriendFeed group that exists specifically for this)?

    Is there a legal or moral difference between a) asking on Twitter with the #icanhazpdf hashtag vs. b) asking on Twitter without using the hashtag?

    Are the laws written decades or centuries ago gradually becoming outdated, counterproductive or even evil?

    Are the copyright laws, publishing industry methods, etc., which were at the time of implementation improvements over the past methods, now a quaint but irritating vestige of the past way of thinking and need to be changed?

    Are the social mores, meanings of “right and wrong”, power relationships, etc., which were just OK 50 or 500 years ago not OK any more?

    Will the next generation, which is incapable of understanding the concept of copyright when asked (and thinks is is vile, authoritarian and just, like, so weird, when it is explained to them) change the laws, as well as the way things are done? And is asking for PDF online, a natural thing to do today, going to become a legal thing to do due to their efforts, then replaced by a world in which all the papers are freely available anyway?

    • I don’t see how anyone can argue that it is legal to download a paper from your university’s library and send it to anyone not associated with that university – whether they asked privately, or on twitter – unless of course the paper is in the 10% or so that were published under some flavor of open access and carry a license that specifically permits redistribution. It doesn’t matter whether it is accepted practice or not. It is not only a copyright violation, but also almost certainly a violation of the terms you implicitly consented to when using your institution’s online access agreement.

      Of course, the fact that it’s illegal does not mean that it’s right. I’m not going to rehash the whole argument for open access here, but it should be pretty clear to anyone that it is ridiculous that scientific knowledge can not be freely accessed and shared by anyone anywhere.

      This is why I think arguing about the morality of copyright laws isn’t really the right question. While copyright laws may mostly suck, they don’t need to be changed in order to allow scientific papers to be freely shared. There is absolutely no necessity for copyright restrictions to apply here. Authors of scientific papers make a choice to limit how their work can be accessed and used when they publish in journals that either demand copyright be transfered to them and wield it to restrict access or use licenses that accomplish the same thing.

      The whole point of using Creative Commons licenses is to allow creators to take advantage of the structure of copyright laws to grant potential users the right to do things like access, distribute and reuse their works.

  10. There’s a BIG difference between Interlibrary Loan and just asking a colleague for a copy. For one thing, there are specific library exemptions written into copyright law. For another, libraries follow CONTU guidelines (http://www.copyright.com/Services/copyrightoncampus/content/ill_contu.html). Libraries also follow the licenses that vary by publisher. For example, we can’t e-mail an article from Elsevier. We have to print and scan it at like fax quality or something equally stupid.
    So yeah, Michael Eisen is right, you can’t download an article and send it to someone at another institution or unaffiliated. Even if publishers often turn a blind eye to this in onesies and twosies, the hashtag might go a bit far. The requester isn’t in the wrong, it’s the provider *and* possibly his or her library.

  11. “Even if publishers often turn a blind eye to this in onesies and twosies, the hashtag might go a bit far.”

    But use of the hashtag is also a ‘onesie’ (one person sends one copy to one recipient)… why not turn the blind eye as well? The only difference is in speed and efficiency of ‘pairing’ the sender and recipient. The end result is the same (and WILL happen no matter what).

  12. Coturnix-

    Same arguments were made for music filesharing

  13. ….and now we subject ourselves to ads or pay what the market has determined are reasonable fees (subscriptions or per-file). What if papers were 99cents each? Or 35 cents each?

  14. “just” like? No but should this ever come to litigation do you really think the judge will disallow such comparisons to be drawn?

  15. Oh for heavens sake….
    No Virginia, “fair use” does not breach copyright (look it up), the question is whether this is fair use (& I think it probably is, same way photocopying & sharing with a colleague clearly is, but IANAL)
    No Virginia, publishing scientific research for money (which the publisher asks the reader to pay for) is not immoral, no matter who funded the research (unless it was the Nazis). It’s just a f***ing business model. Take a deep breath people
    Yes Virginia it is OK to use phrases like “underground railroad” or “slaving away in the lab” or “heads will roll” or a million others with referents of questionable taste… as long as one doesn’t take the analogy too literally (which some obviously do)
    And finally, no Virginia, “civil disobedience” is not much better
    But the good news is YES – there is an #icanhazpdf Santa Claus
    Happy Xmas all

  16. Rightfully or wrongly, I’ve always read ‘fair use’ to mean to avoid use that would cause the copyright holder a (potential) loss of income. Many of the journals ask such high fees for a single paper I’m left feeling that publishers on hand wish to restrict free copies to ‘fair use’ but on the other are not providing a *realistic* alternative, leaving at least some of their intended clients damned either way.

