I got in a conversation recently with a friend who is about to have his first paper published. It’s been through review and is now accepted at a well-respected old-school journal owned by a legacy publisher. It’s not an open-access journal, and he asked my advice on how he could make the paper open access.
We had a fruitful discussion, and we agreed that I’d write up the conclusions for this blog.
First, you can pay the publisher to open-access your paper. That’s a legitimate option at “hybrid OA” journals, which by this point is pretty much all paywalled journals. But even when the journal invites it, that’s not always possible. In this case, my friend has no institutional funds available, and really isn’t in a position to bung the publisher $3000 out of his own pocket.
The second option is to write to the journal saying that you select the OA option, but that since you have no institutional support you have to ask for a waiver. Will this work? It’s impossible to tell unless you try it. Some journals might have an absolutely-no-waiver policy; heck, some might have a “we always give waivers but don’t advertise the fact” policy. My guess is that most have no policy at all, but that editors (who are nearly all researchers themselves) will tend to be sympathetic, and support your case. Anyway, it can’t hurt to politely ask.
If that fails, the the third approach is to use the SPARC Author Addendum. Using this legal instrument (which is freely available), you do not transfer copyright to the publisher, as they usually request, but instead give them a non-exclusive right to publish — which of course is all they actually need. That leaves you legally free to post the accepted (peer-reviewed) version of the manuscript elsewhere: in an institutional repository, your own web-site or wherever. (I’ve never used this myself, but I hear it’s widely accepted.)
If the publisher is intransigent enough to reject the SPARC Addendum, the fourth approach is to dedicate your manuscript to the public domain (for example by posting it on arXiv with the CC Public Domain Declaration). [added by AM: in philosophy the standard site for this is PhilPapers.] Then return the copyright transfer form to the publisher, saying truthfully that there is no copyright to transfer. Publishers are used to dealing with submissions that have no copyright: for example, everything authored by U.S. federal employees is in the public domain. Their copyright forms usually already have a section for declaring public domain.
Finally if somehow all of the above tactics fail — if the journal flatly refuses to give an APC waiver, won’t accept the SPARC addendum, and rejects works that are in the public domain though not written by US Government employees — and if despite their evident hostility to science you still want to stick with the journal that accepted your paper — then you have one final option. You can just go ahead and give them the copyright, but then post the final PDF on your own web-site anyway. Of course, you are not technically allowed to do that, but historically it’s never been a problem. It’s very widely done — especially by old-school professors, because it would never even occur to them that sharing their own work could be a problem.
To be clear, I am not advocating the last of these. The four preceding approaches are better because they are fully in compliance with copyright law. But when dealing with a publisher that is simply determined to prevent your work from being read, then you have to weigh for yourself whether you’re more interested in respecting copyright, or doing what’s right.
This is the situation with several of my own old papers, which in my young and stupid days I signed over to publishers without giving it any thought at all. Having got myself into that situation, it seems to me that making those papers available anyway is the least bad of several bad options. But I would never choose that approach now, since I publish exclusively in open-access venues.
Option zero (not discussed here) is to use an open-access venue to start with: then none of these issues even arise. But failing that: