In this question the OP links to a nature paper that s/he is interested in re-typesetting. The link appears to be to an illegal copy of a non-open access paper that is available (presumably behind a paywall). Should we allow links to pirated papers or force people to link to the non-free version?

For some questions you might need to be able to see the paper in order to answer the question, but in other (e.g., this case) you might not need to see the paper to answer the question. Does this affect or decision?

  • 4
    Before removing such a link, do a quick search to see if the author/their university host a copy or if it has been submitted to a reputable pre-print archive. Those shouldn't be considered pirated. Jan 21, 2014 at 15:30

4 Answers 4


We shouldn't be posting unauthorized links, if at all possible to avoid doing so. I don't want to say "absolutely not," because it can serve a useful purpose under limited circumstances. But "probably not" is eminently reasonable.

  • quoting a long paragraph or (if it's math and figures) posting a snapshot of the relevant section, might be OK if it is needed to answer the question. But I agree we want to avoid illegal links if at all possible. Jan 19, 2014 at 2:46

I think we should allow links, regardless of the legal status of the content they link to.

As a general rule for the Internet, I would prefer to keep as little self-censorship as possible.

Moreover, having too strong policy on removing alleged links to piracy will result in removing some links that are legal (e.g. self-archiving that is legal).


While I think that it is a good practice to use an official link (preferably arXivID (http://arxiv.org/abs/...), DOI (http://dx.doi.org/...), or another id-based link), the direct access is important (without it some questions, or answer, may be incomplete).

So how about using an official link plus (if it is not open access) another link (not synonymous with illegal!)? If it might be illegal, still - IMHO it should be on the conscience of the person who has uploaded it, posted the link or entered the link.

(Again, on the Internet, I prefer under-policing to over-policing.)

  • I think this brings up a good point. I don't like the idea of removing an unofficial link when there is no official link to replace it with. I just don't see the issue with replacing a legitimate link (i.e., a self-archive) with another legitimate link. Maybe we should encourage the person posting the link to say something about why they think the link is legitimate.
    – StrongBad
    Jan 23, 2014 at 9:02
  • 4
    @DanielE.Shub replacing a legitimate link with another legitimate link, where the second is behind a paywall and the first wasn't, effectively removes access to the paper to anybody who isn't at a university!
    – Flyto
    Jan 23, 2014 at 11:22
  • @Flyto no it doesn't (at least generally), it just makes it a little more difficult.
    – StrongBad
    Jan 23, 2014 at 11:34
  • @DanielE.Shub I would disagree with that. Even if the open version is easily googleable and the reader is comfortable with looking for free versions when they hit a paywall (both nontrivial assumptions), I would fear there's a big chance the potential reader will simply not bother. (After all, there's a risk in going fishing for free versions, as you never know how long the search might take.) Even with easily-googleable papers, the replacement quite drastically reduces the usefulness of the link.
    – E.P.
    Jan 23, 2014 at 13:24

Regarding posts that contain links to freely-available versions that are not obviously pirated: I don't see the point to removing links to content. While there is indeed an argument for the permanence of the journal version (though this is only valid if it's a DOI link!), I don't see why both types of link can't coexist.

If a post already has an ostensibly legitimate link, add the journal (DOI!) version instead of replacing the eprint.

  • Good point re DOIs. Ideally, from a permanence point of view, people wouldn't use any other link.
    – Flyto
    Jan 23, 2014 at 13:54

I agree with others that we shouldn't host links to copyright-infringing copies of papers. However, since different journals permit reproduction of different stages of the publication process in different circumstances (e.g. some allow authors to put a copy on their personal website), it isn't necessarily straightforward to determine what is a copyright-infringing copy.

Therefore, I think that we should refrain from flagging this unless it's very clear that the version being linked infringes, and perhaps add something to the Help for the site pointing out that we don't want dodgy links.

  • If we always link to the official journal version, then we know it is a legal version. In some cases this will make it a little harder to answer the question or understand the answer, but I think we need to be good citizens and support copyright even if we don't agree with it
    – StrongBad
    Jan 22, 2014 at 10:02
  • 2
    So (to take an example) if I am the author of a paper, and I have a legitimate copy that I am allowed to share on my website, you feel that I should not be allowed to link to that, but should link to a paywalled version instead?
    – Flyto
    Jan 22, 2014 at 10:19
  • It might be best to ask this as a separate meta question because I can see arguments for both sides. I see two benefits and one drawback from linking to the paywalled version. The first benefit is it makes it easier for mod/editors to make sure we are not prompting piracy. The second is the paywalled version is likely more permanent then a link on you personal webpage. The drawback is that it will take a little more effort from users to get to the free version. The benefits to me outweigh the costs.
    – StrongBad
    Jan 22, 2014 at 10:32
  • 3
    @DanielE.Shub hmm, fair point about permanence. I think that if it were me, I would perhaps provide both links. My general feeling otherwise is that we don't take it upon ourselves, as a community, to check the copyright status of everything that anybody links to on Stackexchange - so why should academic papers be different? While we certainly should not encourage copyright infringement - and might even want something in the site help about not linking to infringing versions - there's no reason to try to police every paper that somebody links.
    – Flyto
    Jan 22, 2014 at 17:04

You must log in to answer this question.

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