:)

I see. So it sounds like the value provided by journals is not anything to do with distribution but rather the “seal of quality” conferred by a journal. If the journals all go out of business how do we know which papers are good any more? My tongue is only half in cheek here. Even perfect search (which we don’t have) isn’t quite the same as good curation.

I don’t know of any such formal proposal, and my comments are admittedly not very clearly thought out since I only spent a couple hours writing them up. As to how to resolve conflicts between multiple authors and institutions, the easiest way to give everyone the independent write to distribute, so that the work is publically available if anyone wants it to be. In other words, if you want to keep something private and don’t trust someone who helped, you have to find a way to not put their name on it. Critically, that wouldn’t authorize one author from submitting a modified version if another author complains, for the same reason I can’t publish any random statement and claim that Rob said it.

Generally it isn’t the research institution making an exclusive deal, but the authors themselves on a paper by paper basis. The incentive to authors is the extra credit one gets for papers in good journals, given that a field has decided a journal is good. Many publications nowadays explicitly authorize authors to post papers online (though often only preprint versions), but there are plenty of journals that still don’t. Or at least, I assume so having never published in one, based on the rather large percentage of recent papers that aren’t available online.

It looks like all Elsevier journals allow self-posting, but here’s an example of a math/cs publisher that doesn’t (found by checking through a few of my recent emails to Eugene asking for papers (he has magical access to all journals for some Canadian reason)):

http://www.informs.org/Find-Research-Publications/Journals/Author-Portal/Publications-Policies/Rights-Permissions#own

Is there a formal proposal somewhere about how this right to distribute would work? In the case of multiple authors, how are conflicts arbitrated? How would you prevent institutions from applying pressure on individuals over how they exercise (or don’t exercise) that right?

I have no background here so I really don’t know why a research institution would choose to make an exclusive deal with a private publisher. But assuming there’s some incentive to do so that’s a threat to individual authors’ right to distribute unless that right is protected fiercely.