Skip to Content.
Sympa Menu

cc-licenses - Re: [cc-licenses] Requirements for ...: toward a practicable definition of "author name" ?

cc-licenses AT lists.ibiblio.org

Subject: Development of Creative Commons licenses

List archive

Chronological Thread  
  • From: Terry Hancock <hancock AT anansispaceworks.com>
  • To: Development of Creative Commons licenses <cc-licenses AT lists.ibiblio.org>
  • Subject: Re: [cc-licenses] Requirements for ...: toward a practicable definition of "author name" ?
  • Date: Wed, 26 Dec 2007 10:14:29 -0600

Benjamin Pineau wrote:
> It's obvious that in order to improve and clarify attribution,
> "author name" (as in "the name to be credited") should be
> clearly defined, if possible. Users should be allowed to credit the
> "author(s) name(s)" as they received them attached to the work
> (or a subset of those names, subset to be defined too),
> everything else I can think off would be impracticable.

IMHO, "Allowed" yes (and indeed I think this is how things sit, though I
agree that a more explicit statement of this could be good). "Required"
could be a problem, though (which I know you didn't ask for, but it's
something the language must avoid doing, so it's a practical problem).

There is a large body of sloppily-accredited CC-By-SA work out there. I
prefer to trace the attribution to a real name if I can do so with only
a few hops and if the author has expressed no obvious intent to remain
anonymous. I feel this represents better practice (at least, it's what
I'd want people to do for me).

Username accreditations -- which are frequently automatic and out of the
user's control on community content like Flickr or Wikipedia -- are
often ambiguous. How many Joe51's are there in the world, I wonder? On
how many sites. No, in order to be more explicit, if I cannot find a
real name or a clear attribution request, I use the username plus an
identifier for the site -- something like "digitante@wikipedia" (which
is me, actually, though I made sure to put my real name on my userpage).

On the other hand, if I have not made an explicit attribution request, I
consider it fair play to use my username.

The real problem is that the CC-By-SA specifies no "default" behavior.
It assumes in the "legal code" that the author has provided an explicit
attribution statement. However, common practice doesn't follow suit and
many sites simply do not provide an adequate mechanism -- neither for
allowing it nor for encouraging it, even if it is possible.

ISTM that there are two possible ways to interpret the existing
language, and an improved license would clarify which is the case:

1) The "correct" attribution is needed to whatever degree the licensee
can find it out (which assumes some standard amount of effort people
should be expected to put in).

2) No explicit attribution means that attribution has been waived.

Number 1 is kind of hard to support and gives us horrible fuzziness of
interpretation (How much effort is expected? How do we measure it? How
do we know if it's been done?). All stuff you'd have to go to court to
prove, and of course, if you have to go to court, the license has failed
to do its job.

Number 2 is much simpler. However, it would probably come as an awfully
nasty shock to a lot of content producers out there! Furthermore, it
would surely put sites like Flickr and Wikipedia in legal jeopardy since
it would arguably render their services fraudulent -- they promise CC
licensing, but if they technically force you to waive your attribution
rights, then that's probably actionable.

So, it'd be an improvement if the license says flat out that you lose
attribution rights if you don't explicitly invoke them.

However, it would probably be far better than that to establish some
kind of limited "required effort" to ascertain proper attribution. That
would be a little tricky to write, and perhaps awkwardly specific. But
since most CC licensing activities take place on the Web, there is
probably reasonable cause to give some specifics at least for that case.

> For practical reasons, a few things should be allowed to the
> recipient/user of the work, that a proper definition of
> "author name" can impact. Here is a list of practical problems
> a good definition of "author name" may prevent.

You've omitted some other problems. I read recently that the *reason*
why some sites place limits on what sort of attribution you can apply is
to avoid cases like these:

1) An "author name" might be obviously fraudulent and malicious. For
example, one might publish a full public recantation of Free Software as
an evil, communist plot, funded by Al Quaeda, to undermine Western
Civilization, and sign it "Richard Stallman". This might violate other
laws than copyright of course -- libel or fraud, perhaps (IANAL!).

2) An "author name" might actually be just a commercial advertisement
(Or is this okay? I don't think the CC license clearly exclude this).

