Development of Creative Commons licenses

[Gisle Hannemyr <gisle AT ifi.uio.no>]

On 2013-10-18 22:38, Sarah Pearson wrote:
> Hi everyone,
>
> Before we lock the legal code and publish the license suite, we have one
> final change to propose for the two ND licenses only.
>
> * We propose changing the ND licenses to make it clear that licensees may
>   create adaptations  so long as those adaptations are not shared
>   publicly.*
>
> This change would make it unambiguous that adapting licensed works for
> personal or private use is permitted under any of the 6 licenses in the CC
> suite. Because of exceptions and limitations to copyright, personal
> adaptations should already fall outside the scope of the licenses in most
> cases.  However, there are reasonable arguments that this is not always
> true everywhere in the world.
>
> This change to the ND licenses would harmonize the
> result worldwide, making it unequivocal that such private uses are
> permitted.
>
> As a matter of policy, we believe this is the right result.

I believe it will create some new problems.  Read on.

> One benefit of this change is clarifying that text and data mining are
> permitted under all 6 CC licenses. Text and data mining activity should
> fall outside of the scope of copyright and the other licensed rights, but
> because such analysis often involves creating adapted material in the
> process, there is some uncertainty with regard to the licenses that do not
> permit adaptations.

Unfortunately, this solves a non-existing problem and introduces
a major one.

First, text and data mining is probably permitted anyway.  After all,
scientific researchers and search engine companies around the globe
has done this for ages with "all rights reserved"-texts and very few
has objected (and none have prevailed).  In other words, you're
trying to fix something that isn't broken.

The major problem introduced is the string: "so long as those
adaptations are not shared publicly".

This implies:

1) that CC thinks that the material that is produced as part of
   text and data mining *are* "adaptions"; and
2) that this material can't be shared.

I, for one, do not believe that the material that is produced as
part of text and data mining *are* adaptions in a legal sense.
I do not really understand why CC think that is a good idea to
assume otherwise.

> We gave serious consideration to pursuing an exception
> within the licenses that would specifically permit text and data mining,
> but this has proven to be especially difficult since the type of activity
> is wide-ranging and the field is continually changing. The definition of
> which activities constitute text and data mining and which do not is simply
> not clear-cut, and attempting to create one risks enshrining a soon-to-be
> outdated definition in the 4.0 licenses.

If you're worried about this, then drop the idea entirely.

While "enshrining a soon-to-be outdated definition" of text and data
mining may be bad, assuming that text and data mining may be
"adaptions" are much, much worse if one wants to promote freedom.

> This particular proposal is generic, which is one of its strengths: while
> it is not specific to text and data mining, it enables these activities to
> the extent there was any uncertainty about whether an adaptation was
> created during the process.

Since temporary or intermediary copies that exists within a computer
as part of some process is *not* considered "works" in a legal sense,
I don't understand why this is stated to be an "uncertainty".  It
should be stated clearly that such copies are *not* "adaptions".

> However, it doesn't solve the problem entirely.
> Under BY-ND, licensees still may not share adaptations. While the output of
> mining will rarely, if ever, constitute an adaptation,  this could
> prevent sharing of the adapted dataset used for mining, for example.

I suggest you look beyond CC.  *If* data-sets used for mining is
to be considered adaptions, then we can forget about mining data-sets
generated from "All-rights-reserved"-texts.  Some of us actually
care a lot about mining texts that are *not* under CC-licenses.
Gaining the right to mine ND-licensed text is a very minor win if
this means we lose the right to mine ARR-licensed texts.

> Also, where
> the content was under BY-NC-ND, licensees still may not create adaptations
> privately for commercial purposes.

I think in the area of copyright, the terms "private" and "commercial"
is mutually exclusive.  If you're making money on something, you're not
in the private domain, even if you're paying audience is a single
individual.  So this is another thing you need not worry about.

> For these reasons and many others, CC
> continues to urge licensors not to use the ND licenses for scientific data,
> and we do not intend this change to make the licenses more suitable for
> this use. However, for data that is made available under the ND licenses,
> this proposed change would remove some possible barriers to text and data
> mining.

I think the effect of this on many licensors will tip them in the 
direction of CC BY-ND. Many academics are very strongly attracted
to the BY-ND-licenses (instead of just BY).  They're not  against
text and data mining, but they are frightened that someone may
create an adaption that will "deface" their precious text.  Telling
them that they can have their text protected against changes, and
*still* be part of the OA text and data mining ecology if they use
CC BY-ND, will just push them in the direction of CC BY-ND.

> Because adapting works for private use often falls outside the scope of the
> licensed rights and because such private use does not meaningfully affect
> the rights holder, we see this change as substantively minor. But to the
> extent it resolves uncertainty over whether such uses are permitted, we see
> it as a clear improvement, with little to argue against it.
>
> We will be posting updates to the html of the legal code to reflect this
> change shortly. We look forward to your feedback.

It is not minor.  It has huge damage potential.

For instance, let ut imagine you go ahead with this, and the 
"all-rights-reserved" publishing industry discovers that the CC, of
all, can be called upon to testify in court that in its opinion, a
text that is crawled and then transformed into a data-set for the
purposes of text or data mining may be regarded an "adaption".
--
- gisle hannemyr [ gisle{at}hannemyr.no - http://folk.uio.no/gisle/ ]
========================================================================
    "Don't follow leaders // Watch the parkin' meters" - Bob Dylan