    For what it’s worth (not related to #icanhazpdf): I’m under the impression that one advocacy forum gathers research papers into an archive they have created, which they then provide to their members.

  17. Many interesting comments. Unfortunately, most scientific journal copyright statements make it clear that virtually all scientists are in violation at one time or another. Permission is granted to subscribers and members of subscriber institutions – including inter library loan members – to produce ONE copy for personal, non-commercial use (not clear what this means if the scientist is earning a salary by the business or institution where the copy has been made). Further, many journals add a clause indicating that the one copy must be destroyed after reasonable use – typically defined as following the experiment or draft of the paper for which the copy was used as a reference. Handing a copy to a friend, storing a copy indefinitely in a file cabinet, sending a copy over the internet are examples of copyright violation.

    The problem is that nobody can subscribe to all journals that might be needed for a research project. This is further exacerbated by the fact that employees of small institutions, start-up companies and individuals may not have access to libraries or inter library loan privileges as intended in copyright statements, leaving them with the choice of paying $25-50 for copyright clearance (for the article copy that must be discarded after reasonable use) or finding a friend/colleague/online source to ‘share’ the article. For most such researchers, the cost of legally gathering reference papers for even a small project is prohibitive…

    Surely, publishers understand that copyright infringement is happening every day, and would prefer that it stop. Historically, scientific journal copyright violation prosecutions have been initiated against individuals and institutions – albeit not too many in the news lately – suggesting a degree of realistic tolerance. It is the unfortunate fact that producing and disseminating high quality, peer-reviewed scientific literature has a cost, whether the publication is issued on paper, online or both. Publishers have to find a way to support those costs if such publications are to continue…and scientists have to find a way to gain access to the papers they need to progress in their work. #icanhazpdf may be one solution, but it comes with the risk that publishers, and courts, may not agree that it constitutes reasonable use as defined in copyright statements.

  18. Asking for a paper – by shouting from a mountain top, or by email, or on Twitter (with or without any hashtag) is legal.

    Sending a copy is legal.

    But, the moment the recipient OPENS that paper, it (due to a stupid court case from a few year back) legally creates a new “copy” (on the screen, yes). At that moment, the act of sending by the sender become illegal – a new copy of the paper was made.

    This is why people using #icanhazpdf are fine – nothing they do is illegal. But they should be careful about thanking the senders. There is a good reason most senders do not say anything publicly about it – they DM to get email address, then mail it, without saying anything on Twitter (and certainly not drawing even more attention to themselves by re-using the hashtag).

    Of course, one reason that publishers are not going after this is that the #icanhazpdf tweet contains a link to their site. More eyes on a paper, more likely it will get read, later cited, thus adding to the IF of the journal. If the recipient uses the paper in some way – to cite in their own paper, or to write an article/post about it – this again brings attention to the paper and potentially increases the IF of the journal. So publishers tend to go only after Big-Time copyright breakers, not after single-paper exchanges.

  19. For those that do use this. Put your email in the tweet and delete the tweet when you receive the document. I see several using the hashtag without email.

  20. Comparison with the Underground Railroad is not wrong at all. Access to research results saves lives. (We can tell that big publishers agree with this, otherwise there would be no Hinari.) It follows that LACK of access COSTS lives. So, yes, access to research is a moral issue, not just a business one.


  1. DIY Fernleihe – die dunkle Seite der Macht « Hatori Kibble - [...] einem amerikanischen Science-Blog gibt es dazu übrigens eine interessante Diskussion, ob das jetzt einfach nur eine simple Urheberrechtsverletzung oder …
  2. #icanhazpdf o el Open Access por la vía de los hechos | Cultura Libre - [...] Se dió un debate muy interesante sobre este tema en un blog de la American Chemical Society #icanhazpdf: Civil …
  3. Interactions: The Numbers Behind #ICanHazPDF | Altmetric.com - [...] but its practice is fiercely defended by many. As such, the hashtag has been the subject of many heated …