Occasionally there some posts on the main site contain links where it is very natural to suspect that the link provides access to some material without having the necessary legal rights. Should the posts with problematic links be flagged? Are MO mods and MO community supposed to be trying to watch for such links and handle those posts in some way?

This naturally brings problems what exactly is considered problematic. (After all, whether or not there is copyright infringement is a legal question and we do not necessarily have the access to all facts needed to asses that.)

But I am asking mainly because there is conflicting information from moderators on what should be done when you see posts which might contain links to material that is copyrighted. So it might be useful to have a place where the "official" policy on this is clearly stated. (Admittedly, some the information I link is rather old - from 2010 - but I did not find a more recent discussion of this issue; that is, until the few comments from September 2018.)

On one hand there is a rather old thread on tea: Links to possibly copyrighted material. One of the moderators (Ben Webster) wrote there:

In all previous intra-moderator discussions on this subject, our position has always been that we won't worry about what people are linking to (as long as its on-topic) until we got a take-down notice (which we never have), but this is in part due to our own ideological opinions on copyright (the original of Naive Set Theory was written in 1960; if the copyright law in force then had not been changed, it would be in the public domain in a couple of years). Do other people feel differently?

Similarly, there is this comment from Scott Morrison (also a moderator):

Several people have flagged this answer as requiring moderator attention. Please use your own judgement as to the legality of any site on the internet, and remember linking is always okay.

However, there was a comment exchange on a relatively recent post where the responses from the involved moderator suggest a different position (I have omitted usernames of the involved users and tried to select only the parts relevant to this issue):

@ToddTrimble: What I don't want to see are links to weird pirate websites, or an insistence on answers that only reference freely available sources ...

TT: if you or anyone else should ever see a link to a weird pirate website at MathOverflow, then you should report it right away; we simply can't have or endorse any legally questionable activity.


@ToddTrimble, I saw such a link just yesterday, and wasn't sure what to do. (Unfortunately, I now forget where.) What is the proper action? Should I flag?

TT: Yes, please flag it if you can find it.

@ToddTrimble "if you or anyone else should ever see a link to a weird pirate website at MathOverflow, then you should report it right away;" Sure. Precise reference more than suffices, doesn't it? :-)

TT: Please raise a flag for any such occurrence.

This post on Meta Stack Exchange seems related: What should I do when I see copyright violations posted on Stack Overflow? (However, I am fully aware that MathOverflow is different from a typical Stack Exchange site in many aspects.) There was a discussion on a similar topic also on Mathematics Meta: To what extent should copyrighted material be made available on math.se?

  • 5
    My own position since becoming a moderator is to avoid exposing MO (as part of SE, Inc.) to any sort of legal liability, and to abide by "better safe than sorry". I am of course aware of the prevalence of sharing information about certain sites where information is freely available, and the widespread resentment that corporate entities have control over content that mathematicians often create for free and would gladly share if it were entirely in their hands. However, some users feel very strongly about copyright law and some even give an impression they might complain if it were violated. – Todd Trimble Oct 30 at 17:08
  • One could ask how links for download of copyrighted math texts is like, or unlike, some other SE site with links for download of copyrighted movies; or of copyrighted software. – Gerald Edgar Nov 1 at 13:03
  • 9
    @ToddTrimble: The OCILLA/DMCA explicitly removes the liability for hosting user-posted links if the notice-and-takedown process is followed. Do you have any evidence that preemptive deletions of links have any legal effect at all? – Dmitri Pavlov Nov 1 at 15:59
  • 2
    @DmitriPavlov I'm not a lawyer, so it may take me some time to be able to study the matter in more detail, and to weigh pros and cons. Judging by the upvote and downvote patterns on answers here, it seems the entire matter is fraught with contention. – Todd Trimble Nov 1 at 19:17
  • 3
    I think in fact I will invite someone from SE Community Management to weigh in. I am uneasy in the sense that once someone brings to my attention links to a website said to engage in piracy, simply to ignore that notice or not acting on it could be viewed as a form of aiding and abetting. Perhaps I am simply naive or not sufficiently informed enough here, but I need to be very sure about what I'm doing; cf. "better safe than sorry". – Todd Trimble Nov 1 at 19:22
  • 2
    @ToddTrimble: Sure, one must take action if the copyright owner submits a correctly constructed DMCA notice. But what's the point of reacting to third-party complaints (i.e., not by the copyright owner)? How could the third party possibly know that the link points to something illegal? This can only be decided by the copyright owner. Some publishers have no objections to websites such as Library Genesis, and there is no way to find out the contrary without a properly constructed DMCA notice. – Dmitri Pavlov Nov 1 at 19:59
  • 3
    @DmitriPavlov I'm listening, but I'm not going to rush to reverse any editing decisions I've recently made without thinking very carefully about what we're doing here. Moreover, it may not just be a matter of law (cf. Andy Putman's answer). I sense we're on very tricky ground here. – Todd Trimble Nov 1 at 20:13
  • 4
    As far as I understand (IANAL), merely editing out these links does not change anything from the legal point of view, because the links are still there in the editing history; they just take two more clicks to reach. Or are we speaking about a different kind of 'editing out' that is accessible to moderators only and leaves no track in the history? In this case it would be good to have a complete picture of how it works, so that we can take a more informed decision when we vote. – Federico Poloni Nov 3 at 13:40
  • There is also an older discussion on tea: are we concerned about people posting illegal material? - however it is rather short. Perhaps also this discussion on meta can be considered related, but aside from a few comments, not much was said there: How does MathOverflow deal with the liability for the content of the site? – Martin Sleziak Nov 5 at 0:01

