Why do we so often talk about ‘best practice’ in digital preservation when we don’t always have clarity on what that actually looks like? Indeed, if it is our policy to try and carry out digital preservation best practice are we simply setting ourselves up for failure? How is it possible to comply with something that is so hard to pin down? What if what is ‘best’ for one organization is suboptimal for another?
I’ve been meaning to write this blog post for a while as these are questions that have popped into my head frequently since joining the DPC as Head of Good Practice and Standards 4 years ago. I didn’t have any say over the job title I was assigned when I joined the team, but I very much appreciated the foresight my colleagues had when they established the post. It already felt like a daunting role to fill, but if it had been ‘Head of Best Practice’ I might have had serious reservations!
Before I dive in with some thoughts on the questions posed above, here is a quick definition of best practice.
“a working method or set of working methods that is officially accepted as being the best to use in a particular business or industry, usually described formally and in detail” Cambridge Dictionary
However, it is also often noted that there are some challenges related to this choice of phrase. For example, Wikipedia states that “sometimes a best practice is not applicable or is inappropriate for a particular organization's needs. A key strategic talent required when applying best practice to organizations is the ability to balance the unique qualities of an organization with the practices that it has in common with others.”
An interesting blog post from Daniel Thornton titled What’s wrong with best practice in organisations? gives some history to the term and its usage and asserts that using terms such as best practice is actually ‘bad practice’. So why are we in the digital preservation community so wedded to it?
What is digital preservation best practice anyway?
The digital preservation community’s preference for the term ‘best practice’ is clear. In a fairly non-scientific demonstration of this, I carried out a text search of the proceedings of the most recent iPres conference in Glasgow. Whereas the term ‘good practice’ appears 32 times (actually 13 of which are in my own and my colleagues papers!), the term ‘best practice’ is used 56 times. This quick and dirty methodology suggests to me that a majority of digital preservation practitioners favour the term ‘best practice' - but what does it mean in the context of our work?
If you asked someone working in our field what digital preservation best practice actually looks like, they might mention OAIS, the Open Archival Information System Reference Model. They might also mention ISO16363: Audit and certification of trustworthy digital repositories. However, both of these standards have their challenges:
- In a recent DPClinic session we held on OAIS, speakers discussed the incomplete nature of OAIS and some of the different ways it could be interpreted. Can a standard that is incomplete and/or easy to misinterpret be considered best practice?
- At the iPres 2022 conference, in a talk by Jessica Tieman on ‘A decade of Trustworthy Digital Repository Certification: Yet there was one’, the U.S. Government Publishing Office’s experiences of achieving and retaining ISO16363 certification were described and there was discussion around what action they took when they realized that the rest of the digital preservation community was not following suit. Can ‘best practice’ really be best practice if the community is barely using it at all?
Referring back to the definition above, does digital preservation actually have a set of working methods that we can say are “officially accepted as being the best”?
Can we define digital preservation best practice if we can’t evidence success?
‘Best practice’ is a term that is widely used in business and industry. The term works well in many contexts. In medicine there is clearly a need to be able to define best practice for the treatment of certain conditions. Assertions of best practice in a medical context are typically backed up by science and with evidence, perhaps from clinical trials or recorded cases with positive outcomes. If I am being treated by a doctor, there is no doubt in my mind that I would like them to follow best practice in the treatment of whatever ailment I might present with - and indeed that this decision is backed up by evidence!
In an industrial setting there may be best practices in place for many processes and workflows - for example safety procedures to minimize employee injuries. Again these things are measurable, and positive outcomes can be demonstrated over a fairly short time frame. Though not without its challenges, best practice as a term seems easier to pinpoint and agree in such circumstances.
But digital preservation is different. Our timescales are longer and we do not have clear and immediate ways of measuring success. The DPC has recently completed a project with the UK Nuclear Decommissioning Authority where we have had discussions about the preservation of vital records concerning nuclear waste that will remain radioactive for 10,000 years. It does not make sense to measure the success of the project or actions taken in the short term, and we will not really know if preservation strategies implemented today have been successful until a time period of thousands of years has passed.
Whilst this doesn’t mean that we shouldn’t try to benchmark and measure success, it does make it hard to come up with clear evidence that any working methods we put in place are “officially accepted as being the best”.
Does best practice give us a false sense of security?
The idea of ‘best practice’ can lull us into a false sense of security. There is something very definite and certain about using this phrase in the context of our digital preservation efforts - “If you follow these best practice guidelines, your valuable digital content will be preserved forever...”!
But we all know that digital preservation isn’t like that. It is an evolving discipline. We do what we think is right in the here and now, and assume that further actions will need to be carried out later. We may even need to reverse actions and decisions that have been taken previously, as our knowledge and understanding as a community evolves.
The DPC’s Digital Preservation Handbook has a page on ‘Standards and best practice’ and I’ll pick out one quote from that section:
“The digital environment is not inclined to be constrained by rigid rules and a digital preservation programme can often be a blend of standards and best practice that is sufficiently flexible and adapted to suit the needs of the organization, its circumstances and the digital materials being managed.”
But if we are flexing and adapting best practice anyway, should we change both our language and the expectations that go alongside it?
If there is such a thing as best practice, how come the answer to every digital preservation is “it depends”?
