Tuesday, July 30, 2013

Proposal: The Dated Creative Commons License

Back on June 15, Peter Suber's book Open Access itself went open access, one year after its initial publication. You can get the ebook for free from MIT Press, but because of the Creative Commons license you can also get it from Internet Archive, and Unglue.it has a page to help you download it. It seems appropriate for this book to be its own publishing experiment, and from what I hear, the book has done well, in addition to doing good.

The "embargoed" or "delayed" model for open-access is tried and true in the scholarly journal business, and arguments about the appropriate length and propriety of embargoes are entrenched. In medical research, funding agencies such as NIH  demand embargoes of no longer than 12 months,  while humanities publishers argue that they need longer embargoes. Recently, the American Historical Association recommended that doctoral students be allowed to embargo their dissertations for up to six years. (Suber's book discusses delayed open access and embargoes in chapter 8, Casualties.)

Delayed open access for books, by contrast, is almost nonexistent. For ebooks, it would seem that an exclusive selling period followed by Creative Commons licensing could unlock a lot of value for society, and not just for scholarly works. Most books do most of their sales in the first year of publication and not much after that. The current duration of copyright, typically more than a hundred years, seems disproportionate in comparison. Used book stores capture some of the residual value of print books without profit to the rights holder, and libraries help to preserve another chunk of value. The lack of first-sale rights for ebooks leaves huge doubts about the viability of these channels for ebooks.

MIT Press accomplished the delayed open access with a promise on the copyright page of the ebook. Readers could rely on the integrity and prestige of MIT Press to make good on that promise. Wouldn't it be nice if doing something similar was easy to do for any sort of work, just like attaching a copyright date? Suppose I wanted exclusive rights to this blog post for five years, it would be nice if I could just write "(CC BY 2018)" with a url to provide the legal code.

I don't think I can really do that easily, today. Without some sort of license language, nothing would prevent me from changing my mind, so my prospective license offer would not be reliable. Today's Creative Commons licenses depend on conveyance of the license and assume immediate effect.  In the publishing world, companies go bankrupt or get acquired all the time. If Elsevier had acquired MIT Press in May, a purchaser of the book in April would have no assurance that the book would really go Open Access in June. This is not such an issue with journals because they're continuing publications.

Although I'm not a lawyer or anything, I've taken a first stab at language for applying a future date to a Creative Commons license. I've used Docracy to make the document public so that anyone can make modifications to improve it. (If you do a lot of contracts and you haven't seen Docracy, I suggest you go check it out!)  Or maybe other people have worked on this and can contribute some better language.

The beauty of Creative Commons is that it gives creators more options for distributing their works in partnership with users. A robust way of granting future CC licenses will allow more creators to vote with their works for mitigation of over-long terms-of-copyright.

Update: a quick comment from Timothy Vollmer points to a thread on [cc-licenses] that's very relevant, including some interesting discussion of the mysteriously named "Founder's License".

Update 2: James Grimmelmann, a real law professor, suggests via Twitter that "It is not out of the question that one could unilaterally enter into a binding future license at present." and points to the language he used on a 2005 Yale Law Journal Note:
Copyright © 2005 by The Yale Law Journal Company, Inc. For classroom use information, see http://www.yalelawjournal.org/about.asp. After June 1, 2006, this Note is licensed under the Creative Commons Attribution 2.0 License, http://creativecommons.org/licenses/by/2.0/. Any use under this license must carry the notation “First published in The Yale Law Journal, Vol. 114, pp. 1719-58.”
This suggests that maybe I'm making things too complicated, which wouldn't be the first time. But I wish Creative Commons or someone would just tell us what to do!

Update August 9: I've written more about what we want to do with Dated CC at Unglue.it.


  1. I proposed this idea in a Twitter conversation on May 13:
    and again on July 1:

    Also, it could be said to be an aspect of Library License http://librarylicense.org/.

    Tim McCormick
    @tmccormick tjm.org Palo Alto, California

  2. Thanks for "Dated Creative Commons License" proposal, this is a great project. I had a few thoughts examining possible terms and cases, presented here for discussion:

    1) Is it necessarily Creative Commons?
    For example, Library License has dating provisions, or one might want to make a Library License that coverts to a CC at some point.

    Or, it may be that the alternative possibilities considered here could be expressed by future CC license versions, or by using Creative Commons' "CCPlus" framework (http://wiki.creativecommons.org/CCPlus), which of course has the advantage of optimally aligning with CC and probably better expressions.

    2) Perhaps generalize 'dated' to, say, 'convertible'?
    (note, copyright-based licenses are in one sense already dated, by the expiration of the underlying copyright). In this concept, a license might change based on conditions beyond just one date. For example:

    a) if not renewed by copyright holder. I.e. avoid 'orphan works' problem.

    b) content unavailability trigger conditions, analogous to those defined by LOCKSS, e.g. publisher goes out of business.(https://www.clockss.org/clockss/FAQ#How_does_the_CLOCKSS_board_define_a_trigger_event.3F).

    c) license converts if some certification condition not met. E.g., a program of depositing at Internet Archive lapses, or journal loses Directory of Open Access Journals [or Books] approval.

    d) Performance/revenue target. For example, the case given by Drew Roberts on [cc-licenses] list of a work license converting once it's earned $5,000 revenue. This may actually be an approach for a Knowledge Unlatched / Unglueit-type crowdfunding model.

    e) While dating is usually discussed for changing toward fewer use restrictions, how about a case of the reverse? For example, say I wish to more publicly promote a book before or at early point of 'publication', but revert to an NC or regular copyright at a point. Possibly, later convert again, say to public domain or CC-BY after 10 years, so there'd be multiple license conversion dates .

    Haven't entirely thought this through, but seems an interesting case to consider. Periodically I've suggested to Springer, Elsevier, etc., that they make new papers available for some short time initially, e.g. for general public, social-media and journalistic dissemination -- which might non- or positively affect later use/revenue.

    Of course all this gets into the issue of what complexity is justified, from various standpoints such as making it legally draftable, non-cooptable, the concept communicable, etc. The [cc-license] thread cited makes interesting points about structuring options to guide toward more standard & desired outcomes.

    Finally, I think it would be useful to have terms/hashtag to track this discussion. I nominate and will use unless otherwise suggested:
    but some other suggestions are #datedlicense, #changerights, #changemark, #changelicense.
    thanks for kickstarting this discussion,

    Tim McCormick
    @tmccormick tjm.org

    1. Lots of things to think about. On (e) however, I think it's a really bad idea to associate CC with anything revocable. There's already too much confusion about this, even among people who should know better. And in general, simple works better; options have to be worth the complexity cost.

  3. well "revoke" implies that one state is 'on' or right, one off. If you have a NC that converts to BY, is that revoking the NC, or opening?

    It seems that many aspects such as legal drafting may be best looked at neutrally as q. of how to make state A convertible to state B, for all combinations of A and B.

    1. The way CC licenses are drafted, you could be conveyed an NC license and a BY license separately and your rights under BY would not be restricted by the NC clause. Except for the "Entire Agreement" paragraph. Hmmm.