Writing in the Sand Encoded Archival Description (EAD) in Digital Libraries

Access to manuscript collections is primarily via finding aids. There are many forms of finding aids currently in existence, such as registers, lists, guides, inventories, card catalogs, indexes, and automated systems. The Encoded Archival Description (EAD), developed by Daniel Pitti, is primarily designed for inventories and registers. Though finding aids have long existed in other formats such as Word documents, text files, hard copy, and even html, each of these forms has drawbacks which EAD was developed to overcome.

In order for computers to be able to determine specific information within a finding aid such as author, works, and items, the text containing this information needs to be descriptively tagged. And in order to build large networked databases that can support control, searching, display and navigation of collections, the tagging system needs to be standardized. Compliance with the DTD (Document Type Definition, which defines what elements you can use where and how) provides consistency for software processing, and the standardization necessary to create federated searching across institutions.

EAD encoding permits the options of searching by name, date, subject, geographic location, physical form or genre, and function. Footnote chasing is possible by linking citations within <bioghist> to the source materials. One can create a dynamic table of contents based on major elements and separate displays for different sections, for top-down searching and browsing. And by providing linking attributes at the item level, EAD offers the ability to access primary source materials if digitized.

There's a terrifically useful tag library online to help you make sense of each of the elements and how and where it is used. The Revision Description (<revisiondesc>) element provides version control for documentation of corrections, deaccessioning, or accessioning of new accruals. The Archival Reference (<archref>) element increases the amount of feasible linkage, and the establishment of interrelationships. Once the EAD is created, the content of the elements can be rendered in any number of ways to meet user needs or preferences.

Here is a sample EAD file from the University of Tennessee Special Collections Online Finding Aids. You will note that within the <eadheader> there is the identifier (<eadid>), the file description (<filedesc>) and the profile description (<profiledesc>, which gives information about how this finding aid was created (adminstrative metadata). The <frontmatter> section provides the basic information that should be presented to users first: the proper title, date, manuscript number, publisher, and encoder, with the date this EAD was encoded.

The <archdesc> section contains the remainder of the file information, beginning with an overview of the collection, including an abstract and the location of the repository where it is found, as well as rights information. Following these is the <bioghist> section, where information is given (if available) about the creators or subjects of the material; this provides context for researchers. The <scopecontent> then describes the scope of the contents, <arrangement> tells how these materials were organized for this EAD, subjects are listed in the <controlaccess> section, and then the EAD begins to describe the materials.

Some EADs only encompass a single letter; some cover hundreds of boxes of materials. The EAD allows for 15 levels of hierarchy in description. Most materials are stored in boxes, so the breakdown often includes box, then file, then folder, then the item level as encountered when leafing through the file. However, time constraints often require simple description at the collection level, with perhaps notes about the types of contents in each box. The amount of information included in EADs vary immensely in use. This is a very time consuming cataloging effort, but exceptionally useful to researchers who might never otherwise know this material exists. Providing this material via the web is a tremendous boon, for no one can visit each and every repository to discover its contents.

Here is the online display of that same file. This particular EAD was encoded to enable users to view each of the constituent files described by the finding aid. As you browse through the contents, you will find links out to the digitized photographs of the William Cox Cochran Collection. A second EAD, this one for the John Shrady Letters, demonstrates linking into text documents in a similar manner. Thus, the EAD is an invaluable tool for promoting primary source materials online. Materials that could only be accessed in person, or not at all, are now both findable and accessible via use of the Encoded Archival Description.

These are the links from the discussion above, followed by a bibliography for further reading:


Recommended listserv: EAD.

Display software used is DLXS (Digital Library Extension Service).

Girl writing in the sand

This information is provided without guarantees as to validity or completeness, particularly in light of the fact that the world of metadata in digital libraries is a world of shifting sands, constantly changing.