There is no one-size-fits-all approach to digital preservation. There is no one solution that has been officially agreed to be ‘the best’. One of Trevor Owen’s Sixteen Guiding Digital Preservation Axioms in ‘The Theory and Craft of Digital Preservation’ is that “The answer to nearly all-digital preservation question is it depends.”
Some of our work at the DPC involves being able to support our members with difficult questions and decisions about digital preservation. We are often approached with questions about how to carry out a particular task. What our members may be looking for is a very clearly defined answer to a question like ‘what does best practice look like for digitisation?’ They may not be prepared for a much more nuanced answer (which very often does start with ‘it depends…’).
The answer to a question about best practice shouldn’t start with “it depends”, but when recommending ‘good practice’ in a specific scenario, we are bound to follow up with questions about what exactly you are trying to achieve, what the value of the content is, who is it for and what do they require, what resources are available and what other relevant factors are at play. Using this information we aim to come to a workable good practice solution that will be appropriate for that particular scenario.
In a blog post after the iPres 2022 conference, Brendan Power from Trinity College Library, Dublin hits the nail on the head:
“I think we are often tempted to assess our practices against advanced metrics which can be beneficial in setting our ambitions high but potentially limiting as the scale of the challenge can be seen as overwhelming when the aim is focused on “best practice”. At iPres 2022 there was a lot of mentions of “good practice”, which I think encapsulates a more flexible approach and one more easily achieved. I think “good practice” implicitly acknowledges that your practices must be guided by the ambitions that you have set and not just a pre-defined set of criteria.”
To me, ‘best practice’ implies there is one best way of doing things - better than all the rest! ‘Good practice’ is more fluid, allowing numerous different approaches depending on the circumstances and that seems far more in keeping with how we actually work as a community.
Best practice - at what cost to the planet?
Increasingly, we as a profession need to think about the impact of our digital preservation actions and infrastructure have on the environment. You should definitely read Toward Environmentally Sustainable Digital Preservation by Pendergrass et al if you haven’t already. Is ‘best practice’ for digital preservation bad practice for the planet? How does digital preservation 'best practice' sit alongside some of the ideas within this article such as the concept of “acceptable levels of loss”?
Rather than rigidly promoting digital preservation best practice I prefer to think of a more balanced way of working. A practice that considers the needs of the digital content (and its users) alongside available resources (and I include within that, the natural resources that we use in order to store, manage and provide access to that content).
If 'best practice' says that you should store 3 copies of all of your digital content and run a checksum verification every month (it might not…this is just an example!), 'good practice' could say that you should do this for really valuable content, but for content that is less critical, you might get away with less copies and less robust or frequent fixity checking. This could be a perfectly appropriate conclusion to reach given resourcing constraints and the carbon costs of storage and calculating and verifying checksums.
Is best practice equitable and fair?
When we set the digital preservation bar too high, does that inadvertently discourage organizations from making progress? What about smaller or under-resourced organizations that will never be able to achieve perfection?
A motto that I have increasingly lived by in my working life is “don’t let perfect be the enemy of good”. Rather than fixating on best practices, it is better that organizations with digital preservation challenges just try to get started with doing something. Taking (even small) steps in the right direction is something we encourage, rather than waiting for everything to be perfect before moving forward.
Waiting for perfection and ultimately doing nothing might actually be seen as bad practice, but taking small but positive steps in the right direction is good practice and something we very much encourage at the DPC.
One of the ideas behind DPC RAM, our Rapid Assessment Model for digital preservation, was to try and encapsulate and describe good practice and present it in a way that helps organizations get started and move forward. The model allows an organization to assess and track their digital preservation maturity over the eleven sections. We specifically state that each organization should set their own targets for improvement. Not every organization needs to strive for the highest level across the sections of RAM (though of course they can if this is appropriate for them).
Is ‘good enough’ good enough?
In a presentation at iPres 2022, Going for Gold or Good Enough: Observations on three years of benchmarking with DPC RAM (page 305) myself and colleague Paul Wheatley stress that not everyone needs to aim for gold standard digital preservation in all areas:
“The flexibility of DPC RAM in particular enables organizations to not only decide how they will carry out a particular aspect of digital preservation but also define what ‘good enough’ looks like for them. Though the DPC regularly stress that RAM is about continuous improvement, improvement should only continue so long as it is necessary.”
Continuing this theme of ‘good enough’ digital preservation, From Theory to Action: “Good Enough” Digital Preservation Solutions for Under-Resourced Cultural Heritage Institutions from the POWRR project also promotes this approach for small and medium-sized institutions with limited resources:
“Understand that digital preservation is an incremental process. Digital preservation is achieved through cumulative activities of increasing efficacy. It is time to embrace a “good enough” approach to digital preservation. Do not begin by aiming for designing and implementing a robust program with a detailed workflow for ingesting digital objects into a technical infrastructure fully in keeping with standards and best practices.”
The role of ‘good practice' and even ‘good enough practice’ is important. Digital preservation should not be an elitist discipline to be reserved for a small number of organizations who are able to put resource into formal certification, it should be possible for all organizations to take steps towards preserving their digital content. As a community we should be conscious of the need to lower the barriers and encourage greater participation and engagement rather than fixating on a ‘best practice’ that only a minority could hope to attain.
A call to action…
Is it time to shelve the term ‘best practice’ for digital preservation and instead prioritize being good?
Thanks to my DPC colleagues William Kilbride, Paul Wheatley and Michael Popham for commenting on a draft of this blog post.
Comments