For Preservation Week 2011, I wrote a series of posts for the UCLA Library Preservation Department Weblog. Those are linked here and archived after the fold.
- Monday, April 25: Tangible Records
- Tuesday, April 26: Projects versus Progress
- Wednesday, April 27: Data!
- Thursday, April 28: Preservation Education
- Friday, April 29:Preservation Administrators and Conservators
1. Monday, April 25: Tangible Records
“Tangible records” is not the most poetic pair of words, but I think it’s a crucial idea and it’s no accident that I’m starting off with this idea for Monday in Preservation Week. Preservation can get complex very quickly, as one finds out all the known constraints on any particular preservation plan and identifies all the information that’s lacking. Remembering to come back to the question of what is actually recorded and how that recording occurs is a critical check and a great guide to priorities.
This is something that we often lose sight of in 2011, when any library worth the name provides extensive digital resources. It’s easy to think of digital resources as intangible, having no mass or presence in the physical realm. A great deal of digital preservation work depends on working with digital objects in this abstract manner.
However, all of that is for naught unless bits are reliably written to media with a know life expectancy. The cloud is still made up of hard drives, even if their specific locations and identities are hidden behind a, well, cloud. A good test of any digital preservation solution is to find out if the clouds can be parted and the question of tangibility can be answered. There has to be someone who can say where the data centers are, what the protocol is for distributing data across them, and how the data center deals with the life expectancy of their media. Absent that, the rest of the project falls apart.
When the tangible thing is the complete and authentic thing, then preservation and conservation are one and the same concept; intellectual and artifactual value are inseparable. This is the common case in museums, and to my way of thinking, this is one of the core issues that distinguishes museums from libraries and archives. It’s no great surprise that museums have very little of what librarians or archivists would recognize as preservation administration, even as they give a great deal of attention to conservation.
Another place where tangibility is a useful touchstone is in distinguishing between preservation and re-creation. The conceptual problems that arise here are Proustian, in an ironically literal way. The book he wrote can be preserved , easy enough. The pastries, should they step out from fiction, could be preserved, as well. Cookies have a tangible, even a delicious form, albeit with substantial mold risks over time. But Proust didn’t record the taste of tea and madeleines, he recorded his response to their taste, and no taste of tea and madeleines can guarantee a return of those memories. The best we can do is to reliably reset the stage with the correct ingredients, and turn things over to a competent actor.
One way of recording cookies is a recipe, and if you’ll allow the literary turnabout, this is an important lesson for preservation of many kinds of performative works. Sight and sound lend themselves to direct capture more easily than taste and scent. A well recorded recipe, competently executed, preserves the opportunity for re-experience or re-creation. As a preservation administrator, when the conversation turns to recapturing the moment, I always look for the script that created it. The score of the symphony, the recipe for the cake, the film of the race, the code for the game: these are all things that can be preserved. The subjective experience of them, the fidelity of one re-creation to another, or to the original, are matters out of my hands.
There’s a story told about the conservation of a Sol LeWitt wall drawing. A museum had one that LeWitt himself had drawn, and they took a conservation approach to it, removing the whole wall to put into storage. LeWitt was irritated. The museum had missed his point, the work of art was a recipe for re-creation, a script for how to make a drawing on a wall. Of course, LeWitt had missed the museum’s point, as well, which was that it’s interesting to have works in the author’s hand. If we had any recordings of Beethoven at the piano, we’d take pretty good care of them.
To summarize, then, when preservation starts to seem overly complex and slippery, I ask myself what kind of tangible thing is at stake. If there is one, start there. If there isn’t, then find out what can be recorded and how. A good way to do that is to find out who the re-creator is, and see what they require for their work.
A string quartet needs instruments (preservable), music (preservable), and players (not the professional domain of the library, though talented librarians are welcome to take part). A cookie requires a recipe (preservable), ingredients and a kitchen (best kept out of the library), and a chef (not our professional obligation, despite the number of conservators who are madly talented bakers).
A scholar requires resources, and those are surely our business. In this light, it’s interesting to think about the developing story of Paul Brodeur and the New York Public Library. In an interesting article on the Reuters weblog, Felix Salmon suggest that
From Brodeur’s perspective, he’s a reader as much as a writer, and his work aggregating and curating a huge amount of material on asbestos and the like is a central part of what one might pompously call his praxis. From the NYPL’s perspective, on the other hand, the value in his archive lies in his “manuscripts, notes, and correspondence” — the stuff that he wrote.
To me, that sounds like a conflict between preserving the recipe and preserving the cookie, and I’m not sure either party has their bearings yet.
