[This is a guest post by Lee Sorensen, the Visual Studies and Dance librarian for Duke.  He selects for the arts and manages the public-interfaces for image collections.]

Recently I was helping a student looking for articles on spolia, recycled pillars, sculpture, etc. incorporated into the art of another era.  It was a particularly common practice in the middle ages.  Though there’s much written on this topic, it’s hard to get at those results.  The reason is because there’s no agreed term for this historic phenomenon.  Searchers using our catalog don’t find those titles, not because modern catalogs are computerized or the holdings aren’t well indexed, but because of a fundamental difference in humanities thinking from science thinking.  Designers of library catalogs conceive of information in scientific-language paradigms.  In the sciences—the social but particularly the natural sciences—things have very specific names.  If you’re looking for microbial genomes, you can type in ‘eukaryotic genome’ or a specific one, such as the microsporidian Encephalitozoan cuniculi and get pretty much all the literature on that topic. But humanists neither name things schematically or coin specialized terms in their research.  Humanities scholars describe their work in a string of common language words. Take for example, the most recent scholarly article in a film journal by a well-known scholar:

Nicola Mann. “Criminalizing ‘The Hood’: The Death of the Projects in the American Visual Imagination.” Afterimage 38 no. 6 (2011).

Every one of those terms is a simple word.  Searchers for blacks represented in urban film culture would never find this important article using adjacency searching (“Google searching”).  Key-word searching was developed post-World War II by the military and government-funded science contracts.  Those people were in no way thinking about looking for eighteenth-century garden theory or early Christian concepts of soteriology.  Though we assign extensive subject headings in our cataloging, the result is always artificial.  Humanities information— humanities knowledge—is idea based, not factoid based.  Perhaps the next time some report concludes that humanities scholars don’t employ technology as much as scientists (an assertion not validated by most statistics), a better question would be to ask why technology isn’t conceived to recognize humanities information as fundamentally different.