I gave a talk on using FRBR group 1 entities as a framework for preservation administration at the FRBR Interest Group during the ALA Annual 2014 conference [Las Vegas, NV; Friday, June 2; 10:30 am PT]. This talk proposed a way of using FRBR to coordinate preservation effort across library networks and to clarify the goals and expected outcomes of preservation and conservation efforts. Special attention was paid to two areas where frameworks for this kind of coordination are urgently needed: managing regional print archives and understanding the role of digitization in preservation management. The slides (download ppsx) and transcript (below) should give a good sense of how this works, and an article that expands on this talk is on its way to Library Resources and Technical Services for review.
The majority of this talk focuses on aligning FRBR group 1 entities with areas of preservation administration, but there are brief remarks on two other frameworks that I’m working on: evaluating scarcity in print collections and describing the focal points for preservation interventions on information carriers. (There’s some clunky language in that last clause, I know. I find these frameworks useful, but I’m still a long way from being able to convey them with any elegance.) These models support the particular use-cases I discussed in the talk, print archives and reformatting, but they are concepts I have found useful in many other areas, and I use them both in teaching preservation and evaluating preservation strategies.
Slides: FRBR For Research in Book Retention (PowerPoint Slide Show/ppsx)
Thank you for giving me the chance to speak with you today. As I start, let me give the disclaimer that I am not a cataloger. Although I can hold forth with tedious authority on many topics, authority control is not among them. So, as I say “FRBR” over and over again today, and follow it with a verb like “does” or “says” or “implies”, please do hear my inner voice of doubt, saying “at least, I think so… maybe?”
[Next]
The framework I’ll speak about today developed over the course of several years, and across several different areas of activity. One was my work as an educator, teaching in graduate LIS programs and in various professional venues, like DPOE and the NDSR. The other was my personal need to think about what I was trying to do as a preservation librarian. The preservation effort is one of the younger parts of our profession and although it has developed many excellent, rigorous standards for specific activities, it is still light on overarching theory. And thank heavens; I don’t want to push preservation very far into the theoretical realm with this framework. This is meant to be a to help get work done and to avoid doing the same work twice.
As a teacher and working preservation librarian, I found myself coming back to four key issues: how to preserve a thing, which things to preserve, for who to keep those things available, and for how long. FRBR has been helpful to me in addressing the second and third items on that list [CLICK] which things and to some extent, for whom.
[Next]
FRBR marries analytic and descriptive bibliography in a way that complements one of the key theoretical constructs in preservation, the distinction between intellectual and artifactual value. This is always contentious in practice. All information is embodied some how – maybe in print, maybe as an electron caught in a gate in the flash drive I used to carry this presentation to Las Vegas – and it’s hard to agree on when that particular embodiment of information is important to or incidental to the information’s value.
For now, let’s acknowledge that there are some works that are strongly to one side or the other. The Gutenberg Bible isn’t interesting because of the text – plenty of sources exist for the Vulgate – but for the way it was made. In contrast, print on demand books or inexpensive re-publishings such as the Dover editions have the bare minimum of artifact required to convey a text to a reader, and the artifacts say little about the creative choices of a book designer, or the material effort of a printer, for example.
[Next]
One can see this playing out poorly over and over again in the history of libraries and their preservation efforts. Nicholson Baker’s Double Fold may be the most famous example. Libraries understood their preservation focus to be the “content,” a series of glyphs in a particular order, while Baker was concerned with the context, a particular kind of artifact that conveyed material culture and aesthetics of its own.
[Next]
It is a short step to seeing how FRBR Work-Item hierarchy is useful for clarification. I suspect that a large group of preservation controversies can be traced to a misunderstanding of which of the FRBR levels are the intended focus of a preservation effort. If so, using FRBR to detect problems and enact solutions is valuable.
Let me give a short sidebar at this point, just to nod towards a model for dealing with the “how” of preservation, the implementation of solutions
[Next]
It is helpful to think of preservation as a set of activities that repair or mitigate against damage in the materials and rendering environment of information. Any information object can have preservation risks in several ways.
