Apologies for my late blogging!
The four policies I chose to review are the Dartmouth College Library’s Digital Preservation Policy, the Public Record Office of Northern Ireland (PRONI)’s Digital Preservation Strategy, Illinois Digital Environment for Access to Learning and Scholarship (IDEALS)’s Digital Preservation Policy, and Rhizome at the New Museum’s Digital Preservation Practices and the Rhizome Artbase. Although these policies come from a range of archives/institutional repositories, libraries, and museums, there were definitely a lot of commonalities.
Digital preservation policies should align with institutional collection policies and mission – identify the types of items most important to the institution. These policies need to establish what types of material the institution will collect and preserve, but also be flexible enough to for appropriate on-the-ground decision-making between policy review periods. Stanford’s web archiving policy focuses on at-risk content while also making sure that the policy supports other collection policies and strengths, and prioritizes what is likely to be useful to Stanford’s researcher base.
So, a good digital preservation policy would establish what the institution’s mission or responsibility for digital collections is (such as in PRONI’s policy, which reads an original remit about paper-based collections to also apply to born-digital and digitized material); explain challenges to preservation and/or risks (such as Dartmouth and Rhizome’s policies); define audiences/users for digital materials being selected and preserved; establish collecting and preserving priorities (all five policies I looked at do this); delineate principles behind preservation (Dartmouth’s policy explains life cycle management and lists several resources they have access to, like Portico, LOCKSS, and HathiTrust); and sets a regular schedule or deadline for the policy to be reviewed.
There were some interesting differences I noted in these policy documents as well: IDEALS’s policy does not list specific formats, PRONI’s mentions that they have a list of accepted file formats but that is not included in the policy, and Dartmouth’s lists their preferred formats within the policy itself. Rhizome’s policy reads more like a whitepaper than a policy document, and goes into more depth on the multiple different directions future actions could take.
Discussion Questions
-
- Stanford’s web archiving policy is for an institution with high staffing levels and adequate funding. While all of the major points still apply to smaller institutions, how do you scale this type of robust, well-defined collections policy to understaffed or all volunteer-run organizations, such as the ones many of us in class are working with?
- Rimkus, Padilla, Popp, & Martin’s analysis of file format policies across ARL institutions brought up that repository managers place more trust in file formats that originate from library reformatting programs. Is some of this built-in trust because many librarians come from humanities backgrounds? Could increasing diversity in library staff’s backgrounds (i.e. more people with media production, art, design, or programming backgrounds) change the level of confidence repository managers and policy creators have in other formats?
I found the IDEALS digital preservation policy to be one of the most useful ones from the list. Not only is it concise, broad, and easy to understand… they also address the challenges to digital preservation that a lot of our organizations face. “Creation of a Development Environment” is a struggle that is almost an understatement for many of these smaller institutions… having the resources or time to dedicate to preservation planning is impossible. I think this will help a lot of us to consider what is in reach and feasible as we draft our own policies… in some cases just having them *think* about preservation is a great step forward for their institution.
In response to your first question, I think part of the way we are all scaling our policies is through the specificity of all of our organizations. Many of the small organizations are very specific in their collecting policy, so there are fewer ways that they branch out. Stanford is doing a lot and succeeding through their staff; we just need to help our organizations do enough to preserve their small holdings. The UK Data Archive, PURR (at Purdue) and other large digital repositories have long and detailed policies because they have huge amounts of data (comparatively) . Our organizations won’t need that level of detail, so the scale is set by the amount of material.
Responding to your second question, I think an increased diversity in backgrounds will absolutely change confidence levels in certain formats, particularly for the media production, art, design, and programming crowd that you cite. That is my background, and it has been interesting for me to note the dissonance between some institutions’ stated goals of providing merely functional access and their more technologically purist choices for preferred formats.
For example, *everyone* seems to love TIFF as an image format. And it’s tough to argue with that position from a standpoint of preserving the maximum amount of data. They’re not wrong; that *is* the most visual information you can possibly save. And it is well documented and nonproprietary. But here’s the thing… it’s unlikely anyone will ever be able to use all that information. A DSLR can see roughly 14EV of dynamic range. A good monitor will give you roughly 10EV, and a good print is drastically less at roughly 6.5EV. You can underexpose the image so that the monitor displays maximum highlight detail, or you can overexpose the image so that the monitor displays maximum shadow detail, but you’re not going to get the maximum detail on both ends of that spectrum without some heavy handed trickery. And that trickery tends to look unnatural to the human eye (do a Google Image search for “HDR” and you’ll find plenty of bad examples). I would argue that such transformations essentially destroy the photographer’s vision for how the image should look. Are there potentially forensic reasons for wanting such extra data? Sure. But I think that’s starting to drift beyond the realm of simple functional access to the artist’s original image.
In comparison, JPEGs are technically lossy, but if you save them at the maximum quality setting you have essentially only thrown out information that your monitor and printer have no hope of using, and the photographer didn’t want you to have anyway. They’ve already made their decisions as to what information to show you and what to hide from you. Is JPEG an ideal solution? No. But to me, it’s adequate functional access.
Your experience is such a good example! I think many places continue to use TIFF because it’s a recognized standard, and moving away from that would require deep knowledge of what your users and organization would need in the future AND of how specific formats may or may not meet those needs. If an archivist, especially one who is early career and probably wearing many hats, wants to argue to move away from something seen as “industry standard” they’d likely need enough experience with the formats creators are using to understand what could be adequate functional access and what could be “overkill” for the organization’s needs.