It is looking like I may end up teaching a graduate seminar on digital preservation for the University of Maryland’s iSchool. There is an existing syllabus, but I will have some flexibility in terms of how I shape and design the course and I am curious what thoughts different folks have for what would be the most effective way to teach a graduate seminar on the subject.
Below are a few of the big picture course design questions I am thinking through and some of my initial thoughts on them. I’m curious for any and all input folks might have.
Organizing Principles: How best to organize a digital preservation course?
- To what extent should such a grad seminar like this be about frameworks and principles vs examples and cases? I’m thinking that I should cover those, but I’m also thinking that too many of those models fail to address the idea that digital preservation is fundamentally about risk mitigation from future loss. That is, it’s less about a process and more about how to make the best use of available resources and identifying the best opportunities to systematically work to further lessen the risk of loss. I also think that the frameworks often get in the way of first grasping a fundamental understanding of the nature and structure of digital information and digital media. So I’m entertaining the idea of getting to the frameworks at the end as a way to understand the issues but working through the core issues first.
- How would you organize and structure such a course? If I don’t start with the frameworks, I’m thinking it makes sense to start by working through a core understanding of digital information and digital media and work from there into the various issues in the NDSA levels of digital preservation.
Particular Tools & Software: What role should they play in the course?
- What approach should I take toward particular tools? On the one hand, it is very pragmatic to leave a course like this understanding how to use particular tools, but at the same time, the tools are always going to be changing and everyone needs to be able to plan for how to swap in and out different tools to meet the underlying objective. In my digital history courses I have required students to each figure out how to use and then teach the class how to use particular tools and software. I like this approach as teaching yourself how to use new software and evaluating it is an important skill in it’s own right. With that said, some of the digital preservation tools out there are complex enough that I’m not entirely sure this method would do them justice.
- How much should a course like this require/push students to develop some basic command line literacy? My sense is that many student’s will not have this, but it is challenging to think through how to do much work in this area without that. With that said, the course isn’t about developing that command line literacy, so I’m not sure how far to delve into this kind of thing.
Kinds of assignments: What would be the most useful for the students?
- I’m curious for what folks think would be the most useful kinds of assignments. I’m thinking that given the context of planning for risks and the need to make such plans inside the constraints of an institution that it might make the most sense to have students serve as consultants for small cultural heritage organizations and have them develop plans for options to improve their approaches to ensure long term access to their digital content. So I think many of the assignments might be fit around that. With that said, I am curious for any other ideas for how to either improve this idea of a course project or for other kinds of assignments.
These are good questions, Trevor. As you mentioned in your post, digital preservation is an exercise in making the best use of available resources to mitigate risk. In my work with libraries, the starting point is always, “What guarantees and promises do you want to be able to make? Can you keep these promises with existing resources? Are there other organizations that can help you share the workload, risk, and benefit?” These basic questions, along with an examination of the benefits of policies to justify and support taking on any level of digital preservation work, seem to be very useful first steps.
Trevor – Good questions and scope. I’d love to chat with you on the phone about this (may be easier and more fluid than writing out my thoughts in a linear method here). For certain, I would begin with the construction of digital information as the beginning and central theme of the course. Digital preservation is 100% about digital information. The preservation component brings risk mgmt and planning, yet everything in a digi pres course revolves around the environments of creating, using, and storing binary information. Just as a paper conservation course would require the students to understand the chemical construction of cellulose in order to design treatments and preservation plans for paper materials, a digital preservation course would do well to ensure students understand the target of preservation before understanding frameworks and approaches for preserving said target over time. Maybe you read this blog post of mine about the chemistry of digital preservation (https://www.avpreserve.com/blog/what-is-the-chemistry-of-digital-preservation/), but if not, take a look.
From that foundation then you can articulate the strengths and weaknesses of various approaches to digital preservation. And focus on the differences of contexts in which digital information is preserved.
Hands-on projects with “clients” is a useful approach. My recent Digital Curation students at UW-Madison worked in groups as consultants with local orgs that had digital curation needs. They worked on the projects throughout the semester and I think (1) they all did great work, and (2) they all applied what was learned in class to the projects.
Finally, yes on the command line skills. I think it’s a must (and something that can be tied into the beginning familiarity with the technical construction of digital information).