- Substrate: material substance(s) of the object (e.g. paper, film, aluminum, silicon)
- Media: the material substance(s) that record information (e.g. ink, silver, ferrous alloys, electrons)
- Transport: the means of moving information from the Media to the reader (e.g. light, light, SATA bus, USB bus)
- Language: the semiotic system(s) that renders the information in a meaningful way (e.g. Latin, Cinema studies, TIFF, PDF)
[Next]
If we align those zones of preservation effort with FRBR, two things are immediately notable. One is that there are two preservation efforts that are tied to the Item. This isn’t any surprise, preservation works on things, but it’s converse is important. You can’t preserve a work directly, only by maintaining the objects that embody information and make its rendering possible. This may be a big clue to both why cataloging is critical to the preservation effort, and to what cataloging can do to further the preservation effort.
The other notable observation here is that there are preservation efforts that don’t focus on specific items. Keeping a transport option available, a VHS player for example, or keeping a language available, Aramaic or FORTRAN for example, supports the preservation of many objects.
[Next]
I won’t dwell too long here. You know better than I that FRBR is a descriptive framework that defines a hierarchy of abstract Works – Expressions of those works in distinct intellectual forms – Manifestations of those Expressions in a particular genre or medium – and instances of those Manifestations as discreet Items.
[Next]
Preservation is best understood from the Item up. The Item can be repaired (what we call conservation) or stabilized through preservation. Items can also be created in a preservation process, what we call reformatting. Those Item-centric activities in turn support preservation at higher levels in the FRBR hierarchy.
[Next]
Because preservation can only act on (or be enacted through) Items, each preservation action supports only one WEMI chain. It is also possible for success at one level of the FRBR hierarchy to coexist with failure at another, so clarifying when an Item is important in itself – the case for special collections – from when an item is important because it instantiates a manifestation is very important.
[Next]
It’s easy to confuse M and I in preservation, or said another way, FRBR can help us have some clearer definitions of preservation activities:
- M-level & up is preservation (risk management)
- I-level is conservation (corrective/preventative treatment)
- I1M1 –> I1M2 is reformatting (digitization, etc.)
A working (not PARS approved) definition of preservation in FRBR terms might say things like:
“Preservation consists in activities that improve the likelihood of ongoing access to or existence of M,E, or W”
“Conservation consists in activities that perpetuate the existence of I”
“Reformatting consists in activities that perpetuate the existence of E or W through the creation of new I and M”
Further, if Ix has unique value and cannot be exchanged with Iy
Implies that I ≡ M: “Item-Manifestation Identicality”
A working (not RBMS approved) definition for special collections could be made in FRBR terms “Collections where all I have Item-Manifestation Identicality” in a way that would help explain preservation’s particular relationship to and obligations in special collections preservation versus general or research collections preservation.
[Next]
This brings me to the issue that’s closest to my own professional obligations, the preservation of large research collections and increasingly, the coordination of print archive networks. In the charts that follow, you’ll see different elements of the IMEW chain blacked out to show preservation scenarios.
Preserving a Manifestation requires at least on Item. On the first, leftmost, branch, you can see what reformatting looks like, with a second manifestation represented by just on item, often with the loss of another from the first manifestation. This doesn’t completely resolve the Baker debate – is a manifestation preserved when any of its Items exist, or does there have to be some critical mass? – but it at least draws a boundary around the discussion that could make for a more productive conversation.
[Next]
Here we see effective E level preservation, despite the loss of some Items, and in once case, a whole manifestation. Imagine a printing of a passage of Shakespeare by the college art department, and held in only one library. Now imagine the thousand shocks that flesh and books are heir to. This is hard at the local level, but plenty of other copies of the play exist.
[Next]
In this scenario, we reverse the trend. This is the last-coy repository model. Although the great majority of duplicative Items are gone, designated instances of each Manifestation remain.
This shows have E-level preservation possible with many I-level loss scenarios, but [CLICK] it begs the question of whether E preservation requires all M, or just one M?
[Next]
Work-level preservation begs the same question – I would like to say that W preservation requires an E, not all E, that is to say that a W is not the sum of its E, but that E are derivative of a Work.
[Next]
If so, however, we may have to consider that a situation like this could represent W-level preservation, even thought it has failed as E-level preservation in 2 of 3 cases. [CLICK] Does W level preservation depend on an E, or require many or all E?
To return to the preservation repository question, you can see how important it is to have a clear mechanism for focusing preservation efforts and evaluating outcomes. I think FRBR is tremendously valuable in that.
[Next to End]
The remaining slides show a series of Item-level scenarios and their implications for print repositories.