Orphan works (and Google Book Search)

I talked to a group of graduate students in history at George Mason yesterday about emerging copyright issues. We spent some time on orphan works, a subject of great interest to all the historians of my acquaintance. The anti-injunction provisions in the Copyright Office's proposal cover instances "where the infringer has prepared or commenced preparation of a derivative work that recasts, transforms or adapts the infringed work with a significant amount of the infringer’s expression."

Especially for historians, who want to show you their evidence (or excerpts thereof), it's important to know whether direct quotation -- or, in the case of images particularly, complete reproduction -- can constitute recasting, transformation or adaptation. A fair number of cases suggest that the answer is yes, if the work in which the quote/copy is embedded has a different purpose than the original. What if a photo is used to illustrate a point, such as "Many Jewish immigrant families were proud of their involvement with socialist causes" (an actual example based on a photo of my husband's family holding up their favorite socialist paper, which was reprinted in a recently published book), but the photo is not otherwise analyzed or criticized in the text? I'd still say that falls within the definition of a transformative derivative work.

The proposed definition adds "with a significant amount of the infringer's expression" to the definition of a derivative work, which suggests a look at what the copier has added and not what s/he's copied, a rule that diverges from the test for infringement. What about a book like Wisconsin Death Trip, a compilation of death-related photos and news reports from the end of the nineteenth century -- does selection and arrangement count as the infringer's expression? I would think so, given that it does for purposes of the copyrightability of compilation works.

But this additional "significant amount" requirement appears to (deliberately?) exclude users like Google (who'd also be barred from using this new provision by the requirement of reasonable search for the copyright owner, since Google is relying on opt-out), who would not seem to be contributing original expression in assembling a searchable collection of documents. The Copyright Office report discusses large-scale users like libraries, but only contemplates them taking down a digitized work whose owner appears (thus avoiding any liability at all) or being able to continue distributing a book or movie that includes the work subject to the reasonable compensation requirement. The report does not consider the possibility that being part of a database is itself a type of transformation, which is obviously an important issue right now.

So, should copyright minimalists attempt to get "significant amount" taken out of the proposal? The idea that adding a short introduction shouldn't be enough to get someone into the reasonable-compensation-only category, which is the other obvious point of that requirement, is persuasive enough, but it might already be covered by the requirement that the orphan work be incorporated into a derivative work. Without sufficient transformation, maybe it's just a reproduction rather than a derivative work.

On the other hand, if the proposal is adopted with this wording, maybe it's more persuasive to say that database uses generally create derivative works (sometimes as fair use) since the implication will be that recasting, transforming and adapting can occur without the addition of significant new expression. Of course the Tasini and National Geographic cases have already gone into this problem in some detail, but I wonder if there are other similar database situations lurking out there.