SAA 2015: making born-digital records accessible

Terminal Tower, Public Square, Cleveland, Ohio, 2015-08-21. Until 1991, Terminal Tower was the tallest building in the city of Cleveland and the state of Ohio. I have loved this building as long as I can remember.

SAA 2015 is in full swing. Today, I sat in on two sessions -- Arrangement and Description and Access for Digital Archives (session 401) and Out of the Frying Pan and Into the Reading Room (session 507) -- that focused on on providing access to born-digital materials. I was tardy in arriving to the first and had to leave the second in order to travel to an offsite meeting, so what follows is a partial listing of things I found interesting or useful.

• One repository is providing access to a born-digital body of materials that is subject to varying copyright and donor restrictions by loading copies of the files onto a laptop that is not connected to any network and has disabled USB ports. This approach isn't perfect, but archivists shouldn't wait for perfection to start making their holdings accessible. (Moreover, as another archivist pointed out, this approach requires minimal IT support.) 
• No two collections are the same, and processing is always time-consuming. Another repository assesses each collection of born-digital materials for quality of data, authenticity of data, complexity of the access restrictions associated with copyright and donor stipulations, and anticipated level of use. Records that contain high quality and authentic data, lack complicated access restrictions, and will likely receive high use receive more intensive processing than those that don't meet these criteria. 
•  The amount of processing work we do will likely vary. One institution has some born-digital collections that consist of flat groupings of items and some collections that consist of files arranged in directory structures. In other instances, collections are mixtures of analog and digital items, and the archives wants the arrangement of the digital materials to correspond to that of the analog. 
•  We don't yet have a firm sense of what our users want. Some of our users are comfortable with doing keyword or other types of searches, and others are accustomed to box-and-folder hierarchies. We may discover that we need to try to meet the needs of both groups. 
• Access solutions are varied, constantly changing, and have a way of emerging in response to pressing user requests. We need to remain flexible and mindful of the fact that solutions that work at one institution might not work at another. 
• We need to publicize our born-digital holdings, and we need to make sure that colleagues who do reference work are comfortable working with these materials and highlight their existence to researchers when appropriate.

The question of making restricted materials available online also came up, and one presenter recommended making use of the redaction functionality being incorporated into BitCurator and informing end users of their responsibilities regarding inappropriate disclosure of information that may be subject to various restrictions. The latter approach was also explored quite extensively in a Thursday afternoon pop-up session that centered on issues raised by recent events at the University of Oregon, and the discussion included making access contingent upon entering into formal, online agreements.

I find this an intriguing approach, but most most government archives will likely be very slow to embrace it. Some state open records laws specify that records creators and archives cannot impose limitations on the use of information that is disclosed in response to freedom of information requests; if a record contains restricted information, the creating agency or the archivist must redact it prior to disclosing it. Moreover, governments tend to be risk-averse -- sometimes excessively, and sometimes with good reason. However, I can envision some scenarios in which government archives might well adopt this approach; using a click-through agreement to highlight the presence of records potentially covered by copyright isn't quite the same thing as hoping a researcher will abide by an agreement prohibiting disclosure of information found within psychiatric case files.

Finally, in response to a question concerning whether we should embed all of the metadata we're creating as we work with digital materials into our finding aids, one of the panelists in session 401 said something that's been on my mind for some time: we need to start thinking about moving away from document-based finding aids. I like Encoded Archival Description (and well-crafted MARC records make me feel as if there is an inner logic and order to the world), but it's high time we stopped thinking of archival description solely in terms of "fast paper."