Development of Creative Commons licenses

[James Grimmelmann <james AT grimmelmann.net>]

Terry Hancock wrote:
> Sorry for the long response time -- I went to a conference in between.
> But I think this is an important question...
> 
> James Grimmelmann wrote:
>> Terry Hancock wrote:
>>> However, it seems to be quite vague on the subject of how the license
>>> affects the use of sampled works. Note that it permits you to "create
>>> and reproduce" derivative works, but it doesn't say what terms can apply
>>> to them. For example, it would seem intuitive that one could re-license
>>> sampled works under other licenses (if not, the effect would be
>>> remarkably confining).
>> Why would there be any restrictions on the terms one could use to
>> relicense?  I don't see any hook in the language that could be read to
>> imply restrictions on licensing terms.  The quoted language grants
>> permission to engage in every single one of the exclusive rights of a
>> copyright holder that might apply to a derivative work.  Sure, it
>> includes some restrictions in 3(a)(i) and (3)(a)(ii), but those
>> restrictions involve the transformativeness of the derivative, not what
>> conditions (or lack) the licensee puts on the derivative works.
> 
> So, let's assume that I am a free software game developer and I want to
> make a game with some nifty sound effects.
> 
> I find an audio track with a lot of cool sounds on it, which is licensed
> CC-Sampling.
> 
> I want to release my source code under the GPL, including resources
> which are essential to the game, which I deem includes the sound effects.
> 
> So, the first thing I do is create a *derivative* of the original
> "Sampling" track, which extracts the sounds I want and re-arranges them
> for use as effects. Possibly I even just snip them into individual
> tracks and collect them in a directory, or I have them in one FLAC or
> Ogg file -- but it shouldn't matter.
> 
> I believe that thus far I have obeyed the "transformativeness"
> requirement of the CC-Sampling. So, now I am completely free of
> restrictions on the derived work, and can re-license it as GPL. (?!)
> 
> This then becomes part of my "source code" for the game, which will also
> be GPL, and so there's no license conflict.
> 
> When I ask to admit this package into Debian, they will be happy with
> the GPL licensing. However, I already know that they will raise hell
> about the CC-Sampling to GPL re-licensing step above (the (?!) step).
> Yet, you say it's legit.
> 
> Who's right?

I am.  :-)

Debian has very good reasons to be very careful about closely checking
the terms of material that has in some way been incorporated into
software included in Debian.  The transformativeness of the sampling
into the game, provided that it really is highly transformative,
suffices to launder away any restrictions on how it can subsequently be
used.  Of course, what counts as "transformative" is not something that
we have extensive experience with, so I can understand erring on the
side of caution for something like Debian.  But if we pass the
transformative test, everything subsequent is kosher.

Everything, that is, except for the advertising clause.

> Okay, assuming that this is not a problem, let me suggest a legal puzzle
> that's related (and perhaps sheds light on this one, but I'm not sure):

I think it does . . .

> 
> Suppose that I transform the original track by completely hashing it
> into tiny pieces, and rearranged them completely, but in fact, all of
> the original work is still there (either intentionally, or by accident).
> 
> Since the resulting work is GPL, another person (not me) can (in
> principle) reassemble the GPL sounds into a work that is either
> identical to the original or at least "not sufficiently transformative"
> from it.
> 
> This could even happen if I didn't copy all of the original work, but
> just some of it, and other GPL authors copied the rest. A later author
> could legally combine our GPL sources to get all the pieces of the
> entire work. So none of us would've been in danger of violating the
> "transformativeness" requirement in Sampling.
> 
> Now, I have a theory about this. I think that the author who transforms
> our GPL work into the original is the only one violating the license,
> and that he does this in the same way as anyone does who re-creates a
> copyright work from public-domain materials.

Exactly.  It's the same way that I can violate the copyright in _Infinte
Jest_ by "re-creating" it from the public-domain alphabet.

> 
> In other words, I could just as easy have said: I receive a novel under
> CC-Sampling, hash it up into words, and then make new stuff from the
> words, which I then release GPL. Someone receiving my work can in
> principle rebuild the original by combining those words. In fact, of
> course, they can't really do that without copying the original work --
> there aren't enough clues in the words chosen to guide the production of
> the original novel. So, in fact, that person isn't simply freely
> combining *my* work, he's copying from the original work (which is not
> GPL'd).

Yes.  And if he combined them into a passable facsimile of a *different*
work, he'd be copying that work, not the original CC-Sampling one.

> 
> If my analysis is true, then this is really quite interesting because it
> shows how GPL work can be legally derived from CC-Sampling or
> CC-Sampling+ work, even though those licenses are generally regarded as
> among CC's most restrictive licenses.

