I have recently taken the step of self-archiving in our institutional archive http://www.dspace.cam.ac.uk/handle/1810/741/ and hope that some details may be useful to the discussion.
In particular I feel that the current colour system, while a useful general label for a publisher or journal is not sufficiently precise to encourage (or even allow) most authors to self-archive with confidence. In fact two different co-authoring colleagues have felt unhappy that even with a publisher labelled as "green" they were not prepared to self-archive for risk of conflicting with publishers' actual legal requirements. Here is a typical example from a "Gold" journal. I have removed the journal and publisher's name but it is on the SOROS/BOAI list. Reprints Printed reprints may be ordered at a nominal cost. Electronic files of the published papers (e-reprint) can be distributed by the authors for noncommercial purposes. and in each paper: 2004 by XXX Reproduction is permitted for noncommercial purposes. In my opinion this violates the letter and the spirit of the BOAI. Because of my commitment to OA I have written to the publisher and asked for clarification. If this list agrees on my interpretation the publisher cannot continue to call themselves an OA Gold journal. However this label will be so firmly embedded in ROMEO, the publishers' rubric, etc. that I am sure it will remain labelled as such. For part of the authors' submission - the supplemental data - it is worse. I know that for some of this list only "full text" is published but for many scientists the supplemental data (without which the paper is not allowed to be published) is at least or even more important. In the present case the publisher sends the data to a data aggregator which releases it with the following added copyright notice (I quote part and gently anonymise the source) This [data] is provided on the understanding that it is used for bona fide research purposes only. It may contain copyright material of [the data aggregator] or of third parties, and may not be copied or further disseminated in any form, whether machine-readable or not, except for the purpose of generating routine backup copies on your local computer system. The material redistributed by the aggregator is (AFAIK) the verbatim submission of the author except for the addition of the aggregator's rubric. AFAIK the authors have not assigned copyright to the aggregator (though they have been told that their material will be deposited with them) In my view this is incompatible with a publisher's status as a "Gold" publisher. I hope that this is an interim problem but I have a fear similar to other list members that we shall end up with such a diversity or approaches and licenses that the colour-labelling is almost useless. Some points: - I personally do not know where the definitive colour statement about a publisher is to be found. It seems that the ROMEO site is not updated (at least not to news announced on this list) and that its colour scheme is in conflict with the Harnad scheme. I had assumed until recently that the spectrum was yellow, blue and green, but a published announced that it was using the Harnad definition from this list where "pale-green" and "bright-green" appear to be used. The announcement on this list that the Royal Society of Chemistry was now a "green" publisher does not therefore, in itself, indicate how and when we can self-archive. Despite the uncertainty we have, in fact, self-archived at the address given above. - I proposed self-archiving to two other colleagues and pointed them to the announcement on this list. They were unconvinced and are, so far, unwilling to allow our joint manuscripts to be self-archived until there is a clear statement from the publisher indicating precisely what can/not be done. I attempted to use the ROMEO site as indication but it had not been updated. It is likely therefore that even if every publisher were "green" there can be copyright Fear Uncertainty and Doubt that discourages authors. Most will, I suspect, simply continue to sign whatever copyright form is offered by publishers. Most are unwilling or unable to challenge publisher's copyright or practices. Has the OA community considered creating their own license system and persuading publishers to adopt this? I use the analogy of OpenSource under which I publish most of my software. "OpenSource", like "green", is a political label, not a technical one, and there any many variants of OS (see http://www.opensource.org). Because I want my software to be re-usable in any manner I do not choose the GPL license (which effectively prohibits use in commercial programs) but use Artistic. The point is that the author can choose from a wide variety of preprepared licenses and can use these labels in a technical sense - the single word "Artistic" is all that is required. The system is policed - I received a mail from the FreeSoftwareFoundation auditor pointing out that a program I had released under GPL carried restrictions which were not permitted by the license. I am aware that Creative Commons, BOAI, etc. might provide some solution but I suspect a lot of hard detailed work including policing is required. May I publicly thank the help that I have been given by my Institutional repository colleagues. However it appears that I am the first person within the institution to request the self-archiving of a peer-reviewed manuscript and this points out the enormous amount of work required in the future. Peter Murray-Rust Unilever Centre for Molecular Informatics Chemistry Department, Cambridge University Lensfield Road, CAMBRIDGE, CB2 1EW, UK Tel: +44-1223-763069