CC-BY vs CC-BY-SA and How to Define Open Content
As an open source and open content advocate, I am interested in the debates in the OER community about licensing, particularly CC-BY (attribution only) vs. CC-BY-SA (attribution and share alike). To admit my bias, I am, and have always been, a copyleft advocate, so I'm a card carrying member of that latter group of those that believe in share alike.
Now, there are certainly many good reasons to choose either license depending on one's priorities in license choice. Nevertheless, I always chuckle when the CC-BY advocates condemn CC-BY-SA while championing that CC-BY allows for users to make choices about licenses they wish to use, whereas CC-BY-SA doesn't allow for licence choice. As Sam Varghese pointed out during a resurgence of a similar debate (2007) in the open source community between the GPL and BSD people,
Curiously, the BSD people, who have no objection to their code being locked up for good in proprietary applications, raise hell when someone uses the GPL instead. Why? The only thing that the GPL ensures is that any code released under it will be free thereafter.
While some CC-BY advocates may indeed believe in privileging private commercialization of IP in opposition to Share Alike, I believe that many instead are not a fan of proprietization of their work, and what is unsaid but implied in their license choice is the following:
if you use and remix my CC-BY content, you have the right to choose how to relicense the work however you see fit, but you really should choose to relicense derivative versions under CC-BY.
How similar this is to the parent who gives a teenager twenty bucks, telling her "she can use it however she wants," when the teenager knows what the parent really means is in a responsible manner that the parent would approve of. If this hidden implication is the case for some content creators, then authors might consider getting off the fence and choosing the license which accomplishes what they really want.
What stimulated this thinking today about licensing was David Wiley's condemnation of Open Knowledge Foundation declaration that SA is more open than content with the Noncommercial (NC) clause. Now, David, in claiming ownership of the definition of "open content" because he coined the term, points to his 2004 definition Defining the "Open" in Open Content. I can partially agree with him. I think it's difficult to define SA as more open than NC when interoperability is a constraint.
However, I don't agree with Wiley's current definition of open content, but rather with the somewhat different implied definition of "open content" in Wiley's OpenContent License (1998) and Open Publication License (1999) (see the 2003 version of these web texts in the Internet Archive) back when he came up with open content. The OCL is a copyleft license very similar to CC-BY-SA, and the OPL is similar to CC-BY. Thus, it would seem that Wiley might have originally thought that CC-BY style licenses were not open content, and that open content is equivalent to copyleft (or, at least, that CC-BY-SA is "more" open). As Rob Meyers has previously written (and as most FOSS advocates would agree), adding noncommerical restritctions to a copyleft license makes it no longer copyleft. Thus, IMHO, NC licensed content would be seen as not truly open content from this perspective.
Regardless, these two fundamental principles, whether openness depends on interoperability or continued openness of the derivative works, cannot be reconciled in one license, and I doubt it would be possible to generate OER community agreement on one priority over the other. Better would be to develop an open content definition much like the Open Source Definition at OSI which does not privilege either of these principles, and in doing so, OSI avoided creating a riff in the open source communinity over over whether or not the GPL is more open source than BSD, or vice versa. Given these considerations, the OKD works well for an open content definition.
Update 6/29/2010: David Wiley has a post up indicating that he did feel that CC-BY/BSD style licenses worked better than his original, copyleft Open Content License.