These are all great questions.

Let me start with the second question:

a) The Creative Commons by Attribution License is certainly not the only option. It is however a very popular one. Many of the Open Access journals have adopted this license. In particular PLoS and the BioMedCentral Journals.

In practice, any other license would work as long as it specifies that you retain the right of distributing your paper in PubMedCentral (PMC) *AND* allowing readers to download it from PMC and redistribute it. Note that just the permission to post it in PMC is not enough, because then, readers who download the paper are still missing an authorization for copying the paper, and you are forcing them to engage in copyright infringement in order to read the result of research that they have paid with their taxes. The public shouldn't be forced to engage in crime in order to learn what was done with their money.



Do you need your institution to agree ?

YES.

In fact your institution *ALREADY* agreed the day you submitted a proposal to NIH. It also ratified the agreement the day you start spending NIH funds in your research. At that point, the new Federal Law requires you to make your papers publicly available.

Note that your institution *always* have to approve *any* copyright transfer agreement, since as your employer, the institution is the real copyright holder of the paper. Many researchers ignore this fact, and they engage in fraud when they sign copyright transfer agreements for papers without the authorization of the intellectual property office of their institution. The same researcher is also engaging in copyright infringement (another federal crime), when they send the paper to a Journal without the authorization of their institution, who, again, is the real copyright holder of the work. The Journal, also engages in copyright infringement when they publish and sell that paper for which they only have a fraudulent copyright transfer form.


About the first question:

No,
I don't know if anybody has done
that to a Journal.

So, I propose we do it. ! :-)


Let's submit a paper to IEEE TMI discussing this issue and let's specify that we will not sign the standard IEEE copyright transfer agreement, but instead we offer two options:

a) Assign a Creative Commons by Attribution license to anybody (including of course IEEE)

or

b) Modify the text of the standard IEEE agreement to specify that the authors will reserve the right to submit the final version of the paper to PMC, *and* permit readers of PMC to redistribute the paper.


I'm attaching a modified version of the IEEE Copyright Transfer form, that will simply assign to IEEE a license allowing them to copy, distribute, perform public display and produce derivative work from your paper. In this modified version there is no copyright transfer involved.



Luis