This is an answer as a private member of MO, without my moderator hat on.

I'll admit I've never understood the argument that there could be something wrong with linking; the internet is a big scary place, we're (mostly) adults, and you can make a link to something for many reasons, only some of which indicate endorsement. That said, lawyers will try all sorts of things to get their way, and it is possible we could receive a nastygram about MO containing links to sites which do outright illegal things. I would not want us to chose that hill to die on, and would recommend promptly removing the links, and of course notifying the community of the reason about the letter. (Personally, I think it would sort of be fun to tell everyone that mean publisher X had been threatening MO, just to see what happened next.)

My personal preference is that we don't formulate explicit policy on this; where individuals have the energy to replace dubious links with links to the arXiv, institutional, or personal repositories, that's great, and should be encouraged (even if only for the reason that these links are less likely to rot).

  • 11
    "I think it would sort of be fun to tell everyone that mean publisher X had been threatening MO, just to see what happened next." <-- I support this policy. – David Roberts Nov 2 at 8:05
  • 2
    Very nice answer. -- As I understand, originally the purpose of copyright law was to protect the rights of the authors and to allow them to financially benefit sufficiently from their works. I don't think the intention was to solely increase the revenue of publishers which get the copyright transferred from the authors for free, and to obstruct scientific progress by de facto hiding publications from a part of the community. – Stefan Kohl Nov 2 at 16:40
  • "I think it would sort of be fun to tell everyone that mean publisher X had been threatening MO" - regardless of my views on the original question, I don't think this would actually rebound to MO's benefit. I can envisage this playing out quite badly in the UK, for instance, or at least being met with indifference and "stop causing trouble, do what they say, and let us get back to the REF/TEF/NSS/Brexit/whatever" – Yemon Choi Nov 2 at 19:23
  • 2
    @YemonChoi: If one wants to make things right for every country which has less than one percent of the world's population, then one probably needs to stop doing anything at all ... . – Stefan Kohl Nov 2 at 21:03
  • 1
    @StefanKohl Are you taking a poll of the world's population, or the world's mathematicians, then? I would always advocate including links to arXiv copies or copies on authors' own websites, so my issue was purely a reaction to Scott's semi-light-hearted comment. Certainly I am not chastising Scott for expressing that comment; I just happen to disagree with it – Yemon Choi Nov 3 at 0:44
  • 1
    Indeed, @StefanKohl having talked to a range of researchers over the years from different backgrounds with different priorities, Scott's line "I would not want us to chose that hill to die on" is probably closest to my actual views – Yemon Choi Nov 3 at 0:48

The OCILLA (part of DMCA) already specifies a process (notice and takedown) that takes care of these issues. This process also removes from us the burden of determining whether a particular link is infringing or not: this is now the responsibility of whoever prepares the notice.

So MO should continue to process DMCA notices (which it already does), and no other action should be taken since this process already shields MO from any liability with respect to these links.

I believe there is conflicting desire in the community (and to me, no surprise if mirrored in the moderation team) as to the degree of availability of mathematical content on the Internet. I imagine we (or most of we) want all libraries digitized and all articles accessible. The reality is that providing this access costs money, and not many want to pay this cost. Further, there are websites which have what I will call bootleg or unauthorised archives which they share in violation of some existing agreement.