3) An "author name" might be itself an offensive, violence-inciting
political or racial statement. I'll leave serious offenses to your
imagination, but mild examples like "ImpeachBush" or "USACanGoToHell" or
"EatThisBillGates" are probably all over the Internet.

Under various countries' laws, I imagine these could possibly be used in
lawsuits against the content site, not just the originator of the
content. Not to mention it would poison the community. So, if these
kinds of abuses are not excluded by the license, then these kinds of
policies are understandable.

> In my opinion (what's yours ?), the recipient/user of a work:
>
> * Should credit work's authors with the "names" he received
> attached to the work (be them real names or pseudonyms).

"Attached" is an extremely vague word on the internet. Although embedded
tagging technology is slowly improving, most (non textual) content
doesn't have any attribution directly in the file, but only an
attribution associated with it on a web page (at best), or perhaps in a
blanket statement on a web site.

> * You should not have to contact authors to ask them how they
> would like to be credited

This I agree with. If you have to get permission to use the work, you've
lost the essence of a free license.

> * If an author decide to change his name/pseudonym (like in an
> user account rename on wikipedia) after you received the
> work, you should not be expected to find this out and fix your
> copy to reflect all those changes aftermath. This wouldn't
> be practicable.

Yes. The attribution at the time you received it ought to be good enough.

However, please note that this directly contradicts one of the promises
in the CC licenses, which is that you can have your name removed from a
work if you don't want to be associated with it.

OTOH, the solution there is pretty obvious -- *they* must contact *you*
in order to exercise that right.

> * You should be allowed to transliterate an author name to an
> other writing system if technically necessary.

As this seems like pure common sense to me, I suspect it would already
be true. I think that transliterating and encoding changes are probably
not considered significant changes, as long as the relevant information
is conveyed.

> * There are indeed many other considerations that matter for
> accreditation, like "how many authors need to be credited"
> (ie. for a massively collaborative work), but I think that's
> an other matter, unrelated to the "author name" definition.

Massively collaborative work is an interesting problem. It's one thing
to roll credits on a 2-hr feature film. But if you make a 1 minute video
clip based on the feature film, must you still roll the entire credits
(possibly as much as 10 minutes of film?!). Or is it enough to provide a
full credits file with it?

More important, I think, is the issue of hierarchical credit. For
example, it's pretty much a convention that writers and directors of
films are credited more prominently than, say, caterers or even foley
artists. Yet, as written, I believe the CC license would
strictly-speaking require equal credit (not fair credit) to everyone
involved. So if I were to re-do Elephants Dream with new foley sound and
voices (but the rest of the film is intact), how should the credits be
modified?

(As a matter of fact, Elephants Dream says only that the credits must be
run in full when showing the film -- which suggests (possibly) that they
are leaving accreditation for derivatives up to your discretion).

> A simple solution would be to require a recipient/user to credit
> "authors names" as he received them attached to the work (or
> a selected subset of name he received).
> ie. for wikipedia, that would mean the names as they appear in
> the article's history (or a subset of those names).

I think the "subset" idea is a problem. Who picks this subset? I can't
imagine that there are standard rules that could be applied in any
consistent way to all forms of content. And enumerating all the known
cases would turn the license into a tome!

It might be better to provide an "attribution by reference" option -- in
other words, allow the full credits to be contained in a separate file,
retaining the URL or other reference to the credits in the work itself.
The credits within the work might then be left up to individual discretion.

I suspect that the sorts of abuses that would crop up with this kind of
approach would be self-correcting. If the true credits file must be
maintained, then there is an easy way for anyone who cares to verify the
displayed credits for "reasonableness". Misrepresentations would then
reflect very badly on misappropriating authors, leading to social rather
than legal solutions to the problem (as well as straightforward
correction of the displayed credits).

As for the credits files themselves -- throwing around plain text files
with credits, even if they contain hundreds of names, isn't that much of
a burden. They're obviously trivial online. On printed books or media
packaging, they might represent a couple of pages of small print.

Anyway, that's my thoughts on the subject.

Cheers,
Terry

--
Terry Hancock (hancock AT AnansiSpaceworks.com)
Anansi Spaceworks http://www.AnansiSpaceworks.com





Archive powered by MHonArc 2.6.24.

Top of Page