Commons:Balancing copyright claims

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

How sure do we need to be that an image does not violate policy or copyright in order to keep it? What certainty threshold regarding policy or copyright violation do we want? Currently, our deletion policy doesn't answer that question; it basically punts and says, "In any case involving doubt, use deletion request" which leaves it to whomever decides to handle the Commons:Deletion_requests but with some guidance which can be found there (but is not actual policy):

[A]pply copyright law and Commons policy. Consensus can never trump copyright law nor can it override Commons Policy. If the closing admin is unable to say with reasonable certainty that the file can validly be kept it should be deleted in accordance with Commons' precautionary principle.
Under the rules of evidence we apply here, the burden of showing that the file can be validly hosted here lies with the uploader and anyone arguing that it should be kept.

Similarly, Commons:precautionary principle, which is policy punts on the question; it's based on "significant doubt", an extremely vague term (much like "material", as in "material adverse effect on earnings") that doesn't really have a generally agreed upon definition and doesn't really belong in policy. It's exactly that vagueness that is causing a lot of common, and problematic friction. Commons:Project_scope/Evidence is also policy; it says:

In all cases the uploader must provide appropriate evidence to demonstrate either that the file is in the public domain or that the copyright owner has released it under a suitable licence. Typically that requires at least that the source of the file be specified (along with the original source if the file is a derivative work.)* Also, the creator or copyright owner should be identified, if known or reasonably ascertainable. If there is any question, evidence may need to be supplied that the copyright owner has indeed released the file under the given licence.

Similarly, this punts on the question, as it's saying in other words that on the one hand it may be OK if there's nothing more than identified source for the image, but on the other, if there's any question, more evidence may be needed. So it's clear to this author that we don't have but we need to have a policy that defines with some precision how sure we need to be that a file does not violate policy or copyright in order to keep it.

Illustrative Examples -

  1. There's no OTRS entry for File:Vista_anterior_del_cos_humà_femaní_i_masculí.jpg (an image of a nude man and woman, with their anatomy labeled), so there's some doubt - who took the photographs? who labeled them? Did they give away their rights? We don't know.
  2. The source of File:Formidable_Sydney_Boom_(AWM_P00444-047).jpg (an old image of a WWII ship) asserts that use of the image is restricted. Who do we trust?
  3. We keep [1] despite not just a copyright claim, but an actual and still outstanding legal threat as well.

What are our options? A threshold standard should be:

  1. comprehensible
  2. consistent
  3. assume good faith
  4. strict enough that we avoid doing something legally stupid - that puts us or the foundation in harm's way
  5. ethical - balance the interests of editors, copyright holders, and the public.
  6. clear
  7. Should the value of the file be factored into the decision? Perhaps. Why stick our neck out for images that have little value to the project?

Deletion Policy Options: (Purpose: define with precision how sure we need to be that a file does not violate policy or copyright in order to keep it, while seeking the above goals)

  1. Iff the balance of the evidence indicates that the work is more likely non-free than free, or there is equally strong but conflicting evidence, it should be deleted.
  2. Iff there is reasonable suspicion of a copyright or policy violation, it should be deleted.
  3. Iff to the best of our knowledge the work is non-free it should be deleted.

To my eye, any of these three would do.

References

[edit]