2. Tuesday, April 26: Projects versus Progress
At the IPI workshop today, we naturally fell to talking about the relationship between collection managers and facilities managers. One of the hurdles in optimizing collection storage environments is that these two groups simply have different ways of working, each one adapted to the general requirements of their work and customs of their profession, and often enough, these are very different cultures. In the course of the discussion I mentioned to my (soon-to-be-emininetly-employable) students, Nora Bloch and Jacque Geibel, that this was an instance of a general problem we’ve talked about in preservation administration, and that’s the subject of today’s preservation week update. Let’s call it problems versus processes.
I suspect that every preservation librarian can quickly rattle off a list of specific problems they have on their to-do list. For me, there’s an incoming Ethiopian poster collection, Nazi-era press leaflets, recordings of Roy Newquist’s interviews, four volumes of Variedades that were somehow missed during a microfilming project in the 1990s; the list goes on and you get the picture. Each one of these is important, each one has a collection manager or three advocating for it, and enticingly, each one has a beginning and an end. These things could get actually and truly done.
And if I did all of them, several months would go by and the preservation objectives of the UCLA Library and its ability to be a trustworthy steward of the human record would be advanced barely at all.
My goal as a preservation administrator is not to fix all the broken things, but to make sure that the library becomes the kind of place that can and does fix things. If that sounds like a politician’s distinction, well, preservation administrators are universally middle managers, and I’d venture to say that successful middle managers have a little of the politician about them. They’re your representative to the powers that be, and the voice of the Capital back in your hometown.
At the risk of going once more down a Proustian path, I’ll give the metaphor one more paragraph. Preservation administrators, like good politicians, have to balance the creation of good government with enough tangible results to keep the voters happy. If the specific collections or items of concern never get tended to, that’s not great preservation, but if they get tended to while the roof starts to leak and the library starts to collect beyond its means, that’s not great preservation, either.
Let me give some concrete examples from my experience at UCLA. ALmost on arrival, I was consulted about a collection of Nazi press agency leaflets. These ended up at UCLA during WWIi. They bear the date-stamps that our old Serials department added as they arrived, with great regularity, it seems. We can’t say exactly how this arrangement was worked out, though there was a large German ex-pat community in LA as a result of the Nazi program, and many of you can attest to the incredible gravity that a library exerts, and the fascinating things that are pulled into our orbits. And many of you can guess that WWII era new leaflets have come down to us in a wretched state of brittleness. These are as fragile as anything I’ve seen, and it’s a huge collection.
Also, this may be the only set in the world.
Another example are the recordings of Roy Newquist’s interviews. Newquist was the host of a radio interview series called Counterpoint, and the nearly 300 recordings in this collection are a who’s-who of mid-20th century American writers and intellectuals. The collection includes the only recording of Harper Lee talking about To Kill A Mockingbird.
It’s an audio archive. It’s unique.
Here’s where the balancing act starts. Two years ago, UCLA didn’t have a reformatting program. We didn’t have a methodology for either prioritizing scarcely held materials or for taking action once a priority was identified. We did have options for digitizing things or even for microfilming, but each project was a one-off. Each project implied a new set of procedures and kicked the can down the road about the long term management of the resulting digitized versions. This went for books and paper, it went double for audio, and video or moving image was an order of magnitude more ill-defined.
UCLA still doesn’t have a great preservation review and reformatting program, but we have the foundations in place, and they support the whole library system. We don’t have the best digital preservation plan, but again, we have some foundations and as a library, we are starting to see a common set of issues and approaches to solving them. Given my preference, we’d still wait a year to dive into these two projects, but that’s why I’m not solely in charge, of course. (Though, note to University Libraries Gary Strong, if you want to knock off for a week or two and leave me with a modest discretionary budget…)
What’s happening here is a balancing act between priorities and process. As a preservation administrator, I want to make sure that if I’m hit by a bus tomorrow (or, more likely for Westside Los Angeles, a Lamborghini), the library will still have some processes in place that will serve it well into the future. I don’t want the library to be left in the situation of having had a few hundred more brittle books digitized, but no idea what to do about the next hundred.
In these specific case, we have had a really transformational year in our readiness for dealing with digital resources and for preservation review. (In an inside baseball note, it’s been transformational enough that we talk about preservation review, rather than brittle books or reformatting, and “preservation review” covers those services as well as scarcity assessment, persistent deposit, and replacement buying.) We’re on the cusp of a big step forward in our capacity to deal with audio, video and moving image materials. This year, I have a real live preservation budget to support projects in all these areas, and staff to see them through.
