On Wednesday, I gave a talk and displayed a number of files that had been produced by others. They were all linked to a dedicated page containing all the metadata necessary for proper attribution of the source, but that page was not shown during the talk, so the audience only saw the images or videos and some rudimentary metadata. The talk was video recorded and is supposed to be released under a Creative Commons license. How would proper attribution look like in this case?
The problem of attribution stacking has already been touched upon briefly in the description of the image featured on December 14 but let’s consider it in some more detail now:
Suppose an artist decorated a wall with hand-painted representations of CC BY-licensed photos made by a number of others. The wall has a plate attached to it that contains all the names of those “others” and declares that the wall painting itself is CC BY.
In front of the wall, there is a tree that is of interest to me (e.g. because it shows a rare pathology), so I am taking a photo. As it happens, parts of the wall are on that photo but the plate is not. If I then want to publish this image in my open lab notebook or as part of a scientific paper (both licensed CC BY too, for simplicity), would I have to attribute the artist and all the “others” too? And would it make a difference if the plate is actually visible in my photo but its inscription is not readable? Or whether the artist had painted or photocopied (or even just projected) the original images onto the wall? And what if I am not aware of the plate and its contents because my attention is solely directed to the tree?
Such multi-level reuse is perhaps not to be expected to be common for most kinds of scholarly works, but for certain kinds of data, it may well be the norm. Having the data under a CC0/PD license would ensure that reuse on any n-th level remains legal, but what would reasonable attribution look like at those levels? I am not aware of a good overview on the matter, so I am referring, for the moment, to a video recording of a talk given by MacKenzie Smith in December 2010, in which she discusses the problem from a data provenance perspective.