Until such things get sorted out, I recommend not linking to such bootleg sites for the simple reason that fewer lawyers is better. As Todd points out, "better safe than sorry". If some part of the community is willing to prepare a legal defense to protect MathOverflow in case of complaints of such links occurring, then we might move toward including such links. Since I don't know the full or even likely ramifications of such linking, I don't do it. If you want to protect this resource, flag the posts and find alternatives.

Gerhard "Just Say 'Not Through MathOverflow'" Paseman, 2018.10.30.

As one of the people involved in the exchange from which you courteously elided names, I think that it's good to get rid of such links. Anyone who wants to find them can do so without our help. Although I admit that what I'm about to say sounds suspiciously like the broken-windows fallacy, I think, all legal issues aside, linking to dodgy file-sharing sites makes us look less like a reputable scholarly forum and more like, well, a dodgy file-sharing site, at best one step removed.

  • 2
    What "dodgy file-sharing sites" do you have in mind? The deletion of links that just began currently seems to affect mostly Library Genesis, which is not dodgy, not file-sharing, and not a site. – Dmitri Pavlov Nov 1 at 16:03
  • 1
    @DmitriPavlov, dodginess is in the eye of the beholder, but I was responding to a question about "suspicious" links, with the target left unidentified. For the LibGen example, I'm not sure under what technical definition it is not a site, and I don't know what it means to say LibGen is "not file-sharing". (I don't disagree; I literally don't know what that means. LibGen certainly offers download links to files. Maybe they don't host them themselves, and that is your point?) – LSpice Nov 1 at 16:11
  • 3
    "I'm not sure under what technical definition it is not a site": Library Genesis is a library, i.e., a collection of files with books and other texts. This library is distributed through various means: torrents, HTTP, etc. One of these means is a web site, but this doesn't mean that the library itself is a site. – Dmitri Pavlov Nov 1 at 17:13
  • 2
    File-sharing means that files are uploaded by users, with little or no curation from the host. This is completely false for Library Genesis: even though users can upload texts, these submissions constitute only a small fraction of all texts in the library, and these submissions are still thoroughly vetted and processed. The overwhelming majority of all texts, however, is not contributed in this manner. – Dmitri Pavlov Nov 1 at 17:18

I think it is completely unprofessional and inappropriate for MO to allow links to pirate websites. If an author chooses to place his/her work in a public repository (like the arXiv or their website), that is great and we should link to it. But allowing links to sites that flout copyright laws potentially does two things: first, it opens MO to legal liability, and second, it could erode the broad support that MO has in the mathematical community.

I should emphasize that this answer is not taking a stance on copyright laws per se. I personally believe that information should be free, and this informs how I act professionally. For instance, I post my own papers to the arXiv and my website, and I have on several occasions allowed people to attend my classes/lectures even though they were not paying customers of my university. But I do support obeying the law.

