Free images and model releases

classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|

Free images and model releases

Gregory Maxwell
It appears that it is widely known and accepted within the Wikimedia
projects that the law in some countries makes images with identifiable
persons whom did not consent to the photography problematic.

What seems to have been far less discussed is that if we are going to
worry about things which are merely civilly actionable that the
problem isn't just limited to "some countries".

In the US persons who have their clearly identifiable image used for
commercial purposes for which they did not consent have little
difficulty getting a judgement in their favor.  Since this is the
case, can we really regard an image with identifiable people to be as
free as an image without identifiable people or an image with
identifiable people and a suitable release?

What really brought my attention to this matter is an image recently
proposed as a featured image on enwiki:
http://en.wikipedia.org/wiki/Image:Sunbathe_Buttocks.jpg ... Now it
may be that the image was created with the consent of the subjects, I
don't intend to criticize this particular image...   But really, lets
consider such an image with identifiable subjects who didn't know the
image had been taken.  What if she were to return to her dorm room and
find the Wikipedia article on sunbathing stapled to it?   Were the
image taken without her consent it's quite possible that she'd be
pissed, and I believe that she'd be justified. There is no reason that
illustrations like this in Wikipedia can't be ones which are created
with the subjects consent.  So even ignoring any possible legal issues
with such images, I think we're breaking good ethics to use images
like this without the subjects consent.

Here is a US centric article discussing the matter:
http://www.danheller.com/model-release.html

On enwiki we already have established practices which absolutely place
free images over non-free images. For example, given two images which
serve the same purpose one which is available under a free license and
one which we could only try to claim fair use for, we will always
replace the fairuse image with the free image and we will not permit a
free image to be replaced by a fair use image... Even if the free
image is of lower quality.

I'm considering proposing an addition to that practice, saying we
should always prefer free licensed images which have no identifiable
people or for which we have appropriate model releases over images
with identifiable people and without releases.

No image deletion runs, no prohibitions against uploads, just a
preference. In the majority of subjects on Wikipedia, any identifiable
people are fairly incidental to the actual subject of the
photograph.... Encouraging people to avoid getting identifiable shots
of people where possible would probably be good for overall image
quality even ignoring the potential legal and ethical implications,
because clearly identifiable faces can be distracting.

Thoughts?  I'm especially interested in knowing how the French
Wikipedia handles identifiable images of people.
_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l
Reply | Threaded
Open this post in threaded view
|

Re: Free images and model releases

Erik Moeller-2
Gregory-

As a counter-example, take a look at:
http://upload.wikimedia.org/wikipedia/commons/c/c5/Karachi_-_Pakistan-market.jpg

This image was (successfully) nominated for Featured Picture status;
during the discussion, someone objected on the grounds that the
photographed person did not give their consent. Such objections, or
acting upon them, strike me as overzealous. A fantastic photo like this
should not be removed or replaced unless there is a personal objection
from the subject.

We should not discriminate on grounds of quality. But I agree that a
person who finds themselves in a Wikipedia article illustrating a
subject where they might have reasonable grounds to find such use
_personally embarrassing_ should give their consent. This includes
obvious matters like human sexuality, but also images like:
http://en.wikipedia.org/wiki/Image:Sexynose.jpg

I would not object to the inclusion of a note that such (and only such)
images may otherwise be replaced or removed, provided the photographer
is given advance opportunity to obtain a model release.

 > Encouraging people to avoid getting identifiable shots
> of people where possible would probably be good for overall image
> quality even ignoring the potential legal and ethical implications,
> because clearly identifiable faces can be distracting.

I don't agree. I prefer an encyclopedia with a human face. Have you
noticed how in Hollywood movies, directors use every trick in the book
to cover frontal nudity, by placing objects or sheets or whatever is
available before the private parts? (Stanley Kubrick wonderfully
parodied this practice in "A Clockwork Orange".) I'd rather not have a
similar culture in Wikipedia with regard to faces, with hundreds of
photos of people turning their face to the side, having their eyes
covered by black bars, or putting their hands over their head.

