I have sympathy with Stevan's view that encouraging authors to retain copyright (and publishers to do without it) may prove a complicating factor in the bid for open access. I also agree that being licenced back the right to self-archive is a big step in the right direction. However, we have to remember that over a quarter of publisher's policies (in the RoMEO analysis) do not give the author the right to do anything at all with their own work, let alone self-archive. And, although 50% of journals on the RoMEO publisher copyright policies listing -- http://www.lboro.ac.uk/departments/ls/disresearch/romeo/Romeo%20Publisher%2 0Policies.htm -- allow author self-archiving, 30% only allow self-archiving of the preprint. Yes, we can all add our corrigenda files, but who really wants to read a research paper that way? As Pam & Steve point out, there are activities that researchers, academics, and libraries, would like to perform that they cannot if copyright is owned by the publisher.
The fact is that publishers do not *need* copyright assignment (or an exclusive licence - often the same thing by the way) in order to publish. They require only a non-exclusive licence. Why then assign copyright? Instead of publishers licensing back to academics the right to perform a few meagre activities with the academic's own intellectual property, why don't academics license the publisher the meagre rights they need to publish the work? The value that publishers add to the publishing process should not, in my view, be paid for by copyright assignment. It should be rewarded and protected in other ways. Elizabeth