Metadata review
Metadata Review
The American Antiquarian Society’s Historical Periodical Collection, hosted by EBSCO, is a database built on content-based metadata. The articles that appear on the results page are ordered by source, publication type, document type, and subjects. Of these, all are exclusively textual but for document type, which acknowledges illustrations and cartoons as distinct categories. The metadata expand, when one opens a given source, to include the original source’s location, its original publisher, and the dates of publication. In sum, the database welcomes all manner of textual questions—about who published what sorts of documents where, on what sorts of subjects. More specifically, users can ask comparative questions to a swathe of publications. Was more being written about milk during the Napoleonic Wars than had been in peacetime? A survey of fifty publications grants a much more reliable answer than a study of one. The added perk of keywords like “milk” and “nutrition” only makes the comparison on EBSCO easier. Articles by and for milk farmers can thus be set apart from poems about milk maids or Bible references to milk and honey.
What EBSCO does not let its readers ask can be answered two ways. One concerns questions that burrow deeper down the holes discussed above. It would be useful, for instance, to ask a question about Irish cartoonists. But the biographical information about artists and authors is not provided. We can ask for names, therefore, but not for other data about the people whom we trust enough to “ventriloquize,” as Tim Sherratt put it. A second obstacle to more nuanced questions concerns what DH commentators lament as the sacrifice of material to content. To its credit, and unlike other databases like LEXIS NEXIS, EBSCO includes images of all of the pages in its database. These images are of a high resolution, and permit users to magnify and even tint as they see fit. What still eludes us, though, is the experience of reading these documents. To use a prosaic example, could you walk down the street with this book in your bag? Irrespective of weight, how thick was it seen from the side? Here, the outstanding images come up one dimension short. They enable us to vet OCR transcriptions, but not to know the text as a physical object.