Because of all that, by spending time to develop processes first and holding projects off, I think we’ll be able to do better work for a long time into the future. That will benefit the next big ticket collection, but it will also benefit that as-yet-unread Bulgarian novel someone needs for their dissertation (a true story) or the millions of other workaday items that make up the whole of a great research collection. Developing the process also means that if a collection manager leaves (they do), or a preservation administrator is out for a a few weeks in Liberia (I’m always hoping), preservation will carry on of its own momentum.
All in all, I could still use a few months, maybe a year, to fiddle with my preservation program. But I’d be in danger of letting process get the better of projects. At a certain point, it’s time to take a leap, and that may be where having a little bit of political instinct matters to a preservation administrator. I could but maybe should not write an entry on that, so let me try to finish this on a note that we can all agree on. Process matters, and I think it’s the primary responsibility of a preservation administrator, but process has to be measures by its ability to generate projects. Why?
Because our collections are jaw-droppingly awesome and like any librarian, I’m eager to show them off.
3. Wednesday, April 27: Data!
Today, I want to take a break from preservation administration think pieces and write a little bit about tools of the trade. Preservation administration is arguably a subset of management, and the old adage that “you can’t manage what you can’t measure” has led to lots of preservation data collection. Various projects have measured costs, environmental conditions, productivity, and condition of materials in the collection. All of this is part of what gets variously called “data-driven” or “evidence-based” practice, the gist of which is that before doing something, you ought to have some measurable indicator of success or failure so that you can assess the impact of your work.
When I started at UCLA, there wasn’t a lot of data to go on. We needed to do a condition survey and since we didn’t have much in the way of staff, we needed a clever way to get data quickly. I decided to repurpose some work I did at NYPL when we needed to get condition information for the Rose Main Reading Room on short notice. This survey started from asking what possible preservation outcomes we had available to us and then collecting a set of data that would guide us through that decision tree. Other condition surveys have tended to be descriptive, collecting data to describe the items in the collection. This was diagnostic, collecting data to sort the collection into categories for preservation action.
Both types have value, of course, so we put some flesh on the bones of that NYPL survey and did a few rounds of data collection and that was pretty good. But most preservation surveys trace their roots back to the Yale survey in the 80s and the methods Carl Drott developed for collection managers to sample for usage in their printed serials collections in the 60s. When I called up a colleague in the UCLA Statistics Department and talked to them about what I was doing, they very nicely suggested that the state of the art in statistics had come a little ways in the last 30 or 40 years. So, they put three undergrads on the project of hacking through the data that I had, and in this post, I want to give a taste of what they did.
If you already know about Differentiation and Fuzzy Cluster, CART, and AGNES, there’s nothing to see here. If not, read on for bewildering charts and sophisticated statistics.
The images below are extracted from a white paper that I’m working on with the students. I’m not going to go into a lot of detail on the methods, because my understanding is purely conceptual at this stage, and I couldn’t explain anything of substance. The point here is to show how valuable it is to collaborate with the smartest people you can find. It doesn’t matter much if they know anything about libraries; that’s what the librarian is there for. What does matter is that they know things you don’t and, hopefully, that they can tell you that you’re wrong in diplomatic ways.
One of the approaches they took to our datasets is Classification and Regression Trees (CART). This groups like data together (classification) by splitting them into branches around a particular value.
Then, the ends of those branches are compared to the decisions that would be arrived at through selection of an arbitrary value. If the classified branches are more consistent with the actual data than the assumption that everything shares a single arbitrary value, then the values used to split the data makes for a good predictor. In this case, splitting the acidity by decade of publication was a better predictor than simply assuming that all books were acidic. That’s a good prediction statistically and useful, since it confirms that there are a range of acidities in our collection and gives us specifically informative points in time where we can expect different gradations.
Another, related approach is Dissimilarity and Fuzzy Clustering. IN traditional clustering, one looks for closely related groups of data and clusters them. At first pass, though, our data looked like white noise (or, in this case, orange and red noise). The wasn’t much granularity of data that we could realistically collect based on a quick physical assessment of a book. This is something we’re going to try to rectify in an upcoming version of our survey tools, but for the moment, it seemed like we were at an impasse.
However, the people doing the analysis are much savvier than I, and they decided to do Fuzzy Clustering. In this method, data are grouped based on their dissimilarity, rather than their similarity, by finding groups of data that are closer to one another than any of them are to the next cluster of data.
From there, hopefully, some patterns would emerge. They did!
