Digital preservation is a daunting challenge. It is a paradox that the cumulative effect of more than a decade of research and development in the topic seems to have made it impenetrable too. We started with predictions of data loss and since then our wellintentioned enthusiasm has produced great reams of blogroll, huge stacks of reports, endless screeds of code and toppling towers of power-point. The new projects, the new agencies, the acronyms, the jargon, the bluster and the debate constitute a discourse that would more likely discourage a novice than reassure them. Initially doom was our only colour scheme but in the last decade we seem to have settled for a literature in two forms: accessible, superficial doomladen premonitions of imminent disaster; or deathless cryptic monotones about partial solutions to infinitesimal problems. Both are prone to exaggeration: neither are very persuasive for terribly long. There are times I pity my students.
actually a diagnosis of health. In a rapidly developing field it is inevitable and useful that research should cover all sorts of ground. It’s inevitable that some of it will appear impenetrable and some of it will get forgotten – what matters is that fragmented research congeals into a common understanding and a coherent set of practices. Novices need not be exposed to it all: teachers need a measured and thoughtful approach that makes sense of the whole and which engages them in the parts that really matter. It’s the job of the teacher to interpret and keep abreast of the increasing specialisation, the evermore recondite detail, the exhaustive scrutiny of tools and services, and to assemble from them a coherence that engages and enlightens. Perhaps it’s my students that should pity me.
So we’ve come a great distance in a short time. The fact that we’ve managed to turn our initial fears about digital calamity into something that now seems pretty boring suggests we’re on the road to taming them. But there are few commentators who would say that the problem is solved. Most imply that apparent solutions tend only to reveal ever more subtle problems. In any case, change is not a bug: it’s a core feature and one of the principle benefits of IT. We’ve proven pretty conclusively that finding solutions is not a problem: we’re almost too good at it. Translating research into practical executable guidance seems to give us problems. So two question arise for those who want to teach digital preservation: how much of this ever-thickening syllabus do we really need to hand on; and how, if we’re not quite sure how to fix the problem ourselves, are we going to show others how to do it?
allowed a brief remembrance of what makes digital preservation an issue. Digital preservation is not like preservation, at least insofar as there are very few natural processes which we need to confront. That’s to say, while traditional conservators are busy fighting an eternal battle with bacteria, chemicals and grot, our enemies – obsolescence, representation and bit rot – are practically always of our own making. So is it possible that we could make obsolescence obsolete? The idea may seem far-fetched but it’s not out of reach and it would transform what and how we teach digital preservation.
research agenda and another work programme. The purpose is to ensure that skills remain current.
The same thought-experiment is possible with some of our familiar metaphors. For example, It’s becoming increasingly clear that what we called a ‘repository’ in 2000 is less of a ‘place’, it’s more of a ‘process’. We talk of ‘trusted digital repositories’ when we actually mean the deployment of trusted services by trusted agents with trusted data and trusted processes. The repository, if it exists at all, is retail data storage for the AIPs. So when we talk about assessing whether something could be a ‘Trusted Repository’ what we actually need to assess are the services, the processes and the people. And in a service-oriented environment, with dependencies on a constellation of remote tools and operators, and where we call on ‘Digital Preservation as a Service’ we package trust along some very long supply chains. So why are we assessing and certifying repositories? Why isn’t there a ‘DP service seal of approval’? Perhaps instead of teaching students about the characteristics of a trusted archives we should teach them how to assess software dependency in a highly distributed environment?