This RCom subgroup is dedicated to developing an open-access policy as a requirement for research projects to receive significant support from the Wikimedia Foundation.
So far, the most influential policy regarding Open Access has been that of the NIH, which states:
The Director of the National Institutes of Health shall require that all investigators funded by the NIH submit or have submitted for them to the National Library of Medicine's PubMed Central an electronic version of their final, peer-reviewed manuscripts upon acceptance for publication, to be made publicly available no later than 12 months after the official date of publication
Receiving "significant support" from the Wikimedia Foundation (WMF) for research projects shall be subject to the following conditions:
If a manuscript describing research from the project is accepted for publication in a peer-reviewed venue, an electronic copy of the final version of the author's peer-reviewed manuscript is to be submitted to a public and permanently archived repository no later than the official date of publication. This public electronic copy must be provided in at least one of the following file formats: HTML, XML, MediaWiki ML or TeX. Other formats are optional.
The authors of the manuscript must retain a non-exclusive right to copy, distribute, transmit and adapt their work, and must use it to grant the WMF a non-exclusive right to do the same.
Data gathered in the framework of such a research project shall be deposited in a public repository under a reuse-friendly license by the time of publication of the respective research article or within twelve months of having been collected, whichever comes first. Immediate deposition of such data is encouraged.
Exceptions to these rules may be granted by the Research Committee as part of the approval process for receiving WMF support.
For other kinds of support from the Wikimedia Foundation, some of these requirements turn into recommendations, as detailed in this table.
Perhaps the permissible file formats can be summarized. Example from the "Free Cultural Works" definition: "For digital files, the format in which the work is made available should not be protected by patents, unless a world-wide, unlimited and irrevocable royalty-free grant is given to make use of the patented technology. While non-free formats may sometimes be used for practical reasons, a free format copy must be available for the work to be considered free."
Eloquence — my own bias is going to be very strongly to tie any access to restricted resources that require WMF time and effort to open access / open data practices and policies within the research community
lucadealfaro — if we required open access to what everybody gets out of running a dump analysis, we would only push underground a lot of results.
jtriedl — I think we should definitely include some notion of open access / open data. We should be very careful about how we define it, though. Do we want to refuse to support research that is published in places that do not allow free distribution of articles? I would be hesitant about that.
lucadealfaro — I think it would be easier to have a policy that said: For some data (dumps, etc): do whatever you want and don't ask. For user studies that require WMF resources, user involvement, or anything like that, we review (or ask the community to review)
I support the addition of an open-data requirement. But to avoid cluttered language trying to cover too many issues at once, I'd put it into a separate paragraph.
I recommend requiring a format like HTML or XML that fosters rather than inhibits reuse. Authors may deposit PDF or DOC (or anything else); but if they do, they must also deposit HTML or XML.
The policy must address copyright. If authors transfer full copyright to a publisher, then complying with the WMF policy will require either the publisher's permission or copyright infringement. The NIH solution to this problem is simple, effective, and battle-tested: NIH grantees are not allowed to transfer full copyright to anyone. On the contrary, when they publish articles based on NIH funds, grantees must retain a non-exclusive right to make their work OA, and must use it to grant the NIH a non-exclusive right to do the same. (NIH grantees may, and typically do, transfer all remaining rights to publishers.) This works because funders are upstream from publishers, and researchers sign their funding contracts before they sign their publishing contracts. NIH puts a sharper point on it: if a given publisher demands full copyright, demands exclusive rights, or in any other way will not allow the author/grantee to comply with the NIH policy, then the author/grantee must look for another publisher.