I’d love to chat more if you want. Just let me know! Good luck with the course! – Bert
Hi Trevor, this is a great set of questions and I can’t wait to see the final syllabus. Just a quick note on the command line. In the course Susanna Allés and I taught for undergraduates this semester, where we wanted to teach them a ‘full stack’ for digital editions, command line took two sessions of 1.5 hours each, with homework of 4 hours. Here’s hoping our numbers can give you some perspective. We did this at the beginning of class, and of course, we had to go back in there and refresh and expand on the basics, but they did just fine. Github on the other hand…
To what extent should such a grad seminar like this be about frameworks and principles vs examples and cases?
20% frameworks, 80% examples. Agreed on your comments — but you know my take on this!
How would you organize and structure such a course?
Historical would be cool. Has #digpres adapted (in any sense) in concert with technological change? Should it? If yes, how? If not, why not? This also allows you to put frameworks in historical context and explicate their own evolution. Plus, chronological readings allow great insight into how people argue with each other (and thus how knowledge/practice evolve, especially via argumentation)!
What approach should I take toward particular tools?
I’d approach it less as being about tools, per se, and more about tool-sense (!). So, say, either “here’s a tool, go find problems/issues it solves/addresses and describe” (application) or “here’s a problem, what did you end up using to solve/address it and why” (diagnosis). The tool itself should be replaceable, uh, pedagogically.
How much should a course like this require/push students to develop some basic command line literacy?
The Command Line Crash Course (or similar) is useful, perhaps, but can be self-driven work. More useful, imo, is understanding how technologies interoperate (or don’t) and things like architectures, front-end v. back-end, protocols/APIs, data models/transformations, abstraction layers, and so on. Some exposure to the potential issues inherent in the complex coupling of systems and processes seems more valuable than how to grep a file or what not. Being able to recognize potential faults in the “tool chain” or workflow is (hopefully) teachable w/o getting into CLI doodading.
I’m curious for what folks think would be the most useful kinds of assignments.
How about assignments that have no tangible success metric or completion point? How about assignments that are about failing gracefully and whose learning experience is not outcome oriented? The curricular conceptualization of MLIS stuff can trend towards a lazy reconstitution of the undergraduate experience (which has drastically different goals and social/cultural intentions) or an on-ramping-to-the-PhD structure that doesn’t make much sense. So how can assignments both give exposure to actual practice and exposure to the work-a-day world in which success/failure are often opaque, circumscribed, or circumstantial? The assignments and outcomes of the school world and the assignments and outcomes of the world world never felt, to me, within the same orbit whatsoever. Anything that can bring these celestial bodies closer together (without interplanetary collision, of course) seems a win.
Hi Trevor
To some extent, it depends on how digitally literate these students already are, and whether they are coming from an ILS background (including RM and archives) or more generally.
For my course in preservation, which I teach at UG and PG level to new to the ILS profession students, I have a four week window (one lecture and tutorial/workshop, per week) to cover digital preservation. As Bertram does, I teach them about the make up of digital information. We cover, or refresh, bits and bytes, etc, and talk about common reasons for deterioration (both in the code and the media). We talk about digitisation, because for many of them, that is the key push for their collections, then look at the different formats; emulation, migration and reformatting generally and discuss open source and open standards, as well as why choosing something popular is also a good strategy. We cover OAIS, and discuss different sorts of metadata, and ways of transferring information ( I call it the alphabet session- METS, MODS, LOCKSS, CLOCKSS, PREMIS…) It’s more strategy and policy than hands on, simply because they don’t have the background and I don’t have the time, and most of them are online and asynchronous.
I get them to look at the Digital Curation game, because making decisions about what to keep is as, if not more, important than how we keep it, and the LOC personal archiving quiz. We used to get them to download a copy of Xena from the National Archives of Australia, and normalise some content, but as the institutional support at NAA is not there anymore, we’ve had to stop. If you have internal students and a computer lab, you could set up an instance of Archivematica or similar, and get them to transfer and ingest some records, creating SIPS and AIPS.
I get my students to complete a fake grant application for the National Library of Australia community heritage grants – they have to complete a Preservation Needs Assessment form for a real or fictional collection, and then decide what to do. You could specify that it is for a digital collection, or for a collection that is to be digitised. They have to identify who does what, in what format, where and how it is to be maintained and in what format(s), and they have to cost it. I’d be happy to provide a copy of the assessment criteria and materials if you are interested.