  • 5
    I think that, except in the broad sense that doing (or not doing) anything potentially opens one to legal liability, it is far from clear that linking to anything does so. – LSpice Oct 31 at 17:36
  • 7
    You might want to familiarize yourself with OCILLA/DMCA before claiming that "it opens MO to legal liability". As long as the notice-and-takedown process is followed, the DMCA explicitly removes any liability for hosting user-generated content. – Dmitri Pavlov Nov 1 at 16:07
  • 6
    "But I do support obeying the law.": Could you tell us what law you are referring to here? As far as I am aware, there is no law that makes posting links to some sites illegal, but perhaps you have something else in mind. – Dmitri Pavlov Nov 1 at 16:10
  • 1
    To play devil's advocate, what if one gave the link lmgtfy.com/?q=free+access+to+papers with the caveat that one may stumble on a website that hosts illegal copies of papers? I always tell people the website in question is illegal, that its mission to distribute free copies of paywalled papers has lead it to being sued for millions and losing. I certainly don't think MO should officially condone people illegally being able to access papers using websites like these that they otherwise would have to pay publishers money to access. I rather promote efforts like freejournals.org – David Roberts Nov 2 at 8:12
  • 3
    To put it another way: if I write a book, I have the moral and legal right to choose whether or not it should be freely available on the internet. It is grossly immoral to claim otherwise. MO should not facilitate that kind of theft. – Andy Putman Nov 2 at 15:36
  • 1
    @DavidRoberts, while I think that DmitriPavlov's statements may represent an extreme end of the MO community's perspective, I think that it's not true that people using these sites are "illegally being able to access papers", depending on how the modifier is read. My understanding is that providing that access may be found to be illegal on the site's part, but availing oneself of that access is not illegal on the user's part. (As far as I know, file-sharing lawsuits against individuals have almost always focussed on uploading, not downloading—Aaron Swartz being a significant exception.) – LSpice Nov 2 at 21:48
  • 2
    "I understand that you don't believe in copyright laws": It is absurd to claim that somebody that advocates compliance with the DMCA does not believe in copyright laws. "have some kind of convoluted argument for why Library Genesis isn't violating them": Please do not shift the focus of the discussion. What is being discussed is whether links to Library Genesis are legal on MathOverflow. You did not address this question. – Dmitri Pavlov Nov 3 at 2:20
  • 2
    @DmitriPavlov: Under US law, the copyright owner is the once who is required to complain. However, under European law, it is not required that the copyright owner be the one who notifies the website of illegal content. It is not clear to me whether flagging a post counts as the appropriate notification, but if the moderators decide they want to force the matter I would be happy to send more official notifications. As for why I brought up Library Genesis, you were defending it on other threads. Finally, if authors were asked whether they consented to their works being hosted – Andy Putman Nov 3 at 2:34
  • 2
    on Library Genesis, then I would have no problem with it. Doing it without their consent is theft. I'm sure that many mathematicians would have no objection to it, but as an author myself (of innumerable papers, and of several books in progress) I can assure you that the number who do object is not insignificant. The point is not money (which obviously few mathematicians make from their writing), but rather control over how one's creative work is presented to the world. – Andy Putman Nov 3 at 2:37
  • 2
    "it has a Russian domain name": The domains of Library Genesis are libgen.io (British Indian Ocean Territory, United Kingdom), gen.lib.rus.ec (Ecuador), libgen.pw (Palau), bookfi.org (USA), bookzz.org (USA), bookza.org (USA), bookos.org (USA). Which one of these is Russian? "it is registered in some combination of Russia and Amsterdam)": It is registered in Seychelles, you are welcome to check the whois info yourself. – Dmitri Pavlov Nov 3 at 2:49
  • 4
    Well, the OCILLA/DMCA prescribes exactly this procedure for copyright owners, and I fail to see how the same procedure for authors would suddenly become "absurd and non-serious", unless the DMCA itself is absurd and non-serious. To the best of my observations, the overwhelming majority of mathematicians actually want their works to be accessible to anyone, in particular through Library Genesis. While a minority that objects to this exists, it is absurd to try to impose its values on the majority. That's why I suggest a compromise: authors can ask for links to their works to be removed. – Dmitri Pavlov Nov 3 at 4:01
  • 5
    I rather doubt that the "overwhelming majority" of mathematicians would support having their work pirated by Library Genesis. A large number, perhaps, but you shouldn't confuse your friends and acquaintances (or people you encounter on the web) with a representative cross-section of mathematicians. And in any case, the right to choose how you disseminate your work to the world is fundamental to the whole creative and scholarly enterprise; even if only a minority of people object, it is gravely unjust to deprive that minority of this basic right (as your "compromise" would do in practice). – Andy Putman Nov 3 at 4:24
  • 6
    I am pretty sure that most mathematicians very much prefer to see their work being cited, even if it means accessing it through Library Genesis, than to exercise their "basic right" to remove links to Library Genesis. "And in any case, the right to choose how you disseminate your work to the world is fundamental to the whole creative and scholarly enterprise": how so? If copyright disappeared tomorrow, very little in mathematics would be affected. Papers and books would written as before, etc. – Dmitri Pavlov Nov 4 at 5:43
  • 2
    @DmitriPavlov: The person making an extraordinary claim is the one who is duty-bound to provide evidence. Anyone with a reasonably broad knowledge of mathematicians and mathematical culture knows that mathematicians have diverse views about everything under the sun. It would be rather remarkable if an "overwhelming majority" agreed about any controversial point. As far as how mathematical communication would function without copyright laws, I don't think there is any evidence one way or another about how that would work. Certainly physical book publishing would decline dramatically. – Andy Putman Nov 5 at 2:43
  • 3
    I never made any extraordinary claims about "overwhelming majority" of mathematicians (but rather only about an overwhelming majority of mathematicians with whom I talked about this). "Certainly physical book publishing would decline dramatically.": speaking of extraordinary claims, this one certainly is one. With no copyright, we will have more physical books at more affordable prices. One only has to look at Dover, a company that republishes mathematical books that went out of print and their copyright was returned to the author. – Dmitri Pavlov Nov 5 at 7:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .