I have always thought sharing a PDF with a limited number of people was completely within the “fair use” realm of US copyright law. Of course, I’m not even close to a lawyer. I can copy a music CD for a friend legally (fair use), but I cannot digitally share it with a zillion people online. So, in my mind, it’s perfectly fine to ask your friend to provide a PDF from their library. But posting a PDF in a way that makes it easy for the entire world to download has crossed the line. - Steve Koch

Steve, I think your definition of 'fair use' clashes considerably with both the opinions of the media companies and that of the law. (Guess who pushes those laws... hrmmm...) A certain amount of low-profile sharing will be tolerated, since they can't stop all of it. If rooms like References Wanted get too popular though, watch out. - Chris Miller

Chris, I did some poking around, and you're right that it's not "fair use," but instead it's the "Audio Home Recording Act" that I was thinking of. http://en.wikipedia.org/wiki... Specifically, section 1008 which says, "No action may be brought under this title alleging infringement of copyright... based on the noncommercial use by a consumer of such a device or medium for making digital musical recordings or analog musical recordings." - Steve Koch

In the wikipedia article, they give the interpretation of this which I had been told a while ago, "Private, noncommercial copies by consumers using "digital audio recording devices" are explicitly protected by §1008. The Senate report defines noncommercial as "not for direct or indirect commercial advantage", offering examples such as making copies for a family member, or copies for use in a car or portable tape player." - Steve Koch

Of course, that's just wikipedia, but I remember reading that digital copies for friends (extension of "family") was OK, especially because it requires a whole lot of human involvement that doesn't scale like Napster did. I even bought a legal CD dubber that had a special ability to only make perfect copies from original CDs. 2nd generation CDs it would switch to analog mode. - Steve Koch

And of course, finally, this has no bearing on legality of PDF sharing. But it does shape my own personal opinion quite a bit. And I personally thinks it's OK and reasonable to send a PDF to someone. Especially on those occasions when it's an author of the paper who can't access the article at his institution! :) - Steve Koch

We often find a difference between legal interpretation and what we think is "right". Like Steve, I think there's a big difference between privately emailing a PDF to an individual and posting a PDF on a public website but like most people, I have no idea about the legality of the former act. It strikes me as another case of the law failing to adjust to "information in the digital age". - Neil Saunders

IANAL but its not copyright that you are breaking but the TOS of the licence that the institution has with the vendor. Its a contract breech. Most likely the institution has the TOS posted somewhere (or should) that you can check. They do differ but most prohibit redistribution especially outside the subscribing org. - suelibrarian

Good point sue, I hadn't thought of that. - Steve Koch

it is not legal to make a copy of a cd for a friend (in the US at least). you can make back up copies, change the format, things like that - but by making a copy for someone else, you are depriving the record company of a sale -- regardless of what wikipedia says. You can sample parts of songs under fair use, parody the songs... but making whole copies of a cd - no way! As Sue says - we are forced to sign a lot more away in the license than the law requires. there's also dmca crap. - Christina Pikas

There are very specific exemptions that allow libraries to do interlibrary loan. Further, libraries comply with CONTU guidelines and the licenses signed by the institution. It is NOT the same for individual end users to email articles around. Everyone knows that this is pretty common - but if done enough, the borrowing institution really won't need to have a subscription or to pay per view, so it really does deprive the publisher. - Christina Pikas

"if done enough, the borrowing institution really won't need to have a subscription or to pay per view, so it really does deprive the publisher" -- the question for the publisher is how to balance the impact on subscription with the terrible PR that legal action would bring. From what I can see, it would take a lot of p2p access sharing to put a dent in library subscriptions, so publishers are very tolerant of the practice. - Bill Hooker

@Christina, it's not merely what wikipedia says. The wikipedia article is citing US law as well as a US Senate Report from 1992. Unfortunately, I don't know how to access Senate Reports. You can find legal briefs all over the internet that reference the 1992 Senate Report on the AHRA which say that the senate clearly indicated that copying for family members is ok. So it's no just what wikipedia says, though wikipedia continues to be a fantastic starting point. Here's a couple links (the first is apparently peer-reviewed) (search for the phrase "family member"): http://www.meiea.org/Journal... & http://news.zdnet.com/2100-95... - Steve Koch

Related discussion here: http://i9606.blogspot.com/2009... - Bill Hooker

family member != friend, and also look at how the law is commonly interpreted. If you want a senate report the easiest way if you're at a university is to use Lexis-Nexis Congressional. It might actually be in the US Serial Set, too - so your Gov Docs librarian at your local depository library can help. - Christina Pikas

In my mind, if I can legitimately show someone else my physical hard copy of Science or Nature (I highly doubt the law prohibits that!), why can't I do the exact same thing (conceptually) of showing/sending someone else a PDF? - Benjamin Tseng