Erik
_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l
Reply | Threaded
Open this post in threaded view
|

Re: Free images and model releases

Gregory Maxwell
On 3/17/06, Erik Moeller <[hidden email]> wrote:

> Gregory-
>
> As a counter-example, take a look at:
> http://upload.wikimedia.org/wikipedia/commons/c/c5/Karachi_-_Pakistan-market.jpg
>
> This image was (successfully) nominated for Featured Picture status;
> during the discussion, someone objected on the grounds that the
> photographed person did not give their consent. Such objections, or
> acting upon them, strike me as overzealous. A fantastic photo like this
> should not be removed or replaced unless there is a personal objection
> from the subject.

I agree, for the most part, but would you not agree that given this
image and one almost identical but with a solid release from the
subject that the image with the release would be more free simply
because any objection from the subject would have far less teeth?  ...
Content is not free content unless we could, in theory, use it against
the wishes of others.

But obviously there are many situations where making a hard demand
would be unnecessary and foolish... Which is why I didn't propose a
war against all such images.   I think we should encourage people to
help make sure their submissions are as free as possible.  It's fairly
normal for commercial photographers to walk around with a stack of
model release forms and I think that we'd be better off if some
Wikipedia photographers were doing the same.

> We should not discriminate on grounds of quality. But I agree that a
> person who finds themselves in a Wikipedia article illustrating a
> subject where they might have reasonable grounds to find such use
> _personally embarrassing_ should give their consent. This includes
> obvious matters like human sexuality, but also images like:
> http://en.wikipedia.org/wiki/Image:Sexynose.jpg
>
> I would not object to the inclusion of a note that such (and only such)
> images may otherwise be replaced or removed, provided the photographer
> is given advance opportunity to obtain a model release.

So, how to set the criteria?  It sounds like what you're propose is
that we only replace images  if their use on Wikipedia more likely
than average to generate complaint. Thats hard to convert into
instructions which are simple to understand... and it makes
assumptions about what all uses of the images will be.

Often the only time to get permission is when the image is taken, so
making a use dependant rule means that we'll often have to throw out
the first illustration for a personally embarrassing use, just because
no one didn't consider that it might be used in an embarrassing way.

There are, I think, two interconnected issues here: The first is the
ethical use of identifyable images of people, and the second is the
implications the impact of civil (and in some places, criminal)
liability of using an identifiable image without permission on the
freeness of our content.

It sounds like you and I are in complete agreement on the first part:
We should attempt to replace potentially embarrassing images which are
identifiable and where we don't know that we have the consent of the
subject.

On the second matter, I'm not sure we have agreement yet. This is less
an ethical concern and more a concern about maximizing the freedom of
our content.

If we don't have some rule that explains that images with identifiable
people and without releases will be replaced, we'll have no incentive
for people to ever go through the trouble of getting releases. The
freedom of our work as a whole will be reduced as a result.

It was my thought that since this is a lower risk issue that we can
avoid any sort of prohibition. But it sounds like you think that
replacement is too aggressive in the general case.

>  > Encouraging people to avoid getting identifiable shots
> > of people where possible would probably be good for overall image
> > quality even ignoring the potential legal and ethical implications,
> > because clearly identifiable faces can be distracting.
>
> I don't agree. I prefer an encyclopedia with a human face. Have you
> noticed how in Hollywood movies, directors use every trick in the book
> to cover frontal nudity, by placing objects or sheets or whatever is
> available before the private parts? (Stanley Kubrick wonderfully
> parodied this practice in "A Clockwork Orange".) I'd rather not have a
> similar culture in Wikipedia with regard to faces, with hundreds of
> photos of people turning their face to the side, having their eyes
> covered by black bars, or putting their hands over their head.

Eh, That wasn't at all what I was suggesting.

We often have images of devices or processes which are in no way
enhanced by people in the shot... but the images have them anyways
because that was what was in front of the camera when the shutter
fired or it was the easiest way to hold the object.  In such cases,
the extraneous human presence is potentially distracting.

