That was the topic discussed recently by OCLC Research Library Partners metadata managers, initiated by Roxanne Missingham of Australian National University and Stephen Hearn of the University of Minnesota. Archival collections are in many ways the jewels in the crown of collections as they are unique research resources that provide insights into the world across many centuries, the fodder for creating new research. Creating visibility for these collections reaps significant benefits for both researchers and libraries/archives. Archives are, however, complex, and present different metadata issues compared to traditional library collections. As institutions turn to ArchiveSpace and other content management systems to provide infrastructures for structured archival metadata, various issues are emerging.
OCLC Research Library Partners use a variety of standards to describe archival collections: Describing Archives: A Content Standard (DACS), Encoded Archival Description (EAD) and MARC are common in the United States; Rules for Archival Description (RAD) in Canada; and the International Council on Archives’ (ICA) International Standard Archival Description (ISAD) in Australia, the United Kingdom and other Western nations. Discovery is through local portals and regional networks, and aggregations such as the Social Networks and Archival Context (SNAC), ArchiveGrid, Trove (in Australia) and WorldCat.
Some highlights from the discussions:
- Identifiers: Many would like to assign identifiers to parts of a collection in addition to the overall collection. The Smithsonian has been using ARKs (Archival Resource Key), assigning the same stem to parts of a collection providing a type of hierarchical approach to constructing identifiers. The number of personal names in archival collections can be so large that most are uncontrolled and without identifiers. George Washington University is experimenting with automatically generating entities from finding aids with good results.
- Improving exposure to archival collections: RAMP (Remixing Archival Metadata Project), developed by the University of Miami Libraries, is a tool that extracts biographical and historical data from EAD finding aids and then generates enhanced authority records and publishes the content as Wikipedia pages. (OCLC Research sponsored a webinar in 2014, “Beyond EAD: Tools for Creating and Editing EAC-CPF Records and ‘Remixing’ Archival Metadata” featuring a demonstration of RAMP.) The Wikipedia Library’s 1Lib1Ref campaigns encourage librarians to add one reference to an existing Wikipedia article, which could be to an archival collection. OCLC recently won a Knight News Challenge to promote collaboration between public libraries and Wikipedia, which could also include more exposure to archival collections. (See 16 June 2016 news release and WebJunction’s project page, Wikipedia + Libraries: Better Together.)
- Differences between archivists and librarians: Archives have had more autonomy than libraries within their institutions because they have unique collections with their own population of users. Some institutions have integrated archival processing within technical services, but most maintain a separate unit. Even so, a significant shift to metadata standardization from “artisanal archival approaches” has been occurring. Archivists do not have the tradition of creating authority records and sharing identifiers for the same entity as is common among librarians. Archivists tend to use the fullest form of a name based on the information found in collections, while librarians focus on “preferred” form found in publications. Some differences arise from the technology used; for example, ArchiveSpace does not connect authority records to collection descriptions, a major hindrance to data integration.
Among the questions raised:
- How can archivists and librarians best integrate their data and name authority practices when their administrative units are separate?
- The contextual information that archivists provide for personal and organizational entities would enrich the information provided in authority files – how could they be linked?
- Do different needs arise when describing born-digital archival materials? Physical “extent” does not apply, while the original “carrier” may be considered a crucial element. The work underway by the OCLC Research Web Archiving Metadata Working Group led by my colleagues Jackie Dooley and Dennis Massie may address some of these differences.
Karen Smith-Yoshimura, senior program officer, topics related to creating and managing metadata with a focus on large research libraries and multilingual requirements. Karen retired from OCLC November 2020.
Alex: I consulted with Mark Custer of Yale, the source of the statement about ArchiveSpace, asking him for the more details you requested. His response:
1. ArchivesSpace does not include a biographical note that’s attached to an “Agent” (or archival creator /entity) record when it serializes an EAD file for the finding aid description. This has been a feature request from the community to further enable the separation (when most efficient) of description about archival creators from the description about records. Max Evans had a nice article about this back in the 1980s: http://americanarchivist.org/doi/abs/10.17723/aarc.49.3.0862585240520721 You definitely can use ArchivesSpace this way, ever since the first release; the only issues about that are with how that data is then exposed elsewhere, like with ArchiveGrid — since, A) the contextual notes about the Agents wouldn’t be included with the finding aid by ASpace right now automatically, and B) since ArchiveGrid doesn’t ingest EAC-CPF records, and, most importantly, C) since ArchivesSpace isn’t very good at ingesting or exporting EAC-CPF records just yet. All that said, ArchivesSpace has the foundation to support authority-based description; it just needs a few tweaks to how it handles that data to make it a practical choice, in my opinion. But even right now there’s nothing stopping anyone from including and linking authority descriptions in ArchivesSpace, and it’s definitely something that anyone using the system is already doing.
2. ArchivesSpace only permits one “authority ID” to be attached with any “Agent” record. I just brought that one up since we were currently embarking on a project to store a lot more authority URIs in our ArchivesSpace database, and it would be nice, I think, to include both the LCNAF URI as well as other IDs, like a Wikidata URI when those also exist.
So, I’d say that not only does ArchivesSpace permit the connection of authority records to collection description, it actually requires the linking of agent records (which could include details far beyond a normal authority record — such as narrative description, a family tree, etc. — but might also just include a name string!) to all aspects of material descriptions (from accession records, collection records, series-level records, item-level records, etc.). But, like most systems, there could be improvements with how that data is exposed outside of ArchivesSpace, which is why I’m especially glad that the SNAC project envisions working with ArchivesSpace and other archival management systems in the future. Just exactly how those systems will work together has still yet to be determined / implemented, but that integration could have an interesting impact on how processing archivists do their day-to-day work.
I don’t have any more information about GWU’s name authorities work.
Thanks for posting this! I’m not entirely clear what you mean by “ArchiveSpace does not connect authority records with to collection descriptions” – our experience with the application is that its Agents module allows for connecting local representations of authorized entities (complete with authority identifiers to reference against the authority file) to archival description. Could you go into more detail about this?
Do you have any more information on GWU’s name authorities work?