[Foundation-l] Spectrum of views (was Re: Sexual Imagery on Commons: where the discussion is happening)

Aryeh Gregor Simetrical+wikilist at gmail.com
Thu May 13 19:50:36 UTC 2010


On Tue, May 11, 2010 at 7:53 PM, Anthony <wikimail at inbox.org> wrote:
> On Tue, May 11, 2010 at 12:45 PM, Aryeh Gregor <
> Simetrical+wikilist at gmail.com <Simetrical%2Bwikilist at gmail.com>> wrote:
>> [[Daniel Pearl]] does not contain an image
>> of him being beheaded (although it's what he's famous for), and
>> [[Goatse.cx]] does not contain an image of its subject matter.  Why?
>>
>
> Primarily because of copyright issues, at least with regard to the latter, I
> believe.  I used to joke about "what's next, a photo on Goatse.cx?" when
> arguing with the "when you look up X you expect a photo of X" crowd, but
> then, for a while it actually came true.

This is a standard fair-use case: it's a notable image, and our
informative article about it is not complete without a copy of the
image.  It's the same reason we can put an image on, I don't know,
<http://en.wikipedia.org/wiki/Dora_Maar_au_Chat>.  Except we don't,
because it's revolting.

On Wed, May 12, 2010 at 3:53 AM, Ray Saintonge <saintonge at telus.net> wrote:
> Each such issue will have its own spectrum of supporters and
> detractors.  It should not be our role to decide for them; we can only
> make it easier for them to make decisions consistent with their own beliefs.

Okay, fine.  Currently, people have no ability to decide whether they
want to view a particular image.  They get to see it with no warning
whether they like it or not.  I propose that if we think it's
reasonably likely they wouldn't actually want to view it, they should
be asked first, so they can decide not to see it if they prefer not to
see it.  Do you disagree?

> Not necessarily. Supermarket tabloids still sell well. Sometimes it's
> the advertisers, and not the readers who determine this.

Because tabloids aim to provide entertainment more than information.
If your goal is entertainment, then yes, more titillating content will
scare away some readers; but it will attract others, because that sort
of content does entertain people.  So if you're a tabloid, the balance
tilts more heavily toward prurient content (as well as exaggerated
content, content based on shoddy evidence or rumor, . . .).  If your
goal is to provide information, on the other hand, you don't care so
much if people are more entertained, and the balance leans more
heavily in favor of the socially conservative.  We should take our
cues from reputable publications, not tabloids.

(That said, I'm pretty sure tabloids in America don't routinely
contain even topless pictures, let alone full-frontal nudity as at
[[Human]].)

> Each project will be left to determine its own standards.  When dealing
> with Commons "relevant language" is a meaning less term.

I'm not talking about Commons.  I'm talking about the Wikipedias,
mainly, particularly the English Wikipedia.

On Thu, May 13, 2010 at 12:17 AM, Delirium <delirium at hackish.org> wrote:
> I don't actually mind this proposal, and would like it myself for a lot
> of pages. But I'm not sure naked people are actually at the top of the
> list (perhaps someone should try to determine it empirically via some
> sort of research on our readers?). If I personally were to list two
> kinds of images I would want hidden by default, it'd be: 1. spiders; and
> 2. gory medical conditions. Do I get that option? Do I get that option
> if more than x% of readers agree?

We have a perfectly good categorization system already, so any
implementation would likely permit you to blacklist any category you
like.  Of course, this depends on someone maintaining an accurate
[[Category:Spiders]] . . . this scheme doesn't work well if people
remove "redundant" categories from images.  It's also pretty ugly if
every image has fifty categories, on the other hand.  I'd hope that if
people widely blacklist particular categories (especially if some
projects do so by default), that would create the right incentives for
people to categorize things in a way that's useful for the system.



More information about the foundation-l mailing list