Development of Creative Commons licenses

["Luis Villa" <luis AT tieguy.org>]

On 3/7/07, Terry Hancock <hancock AT anansispaceworks.com> wrote:
> Hi all,
>
> Okay, several people on this list (including me) are quite vocal in
> claiming that NC terms are harmful to the mechanisms of the commons. In
> other words, the claim is that they are either "do not produce a
> commons" or "produce an ineffective commons". This is mostly based on
> theory, though.
>
> Now I want to prove it. You know, with *evidence*. ;-)
>
> No facile "well it's not 'free' so it's not in the 'commons'" word
> games. I need actual empirical evidence that NC (including NC-SA) works
> do not get reused, improved, disseminated, etc. as efficiently as By or
> By-SA works do.

[Sorry, this is too short, deserves a lot more, but I'm busy ATM :/

To get actual evidence, you're going to have to define your terms much
more precisely than you have; your current definitions are very vague.
This is unfortunately very hard :) No one has particularly good
metrics on community size/code dissemination; certainly nothing
involving any sort of scientifically valid control. The least bad
studies I've seen in the open source space involve statistical
analysis of sourceforge projects. You could perhaps measure re-use of
samples in ccmixter for a similar study?

The obvious examples, if you're OK with evidence-by-anecdote are all
around- open source projects with commercial involvement are the norm
because they work. There are some variants of open source software
licenses with NC terms, but they aren't popular. Some of that has to
do with GPL network effects, of course, but that isn't the whole
story.

I think the most powerful reason for this is that in an NC commons,
only the original creator has a commercial motivation to contribute to
the commons. Every other contributor must be motivated by some other,
often less powerful, motivation. In a liberally licensed world, many,
many people have a potential strong commercial motivation to
contribute to the commons. The more people with a potentially strong
motivation to contribute, the more people who will contribute. (Mark
Lemley discusses some related concepts in his "Ex Ante Versus Ex Post
Justifications For Intellectual Property"; I recommend reading it.)

I disagree with Greg's suggestion that this is psychological; if
people didn't want to contribute under restrictive licenses then the
BSD and Apache licenses would be much more commong than they currently
are. They aren't, so we have to look for non-psychological reasons.

I might add that I think that share-alike licenses are generally more
successful than more permissive licenses (compare BSD and the various
proprietary BSD forks vs. Linux) because they solve the prisoner's
dilemma and guarantee that all players always cooperate instead of
defect. (I think this particular thought is worth a whole thesis from
someone with a game theory or organization theory background.)

Hope those thoughts help- sorry that they aren't as organized/clear as I'd 
like-

Luis