Yes.  If you're highly transformative, you're free and clear.

> In fact, there are things I can do with them that I can't even do with
> CC-By or CC-By-SA licenses.

To some extent.  Note that something which is a de minimis copy, or
which is a fair use, or which copies only uncopyrightable ideas, can
also be "derived" from a CC work but will nonetheless be free and clear
even of the CC license.

> This would be a good thing for the commons, but I'm not sure if it's
> really true. I already know that opinion leaders on debian-legal think
> it is NOT true, for example:
> 
>> Ben Finney wrote:
>>> Terry Hancock wrote:
>>>>>> [CC Sampling Plus 1.0 license clause re. derivative works]
>>>>> Trivially non-free. The DFSG freedom to create a derived work is
>>>>> the freedom to create *any* derived work, not some limited subset
>>>>> as defined in this license.
>>>> I agree.
>>>>
>>>> However, I do have a question about this ...
>>>>
>>>> If I myself have sampled from a CC sampling plus work in producing a
>>>> new work, I am under the impression that I'm pretty free to
>>>> re-license as I please.
>>> I don't see anything in the license that grants that freedom. The only
>>> part that allows redistribution is quite explicit that it must be
>>> under the terms of this license, and not "sublicense":
>>>
>>> =====
>>> 3. License Grant & Restrictions. Subject to the terms and conditions
>>>    of this License, Licensor hereby grants You a worldwide,
>>>    royalty-free, non-exclusive, perpetual (for the duration of the
>>>    applicable copyright) license to exercise the rights in the Work as
>>>    stated below on the conditions as stated below:
>>>
>>>    [...]
>>>    e. Attribution and Notice.
>>>
>>>       [...]
>>>          ii. You may distribute, publicly display, publicly perform or
>>>          publicly digitally perform the Work only under the terms of
>>>                                              ^^^^^^^^^^^^^^^^^^^^^^^
>>>          this License, and You must include a copy of, or the Uniform
>>>          ^^^^^^^^^^^^
>>>          Resource Identifier for, this License with every copy or
>>>          phonorecord of the Work or Derivative Work You distribute,
>>>          publicly display, publicly perform, or publicly digitally
>>>          perform. You may not offer or impose any terms on the Work
>>>          that alter or restrict the terms of this License or the
>>>          recipients' exercise of the rights granted hereunder. You may
>>>                                                                ^^^^^^^
>>>          not sublicense the Work. You must keep intact all notices
>>>          ^^^^^^^^^^^^^^^^^^^^^^^^
>>>          that refer to this License and to the disclaimer of
>>>          warranties. You may not distribute, publicly display,
>>>          publicly perform, or publicly digitally perform the Work with
>>>          any technological measures that control access of use of the
>>>          Work in a manner inconsistent with the terms of this
>>>          License. The above applies to the Work as incorporated in a
>>>          Collective Work, but this does not require the Collective
>>>          Work apart from the Work itself to be made subject to the
>>>          terms of this License. Upon notice from any Licensor You
>>>          must, to the extent practicable, remove from the Derivative
>>>          Work or Collective Work any reference to such Licensor or the
>>>          Original Author, as requested.
>>> =====
>>>
> 
> I think the trick above is that he's conflating "Work" with "Derivative
> Work".

In the context, when I make my transformed copy and someone else use it,
I am no longer "sublicensing" the Work as such.  But I think it's
legitimate to note this ambiguity, and a cleanup of the language
whenever the next revision pass comes along might be worth considering.

> There's more problematic language further up, though:
> 
>>    c. Prohibition on advertising. All advertising and promotional uses
>>    are excluded from the above rights, except for advertisement and
>>    promotion of the Derivative Work(s) that You are creating from the
>>    Work and Yourself as the author thereof.
> 
> This attempts to place a restraint on derivative works -- presumeably
> even those which fall within the 3a "recreativity" clause. I think that
> is a serious failing.

I do, too.  But if you go over to the cc-sampling list and bring up this
point, folks over there will be VEHEMENT on this point.  They come from
a different community, with different (albeit overlapping) values, and
they see freedom and integrity in very different terms.  So the
restriction stands.

> It's unclear to me whether the attribution and other requirements
> continue to apply.
> 
> Or is there something that I'm missing here?

Not in the same way that the advertising clause continues to apply.


So, I need to retract part of what I said initially.  The no-advertising
clause does create a restriction on how the work can be relicensed.
That's the kind textual hook I was talking about.  There is not,
however, a general restriction on how derivative works can be
relicensed.  The default is permit.

James