Firstly, can I say a big thank you to Heather for stimulating such weighty discussions. I certainly feel a blog post coming on. In the meantime, does anyone remember this from the 80's ? http://en.wikipedia.org/wiki... - Graham Steel

@Chris Miller: can you tell me more about your view of Fair Use? IANAL and I have not been able to find any definitive information on whether (or under what circumstances) p2p access sharing breaches Fair Use. The relevant statute does not provide much guidance; as Christina points out, precedent is very important, and I have not been able to find any test cases that didn't involve commercial entities being sued by other commercial entities. If you know any such, please share! - Bill Hooker

@benjamin it's like first sale or non-rivalrous use or something... when you have a print copy, and you give it away, you no longer have it. if you give away a pdf, you still have the original. that's why it's different. - Christina Pikas

Of course you could always do what I do whenever I need a copy of an article in a journal that my institution doesn't subscribe to: send a "reprint request" to the corresponding author. They'll usually email you back a PDF all legal and stuff. - Miss Elle

@benjamin also the contractual arrangements for a physical object are very different to those for access to a PDF. Essentially physical copies are sold, electronic copies are leased. @miss elle, yes that is usually the best approach, almost certainly a technical violation of use terms though in most cases, unless the authors have bought or been given a specific set of PDF downloads (generally in this case they will send you a link and you have to go to the journal site to download). - Cameron Neylon

I've embedded this thread on my blog, here:- http://is.gd/11FUk And then of course, there's self archiving. I recently wrote to a Prof to congratulate them for a very impressive record of self archiving their work (~200 PDF's). They responded with "Thank you for your interest in our research. We are probably violating copyright assignments but I don't think anyone will object". The overall theme here to me is that science WANTS to be out there, widely read/seen/cited without restrictions. - Graham Steel

@Graham - thanks, but the credit actually goes here to Craig Rowell, who launched a discussion on NN in which I decided to word my response more carefully than usual: http://network.nature.com/people... - and Steve has linked to this thread. - Heather

Oh, and also @Graham - I wouldn't anthropomorphize science, as that would not stand up in a court of law facing a breach-of-contract lawsuit. Some scientists, like myself, would like unrestricted distribution. Then again, I benefit from having journals in which to publish and I don't want them to go bankrupt. - Heather

A fascinating question and one that medical librarians struggle with daily. - Hope Leman

@Heather, I had to look up 'anthropomorphize' and catch your drift. Also, thanks for pointing out that it was Craig who launched the discussion. As such, I've left a comment over at NN, taking Hope's comment here into consideration. (FTR, the comment over at NN didn't come from Hope, but another medical librarian I have had contact with over the years). - Graham Steel

So where do document sharing sites like Mendeley and Lab Meeting come into play on this? They can archive all of your PDFs and in turn allow sharing of them with colleagues and site members. - Brian Krueger - LabSpaces

As Victor & Jan will tell you, this was a concern of mine. Their primary response is to only allow sharing within a limited group of friends - but we'll see when this is tested. - Christina Pikas

@Christina - I completely agree. Here's the Mendeley - "How It Work's" page as it currently stands:- http://www.mendeley.com/how-it-... - Graham Steel

I don't think either of them are very worried about it. Had a "discussion" on twitter about this with Ricardo Vidal and he essentially just told me to submit a comment about it in their community issue forum and if it got enough interest they'd "look into it." It's a great business model, and until legal action is taken I don't see them changing that aspect of the site. The thing that irks me is that they have a "Submit a DMCA" link on their main page. They know the capability of their technology and do nothing to stop it. Their software can determine title, author and publisher of a document. Therefore, they know what is copyrighted and not copyrighted, yet still allow sharing. That's unethical at best. It's as if they're saying, "Sure, here's the DMCA link, now prove sharing of your material." - Brian Krueger - LabSpaces

Personally, I only allow Mendeley sharing of PDFs of papers on which I was an author. I consider I have more of a right to distribute those. But of course, given they are typeset, I don't really always - it depends on the copyright agreement. Nonetheless, it saves me from answering reprint requests in theory, and I feel ready to defend the practice if need be. I always do send journal-typeset reprints in PDF form, and so does everyone I know. - Heather

We can, of course, solve this problem completely for the entire future, by only sending our work to OA journals, or choosing journals that don't require copyright transfer. It would also help if we no longer provided free peer review to toll-access, copyright-snarfing journals. That still leaves us with a decade or so of backlog problem (meaning, after about ten years the backfiles won't be such a big deal) -- except that the majority of journals will probably switch models, and may even open up their backfiles as well. - Bill Hooker

This will never work, though, because at least 50% of scientists will say to themselves "ah, the kum-ba-yah save-the-world fools have handed me a huge advantage by cutting themselves off from so many high-IF journals, which will also be getting fewer submissions". This is why science cannot have nice things; see also http://friendfeed.com/alethea.... - Bill Hooker

Thanks for the shout-out, Bill, and I am afraid you are right... - Heather