Board Thread:Consensus Track/@comment-34115-20180618163709/@comment-34115-20180618175022

We have a valid fair use argument for images.


 * We are attributing the source mentioning who actually owns it, and even exactly what series and episode/chapter and page it comes from.
 * They are small clips from a larger work and do not replace the source material.
 * We are providing commentary on the image, using an image in order to describe what is in it. Instead of using the image in an unrelated context.
 * We cannot replace a copyrighted image of a character with a non-copyrighted image because we are depicting the character in a tv series/manga and an original work we have rights to would not be an accurate depiction of that character.

However we don't have a valid fair use argument for databook text:
 * We are not bothering to attribute the author that has the original copyright to the databooks and are using it in a context that unless you investigate deeply looks as if we are claiming we own the rights to the text.
 * We are providing a database of knowledge similar to databooks. In copying the text we copy almost all of the unique expression in that databook, instead of just a little bit of it. And reading that text from our articles is sufficient enough that it replaces the source material and you do not need to buy a databook.
 * We are not providing commentary on the databook, we are providing commentary on the anime/manga and copying the databook's commentary on the anime/manga.
 * It is entirely possible for us to write text in our own words and still cover the same facts and still accurately depict the commentary we are making.

If you look over the factors on how fair use is decided, a judge would likely uphold a fair use claim for our use of images clipped from the anime and manga, but throw out a fair use claim for use of translated databook texts.