And what did we learn? Happily, that many of our rules of thumbs about the preservation problems related to various eras of printing are borne out by independent data analysis. Mid 19th to mid 20th century (the groundwood pulp era) shows up as a poor condition group. The mid 20th to the end of the 70s (the more recent groundwood papers) show up as a group and, enticingly, look like good candidates to stabilize by environmental control or deacidification. Finally, the post 1980′s book (from the era after the introduction of alkaline paper standards), all look pretty good.
The last thing they tried was an Agglomerative Nesting (AGNES) dendogram. This one still needs some work, but I think it’s work looking at for two reasons. One, I think it quickly drives home the point that there are more sophisticated approaches to these data than finding the average and median.
The other is that it refines the fuzzy clusters in interesting ways. Keep in mind that the analysts don’t know anything more about preservation than what I could convey in a one hour guest lecture for their class. They presented this table to me and asked, “so, does this mean anythign to you? Because these are important groups in your data.”
They do mean something to me.
1. Major damage, acidic and brittle, high use: Preservation Review, probably leading to expedited reformatting.
2. Minor damage, acidic and brittle, low use: Preservation Review, leading to a balance of replacement, storage, reformatting and withdrawal.
3. Minor damage, acidic but not brittle, low use: Deacidification or environmentally optimized storage.
4. Incidental damage, good paper condition, high use: Repair in house.
5. Incidental damage, good paper, low use: Outsource repair to our library binder.
So there you go. Five categories discovered purely though data analysis, that can really inform our preservation planning. For me, this was especially valuable since we’re in the early stages of developing a reformatting program and we’ve been debating the role of deacidification in our strategy. Obviously, there’s a lot more to do in our collaboration with the Statistics Department and a great deal more technical writing to be done to make these methods and our results useful to others. I hope this gives a sense of what’s possible in a good collaboration, though, and the value of evidence-based practice.
4. Thursday, April 28: Preservation Education
Since I’m posting from my alma mater, we’re coming up on the deadline for our summer internships, and have several of our student assistants about to hit the job market, I thought I would share some thoughts on getting an education for preservation and conservation.
I got into this line of work by happy accident. I decided on Indiana University because I could dual enroll for music and library science. I took a part-time job in an early book digitization project that was co-funded by the Preservation Department and the Digital Library Program. One thing led to another.
At the time, UT Austin was the standard-bearer for library and archives preservation and conservation education, and the roots of that program were in the program at Columbia University, the first of its kind in the US. Outside of UT, Pittsburgh was the other center for preservation education, though to my knowledge Pitt didn’t have the same conservation options as UT. In the last few years, though, the preservation tracks in both of these programs have diminished or disbanded. In 2011, it’s very hard to give a concrete answer to the question of how a library and archival science student should get an education in either preservation or conservation.
have taught many graduate preservation classes over the years, at Indiana University and the Pratt Institute, supervised at least a dozen internships and independent studies, and taught plenty of workshops and short courses, online and in-person. And, of course, I’ve been working in preservation administration all that time, at a few different institutions. My impressions from all of that are that the things that every librarian or archivist ought to know can be easily covered in a semester, or a couple weeks of online modules.
When Karen Brown and I developed Fundamentals of Preservation for ALA, we drew on all our old syllabi, which covered our own work at NEDCC (and the NEDCC/IMLS core curriculum project), IU, and Pratt, as well as bits and pieces from our colleagues all over the nation. We needed to use a heavy editorial hand, sure, but I think that the course gives a solid foundation in four weeks. Likewise, I think that the students who went through my graduate level classes will be able to have a functional conversation about preservation, no matter what their on-the-business-card title may be.
However, none of them could step into a preservation job on that education alone. The array of skills and the knowledge-base that my peers across the country have to hand are simply not covered in a one-term course and by and large, they’re not covered in library school curricula. The unfortunate outcome is that students come to LIS programs and have an initial experience of preservation, through a class or occasionally a student job, and they want to learn more and perhaps pursue this as their career direction. The first part of the educational plan, then, is working just right. Our graduate programs are exposing students to new horizons and helping them envision a professional future.
At that point, we abruptly fail. Even when UT and Pitt were going strong, telling a student who had recently discovered preservation that they’re best bet was to apply to a different program and relocate was a pretty shabby sort of advice. Without those programs, advising students to try to string together a few extra courses if they’re offered, plus a student job and an internship or two, and then hope for the best doesn’t seem much better. If we were to consider a viable early-career preservation librarian to be someone with at least two graduate level preservation courses, basic bench skills, and an internship in preservation administration, I’m not sure that we have the capacity to train more than a handful each year across the entire United States. And even then, I’m not sure we have enough entry-level jobs to get their careers started.
