Development of Creative Commons licenses

[Mike Linksvayer <ml AT creativecommons.org>]

On Mon, Sep 14, 2009 at 12:46 PM, Gregory Maxwell <gmaxwell AT gmail.com> wrote:
> Regarding http://creativecommons.org/weblog/entry/17127
>
> I don't know that "strong agreement on the commerciality of certain
> use cases" is very well supported by the NC survey data.
>
> There are a couple of different data points which suggest that
> agreement on the meaning of the license itself, or on the meaning of
> "non-commercial", is on the order of 90% at best. (For example,
> agreement on 'making money' as a criterion, the statement that the
> license text matches their expectations, or the number of users who
> changed their definition after the survey.)
>
> While 90% is surely a significant supermajority, licenses can cause
> harm if they do not result in a very clear understanding. Even a few
> hours of attorney time can easily outweigh the benefit gained by using
> a work and statutory damages start at hundreds of dollars. Because of
> magnitude of the harm when there is a misunderstanding is non-trivial
> I think the potential for misunderstanding must be very low for a
> license to achieve its intended purpose.

Sounds logical, but I don't see evidence. Given weak understanding of
copyright, the potential for misunderstanding of even the clearest
license starts out very high. Yet we see few suits and many efforts
that rely on public licenses, including ones with high potential for
misunderstanding (Artistic License 1.0 and GFDL come to mind), the
latter including CC NC licenses (probably most significantly
OpenCourseWare)., and especially in the CC NC case, lots of
non-"project" uses that don't involve suits.

> Consider this kind of overwhelming majority: 90% think X, 10% think
> !X. 18% of random pairings of creators and users will have conflicting
> understandings of what the license means, or almost 1 in 5.
>
> Even if an author is in the 90% majority—if that author has only 7
> users, there is a 52% chance of at least one misunderstanding. With 29
> users there is a 95% chance, and with 44 users there is a 99% chance.
> If the author is in the minority, a 99% chance of having a
> misunderstanding over allowable uses happens with only two users.
>
> If an author only expects to have 7 users, why not just consider each
> case on its own merits and permit uses individually?

How is an author to predict number of users (0 would be the best
prediction for many)? The author may not wish to consider individual
cases. The author may wish for uses they don't know about to be
allowed under some broad if fuzzy parameters. The author may have
expressive reasons.

> If an author
> isn't concerned about license violations why don't they use the most
> permissive licenses?

Because they are concerned about license violations by "big
corporations" -- just a guess, subject for further study.

> Most authors reported that they have been
> contacted. I am frequently contacted requesting permission for
> activities which are unequivocally permitted by the licenses I use.
>
> With only 90% agreeing on a particular criterion… I can only guess
> that there aren't masses of disputes because people don't enforce,
> even when they don't agree. People tolerate the 'misuse', or they are
> not paying attention; sometimes they give up publishing. Nice guys
> don't sue (or issue takedowns).

Off the top of my head, some other possibilities:
- Use is rare
- Licensors are not aware of most uses
- Licensors don't agree that many uses are noncommercial, but they do
want to allow many marginal cases
- Use does not match professed understanding due to uncertainty
generally or concerning what licensor might think
- There are masses of disputes and they almost never go to court, some
proportion resolved by gentle request, some by takedown, some further
threat, some by disk failure eliminating contested use, some by death
of licensor and licensor's lineage ... some unresolved

> But ifone off negotiations and "nice
> guys don't sue" are what really keeps the system working, why bother
> having a public license?
>
> It would have been nice if the survey had some calibration points: do
> people understand and agree on what CC-BY and CC-BY-ND require?  I
> strongly believe that they do, at least after minimal exposure to
> those licenses, but I have no data. The only disagreement I've seen
> for CC-BY is confusion created by the 'human readable' "must attribute
> the work in the manner specified by the author" being understood
> literally by authors instead of as "must attribute the work in the
> manner specified in the license text", resulting in claims of license
> violation because attribution was not provided in 40pt blinking text.
> Though "anecdote" isn't a synonym for "data", so I really don't know.
> (And… it's hard to tell from experience because most problems are
> created by people completely ignoring their copyright-related
> obligations)

Good ideas for future research.

> At some level of disagreement about the meaning of a license, using it
> will do more harm than good by inspiring disagreements. Where is that
> threshold?

I don't see any evidence it has been reached by any public license in use.

There are probably other thresholds reached first, indeed which may
have been reached, such as underuse due to adoption of relatively
restrictive licensing.  Would love to see future research model and
test such things.

> As far as I can tell, personal use is by far more popular than than
> other forms (i.e. remix, verbatim redistribution of someone else's
> work); while this doesn't appear to be addressed directly in the
> survey, most respondents classified themselves primarily as users.
> Wouldn't a "private, personal use only + contact me for other options"
> cover the overwhelming majority of the need for NC, while closing off
> many opportunities for misunderstanding?

"Private,  personal use only" obtains just about nothing, and even if
it did obtain something, CC wouldn't want to contribute to a culture
of thinking one needs permission to remix for private, personal use,
ie to use a computer. Carving out verbatim distribution, including
filesharing, and doing something other than remixing and never
distributing would probably get one close to NC.

Mike