Sharing is caring
(the following has been dashed off rather quickly. Suggestions for improvement would be good) #Elsevier
has updated its 'sharing policy' . I like some bits, but others make me pause. I find it amusing that the company is saying to researchers 'you may do this and that', when, in my corner of mathematics at least, we've been doing this for years without need for permission. It feels a bit like catchup to me. But there are other areas of science etc that don't have a culture of this, so it's a new world for them to even think about sharing preprints online.
On the positive side, if you haven't signed #thecostofknowledge
boycott and are still publishing with Elsevier, then this is the best point to my mind (not sure if it is new, perhaps +Alicia Wise
can tell me):Authors can share their accepted manuscript...Immediately... via their non-commercial personal homepage or blog. [or] by updating a preprint in arXiv or RePEc with the accepted manuscript...
This means it is all the more important to actually have a preprint version
of the paper on the arXiv (or other public repository). Otherwise one must obey an embargo before posting it there.
What I don't like is this:In all cases accepted manuscripts should: ...Bear a CC-BY-NC-ND license...
Perhaps Elsevier has an argument about applying NC (i.e. non-commercial) if they want exclusive economic rights. But note that at this point one has not signed any contract, so what you do with your manuscript is your own business (unless Elsevier adopts the pernicious policy of having EULA-style 'submit and you agree we own all copyright in the case we accept' terms - which I've seen some publishers use). However, the ND (no derivatives) is a bit flaky, since this basically removes a lot of options for actually making use of the fact it's open access. Suppose I am a non-sighted person without a subscription to some Elsevier journal, and I find the articles on the arXiv. I am unable to convert them to, say, XML+MathML and a form suitable for screen readers etc and post them online for others in my position. Witness the recent controversy about digitising works for the vision impaired. Even the author wouldn't have these rights after signing the copyright to Elsevier and having to abide by the ND clause of her own article.
I know why this is done, since it is to protect Elsevier's commercial interests in, for instance, rights to figures that you might want to reproduce, but in a different form, or to various adaptations that open and sharing academics might want to do. +Peter Murray-Rust
would be livid I'm sure, since it's taking a free resource (making the big assumption authors used a CC license on the arXiv to start with with) and binding it up.
The funny part is that the arXiv doesn't offer CC-BY-NC-ND
in the license metadata, only CC-BY-NC-SA ; so it's 'share alike' (the license follows derivatives and redistribution) and not 'no derivatives'. Such a thing -- the ND clause -- would have to be placed in the comments or in the text of the article itself, not very good from a technical perspective, since one can query the arXiv API for license data (+Mike Taylor
was asking me about this), but making authors add licenses where they will easily be missed is not a smart move. Also, copyright lines cannot be included in the metadata and apparently will be removed!  Articles uploaded without Elsevier's chosen license cannot be removed in any case, so if one accidentally uploads an accepted manuscript, good luck in getting it off.
cc: +Scott Morrison
 http://arxiv.org/help/license #publishing #openaccess