By all means, if the shot is enhanced by having humans in it... we
should have them. If it's enhanced by having humans in it in a way
which makes them identifiable, then we should do so...  But if they
are, I think we should ask our photographers to get releases...   If
they don't,  we can still use the image until someone objects.
_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l
Reply | Threaded
Open this post in threaded view
|

Re: Free images and model releases

Erik Moeller-2
Gregory Maxwell:

> I agree, for the most part, but would you not agree that given this
> image and one almost identical but with a solid release from the
> subject that the image with the release would be more free simply
> because any objection from the subject would have far less teeth?  ...

These are slight variations on "freedom", similar to a reproduction of
public domain 2D artwork without a source vs. one made by a Wikipedian.
In both cases I would consider image quality the more important
criterion. Only if the quality is truly equal I would then distinguish
based on these criteria. In most such cases we could also simply allow
both images to be used.

 > I think we should encourage people to
> help make sure their submissions are as free as possible.

Certainly Wikimedians should be _encouraged_ to obtain (formal or
informal) model release whenever possible, just as they should be
encouraged to provide good metadata, cite sources, use dual-licensing, etc.

> So, how to set the criteria?  It sounds like what you're propose is
> that we only replace images  if their use on Wikipedia more likely
> than average to generate complaint. Thats hard to convert into
> instructions which are simple to understand... and it makes
> assumptions about what all uses of the images will be.

It's more about the content of the image than about its use. An image
which would be typically considered a private or intimate shot is
clearly different from a photo taken e.g. in a public stret. Rather than
having very precise guidelines on this, I suggest an addition to the
policy which refers to common sense and consensus, e.g.:

    When a photo shows identifiable subjects in a setting which could
    be considered private or intimate according to the predominant
    cultural norms where the photo was taken, such photos should be
    replaced with ones where model release has been obtained. The more
    private or intimate the setting, the stronger is the case to
    completely remove the image if there is no model release.

    Use  common sense, seek consensus, and give the uploader an
    opportunity to obtain model release before removing or replacing an
    image for this reason.

Note the reference to cultural norms. The Dutch have just introduced a
video to educate immigration applicants about "Dutch culture"; it
includes scenes such as topless sunbathing, a homosexual couple and nude
hippies (yay, hippies!). I wouldn't at all be surprised if not all the
persons shown in this way gave permission. Similarly, a pot smoker in
the US may be less happy about being photographed than one in the
Netherlands. ;-)

Yes, you have to make an educated guess about these matters, and that
guess might be very off-base. The idea is that you mostly act when
someone complains.

> We often have images of devices or processes which are in no way
> enhanced by people in the shot...

Absolutely. Or the typical tourist shots with the whole family posing in
front of the medieval castle. Certainly here it is a matter of
encyclopedic quality to get the damn people out of the picture. ;-)

Erik
_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l
Reply | Threaded
Open this post in threaded view
|

Re: Free images and model releases

Anthony DiPierro
In reply to this post by Gregory Maxwell
On 3/17/06, Gregory Maxwell <[hidden email]> wrote:
> In the US persons who have their clearly identifiable image used for
> commercial purposes for which they did not consent have little
> difficulty getting a judgement in their favor.  Since this is the
> case, can we really regard an image with identifiable people to be as
> free as an image without identifiable people or an image with
> identifiable people and a suitable release?

I'm fairly sure that statement actually isn't the case.  You *can* (in
the US) use an image of a famous person for commercial purposes, so
long as that purpose is also educational (such as in an encyclopedia).