All of this to say that we ought to be giving substantial thought to education, retention, and succession planning in our profession. To bring these threads together, I’d summarize by saying that:
- We ought to suspect that monolithic programs are vulnerable. It may be better to have a slightly more robust curriculum at many places than a single highly focused program at one school.
- If you’re a mid-career or senior preservation librarian, and you don’t have an intern, get one. I have several each year and yes, the startup period takes a lot of time and energy. But every internship term I’ve completed was more productive with the two of us than it would have been for me alone. The early effort pays big dividends.
- If you’re not having this experience, if internships just feel like a time sink, consider pushing your intern out of the nest sooner, and from a higher elevation. They didn’t get into library school by accident, and they have a powerful incentive to succeed. I have seen a few of mine stumble, of course, but no one completely crashed and several really took wing.
- We need to start thinking about succession: assistant preservation admin jobs, grant projects that allow for substantial management experience, and deeper pockets for professional development of our early-career colleagues. I think most places provide maybe $1K. We should aim for $3-5K per year.
- And prologue to tomorrow’s post, where I intend to write about the preservation administrator-conservator relationship: we need to re-think our expectations for conservators. There’s a big discussion going on about the need for an ML(i)S for library and archives conservators. We need to engage with this and see what we can do to create more pathways for conservators to work in libraries and archives
5. Friday, April 29: Preservation Administrators and Conservators
The preservation administrator’s most important working relationship is with the conservators. This is such a core part of the work that I have to constantly re-correct myself from writing with possessives: my conservators, our conservators. A good conservator is a treasure, well worth coveting. A critical part of my formation as a preservation administrator was the opportunity to work with conservators who really knew their business. I consider myself blessed to have had a chance to work with many of the finest, at Indiana, New York Public, and now at UCLA.
There is a special dynamic in the library preservation model that mirrors what happens in health and medicine. Lewis Thomas observed that the increase in human longevity was more attributable to plumbers than doctors, alluding to the importance of sanitation and public health, rather than medicine, in effecting broad improvements in our quality of life. That said, Lewis Thomas was one of a legion of people who have made important advances and observations about the power of medicine to cure and correct. The one is best with the other. I think this is true of preservation and conservation, as well.
This relationship is also becoming more important.
I think that as preservation administration finds its feet in the 21st century, one of our key roles will be to provide the administrative and public health environment for a growing and more diverse group of conservators. Book and paper conservation are already part and parcel of the library preservation effort, and I would venture to say that all of us could keep a photo conservator busily employed for life and then some.
It is increasingly common for libraries to engage with audio and video media preservation, and I submit that recording engineers are conservators by another name. By way of proof, I suspect that some people read that last sentence and thought, “well, it’s really grooved media and magnetic media, as much as audio and video, and of course motion picture films are really photographic transparencies.” It’s exactly what happens when you say “leather” or “book” around a conservator. Which animal? What sewing structure? Alum-tawed or chrome tanned?
This goes further, too. Most preservation administrators will also be called on to work on digital preservation. But no one has a deep theoretical understanding and practical competency in every one of these areas. When I look at my IT department, I see digital conservators waiting to be formed. In this area, as in every other area of our program, we will need to channel the format and technology specific expertise of specialists into service of the library’s preservation mission.
You can see where this comes back to the hospital metaphor. No hospital administrator or public health official can slip on the gloves and stand in for their surgeon, their anesthesiologist, their nursing staff, their GI doctor, oncologist, and so on. But they must know enough to have a productive conversation and be able to martial those forces to save lives and improve health.
I think that modern preservation administrators are being pushed and called to that same role and that gives me a sense of hope and excitement. There’s a real set of theory and praxis to be learned and developed here: how to get the best out of expert practitioners; how to determine when a conservation speciality needs to be inside the organization, and when it can be an external provider; how to see trends and track progress; the list, as always, goes on.
This way of thinking obviously bears on what education for preservation administration should consist in. I think it also bears on one of my habits for self-assessment and program development. I’ll go out on a limb and say that the projects that have really worked have been the ones where I looked outside my department and pulled in experts: the Anderson school and the Statistics department at UCLA, the Image Permanence Institute, the library schools (over and over again), and happily this list can go on, too.
Summing up? Find the smartest people you can and get them in a room together. Make sure they have plenty of big tables. Ask them to fix some things. That’s probably as close to a philosophy of preservation administrations as I’ll get today.