Group items matching

in title, tags, annotations or url

The ACLS Commission on Cyberinfrastructure’s report points to five manifestations of digital scholarship: collection building, tools to support collection building, tools to support analysis, using tools and collections to produce “new intellectual products,” and authoring tools. 

Tara McPherson, the editor of Vectors, offered her own “Typology of Digital Humanities”: •    The Computing Humanities: focused on building tools, infrastructure, standards and collections, e.g. The Blake Archive •    The Blogging Humanities: networked, peer-to-peer, e.g. crooked timber •    The Multimodal Humanities: “bring together databases, scholarly tools, networked writing, and peer-to-peer commentary while also leveraging the potential of the visual and aural media that so dominate contemporary life,” e.g. Vectors

My initial diagram of digital scholarship pictured single-headed arrows linking different approaches to digital scholarship; my revised diagram looks more like spaghetti, with arrows going all over the place.  Theories inform collection building; the process of blogging helps to shape an argument; how a scholar wants to communicate an idea influences what tools are selected and how they are used.

I looked at 5 categories: archival resources as well as primary and secondary books and journals.   I found that with the exception of archival materials, over 90% of the materials I cited in my bibliography are in a digital format.  However, only about 83% of primary resources and 37% of the secondary materials are available as full text.  If you want to do use text analysis tools on 19th century American novels or 20th century articles from major humanities journals, you’re in luck, but the other stuff is trickier because of copyright constraints.

I found that there were some scanning errors with Google Books, but not as many as I expected. I wished that Google Books provided full text rather than PDF files of its public domain content, as do Open Content Alliance and Making of America (and EAF, if you just download the HTML).  I had to convert Google’s PDF files to Adobe Tagged Text XML and got disappointing results.  The OCR quality for Open Content Alliance was better, but words were not joined across line breaks, reducing accuracy.  With multi-volume works, neither Open Content Alliance nor Google Books provided very good metadata.

To make it easier for researchers to discover relevant tools, I teamed up with 5 other librarians to launch the Digital Research Tools, or DiRT, wiki at the end of May.