Anthony
_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l
Reply | Threaded
Open this post in threaded view
|

Re: Free images and model releases

Anthony DiPierro
In reply to this post by Gregory Maxwell
On 3/17/06, Gregory Maxwell <[hidden email]> wrote:

> What really brought my attention to this matter is an image recently
> proposed as a featured image on enwiki:
> http://en.wikipedia.org/wiki/Image:Sunbathe_Buttocks.jpg ... Now it
> may be that the image was created with the consent of the subjects, I
> don't intend to criticize this particular image...   But really, lets
> consider such an image with identifiable subjects who didn't know the
> image had been taken.  What if she were to return to her dorm room and
> find the Wikipedia article on sunbathing stapled to it?   Were the
> image taken without her consent it's quite possible that she'd be
> pissed, and I believe that she'd be justified. There is no reason that
> illustrations like this in Wikipedia can't be ones which are created
> with the subjects consent.  So even ignoring any possible legal issues
> with such images, I think we're breaking good ethics to use images
> like this without the subjects consent.
>
In a case like the one you give, I agree.  Wikipedia should respect
people's right to privacy (which applies in cases where the identity
of the person is not important for any educational/newsworthy
purpose).

> I'm considering proposing an addition to that practice, saying we
> should always prefer free licensed images which have no identifiable
> people or for which we have appropriate model releases over images
> with identifiable people and without releases.
>
I think this should only apply to non-famous individuals, and that it
should be a strict rule, not just a preference.

> No image deletion runs, no prohibitions against uploads, just a
> preference. In the majority of subjects on Wikipedia, any identifiable
> people are fairly incidental to the actual subject of the
> photograph.... Encouraging people to avoid getting identifiable shots
> of people where possible would probably be good for overall image
> quality even ignoring the potential legal and ethical implications,
> because clearly identifiable faces can be distracting.
>
> Thoughts?  I'm especially interested in knowing how the French
> Wikipedia handles identifiable images of people.
_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l
Reply | Threaded
Open this post in threaded view
|

Re: Free images and model releases

Ray Saintonge
Anthony DiPierro wrote:

>On 3/17/06, Gregory Maxwell <[hidden email]> wrote:
>  
>
>>What really brought my attention to this matter is an image recently
>>proposed as a featured image on enwiki:
>>http://en.wikipedia.org/wiki/Image:Sunbathe_Buttocks.jpg ... Now it
>>may be that the image was created with the consent of the subjects, I
>>don't intend to criticize this particular image...   But really, lets
>>consider such an image with identifiable subjects who didn't know the
>>image had been taken.  What if she were to return to her dorm room and
>>find the Wikipedia article on sunbathing stapled to it?   Were the
>>image taken without her consent it's quite possible that she'd be
>>pissed, and I believe that she'd be justified. There is no reason that
>>illustrations like this in Wikipedia can't be ones which are created
>>with the subjects consent.  So even ignoring any possible legal issues
>>with such images, I think we're breaking good ethics to use images
>>like this without the subjects consent.
>>    
>>
>In a case like the one you give, I agree.  Wikipedia should respect
>people's right to privacy (which applies in cases where the identity
>of the person is not important for any educational/newsworthy
>purpose).
>
What is unclear about the picture is whether the subject is
identifiable.  She has a towel (or something) over her head which would
hide her identity.

Ec

_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l
Reply | Threaded
Open this post in threaded view
|

Re: Free images and model releases

Gregory Maxwell
On 3/20/06, Ray Saintonge <[hidden email]> wrote:

> What is unclear about the picture is whether the subject is
> identifiable.  She has a towel (or something) over her head which would
> hide her identity.

The woman in the back's face isn't obscured.
_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l
Reply | Threaded
Open this post in threaded view
|

Re: Free images and model releases

Anthony DiPierro
On 3/20/06, Gregory Maxwell <[hidden email]> wrote:
> On 3/20/06, Ray Saintonge <[hidden email]> wrote:
>
> > What is unclear about the picture is whether the subject is
> > identifiable.  She has a towel (or something) over her head which would
> > hide her identity.
>
> The woman in the back's face isn't obscured.

Yes, that's the woman I was talking about.

Anthony
_______________________________________________
foundation-l mailing list
[hidden email]
http://mail.wikipedia.org/mailman/listinfo/foundation-l