This week’s readings firmly demonstrate that the role of policy in forming the approach to digital preservation is more than just curatorial navel-gazing. The “what”, “why”, and “who for” questions have dramatic implications for the choices institutions make. Spelling these out in writing can serve as a vital touchstone for institutions as they wrestle with the minutia of individual decisions, because it’s easy to lose sight of the bigger picture when you’re down in the weeds. It’s occasionally time well spent to pause and ask “wait… why are we doing this again”?
Sheldon kicks off this week’s readings with a broad examination of the trends and history of digital preservation policies in libraries, archives, and museums. One of the most notable points she makes is the large gap in the number of policies created by libraries and archives as opposed to museums. Libraries in particular led the charge in creation of digital records as early as the 1960’s with the creation of MARC records, and this early start carried libraries and archives – who share a functionally similar service model with libraries – forward into the digital age with somewhat greater momentum than museums. As Sheldon points out, both libraries and archives are tasked with processing and providing access at scale. Museums on the other hand have to wrestle with preservation concerns that are far less cut and dry. Artifacts in their care are often exceedingly unique, sometimes consisting of both analog and digital properties that must both be preserved in ways that don’t scale in the same reproducible manner that libraries and archives are able to employ with simpler objects. As such, Sheldon suggests that the dearth of policy documents from museums could potentially stem from the fear of forcing a “one-size-fits-all” approach into a practice that more often requires flexibility.
The four policy documents referenced in Sheldon’s study that I chose to examine were from the National Archives of Australia, the UK Parliamentary Archives, the Dartmouth Library, and HathiTrust Digital Library. I started with Australia simply because I’ve been given the impression at various points in my academic career that Australia really has their stuff together and is often in the vanguard of curatorial approach. As such it was interesting to note how succinct their policy was. Not that I felt they ignored important areas, but they were more brief and to the point than the UK Parliamentary Archives. At first pass it was notable to me that NAA expressly valued functionality over bit preservation. I don’t know why I was surprised to see such a decisive stance on one side of the spectrum, but I was. The only thing I thought was odd about NAA’s policy was that they paused in the middle of it to list their guiding principles. That seemed like it would have made more sense in the beginning.
Enforcing my impression that Australia is an industry leader it did not escape my attention that the UK Parliamentary Archives policy directly referenced NAA’s policy, and indeed much of the structure and content of Australia’s approach was reflected in the UK policy document. The Parliamentary Archives went to a great deal more trouble in explaining the problems and potential costs of failure in their document than the NAA did in theirs, but the influence of NAA’s approach was clearly evident in areas such as the emphasis on functional access over bit preservation and the sections on collaboration with external entities. One point of divergence that I appreciated in the UK document was the discussion of archival considerations at the point of creation in an asset’s lifecycle. I found this focus on asset lifecycle to be a very pragmatic lens through which to view the preservation concerns of digital objects.
I was interested to compare UK and Australian approaches to an American institution, so I chose The Dartmouth Library next. It was interesting to see that they also used life cycle management as a lens for outlining their approach. The rest of the document struck me as being somewhat less rigorous in terms of outlining things such as information security and data integrity, but it thoroughly covered the organization’s scope, even specifically calling out items it would not manage such as subscription-based resources.
Lastly, I was interested to see what HathiTrust had as a policy document since they had such a drastically different purpose than the previous three institutions. Not surprisingly, their very brief document reflects an equally narrow business model. And I use the term “business model” intentionally since this entirely online institution is largely subject to the same usage dynamics that any other online business is subject to. A concise and differentiated purpose is what sets them apart. As such, their preservation policy is focused on a very specific use case requiring the preservation of original appearance and functional access via OCR-enabled search capability. Beyond this they don’t get a whole lot more detailed than saying “we adhere to industry best practices and standards”. The whole thing is so brief it’s almost funny. It’s sort of like HathiTrust just stood up after the UK Parliamentary Archives spoke for an hour and said “uh, yeah… ditto.”
Rimkus et. al. wrap up the readings with a look at some of the technological forces that shape our policy decisions. In particular they note the role that preservation software has played in guiding (perhaps even dictating?) the archival format choices for institutions. I’m particularly interested in the tension here between openness of format and adoption rates. Owens tells us that “adoption is the core driver for continued accessibility.” Rimkus et. al. agree with this as well, noting that even a proprietary format can be considered a reliable bet in terms of longevity if embraced by a broad community. This makes me wonder to what extent the reverse can be true. Can openness by itself support embracing a format as a matter of policy even when adoption levels of that more open format are middling, or even low? Using my own project as an example, I have seen parts of our organization stick with DNG vs. native RAW formats for photography simply because DNG has more open documentation. The low adoption rate for DNG has been somewhat ignored. Native RAW formats on the other hand, are everywhere. And nearly every RAW converter works with them. This is not true of DNG. Even though these native RAW formats are highly proprietary and not even remotely open in nature, their ubiquity makes them a safer bet even in today’s world. I’m having a hard time coming up with a scenario in which DNG’s openness wins the day unless still photography itself completely collapses as a prevalent representational format. Which I concede, is possible.
Has anyone else encountered any such conundrums with no clearly good options? How would it shape the policy recommendations you might make for your institution?