Report from ‘Improving access to archives and other records’ – Melbourne edition
by Belinda Battley
This was the question posed by Anne Gilliland at the end of the discussion this week of Chris Hurley’s Modest Proposal, hosted by the Centre for Organisational and Social Informatics at Monash University, in association with the Recordkeeping Roundtable. The talk began with a suggestion that we need to determine functional requirements before developing a federated system for discovery and access. It ended with the suggestion that even before determining functional requirements, we need to look at the principles and rights involved, and deal with multiplicity by supporting contested territory, honouring the different narratives, and allowing people to arrive at the records they need from any direction, any description that meets their needs. At the end of the evening, Luke Bacon set up a space where the discussion and collaboration can continue and develop, beginning with a new set of archiving first principles.
The evening began with Chris reprising his talk to the Recordkeeping Roundtable in Sydney, with his proposal for federated access to archives and records to address the need to reflect the real world of records today. Next Luke Bacon, Ailie Smith and Kirsten Thorpe responded to his proposal from their differing perspectives and interests, as described in the previous post to this discussion group. There was a lot more discussion following these responses, and I’ve tried to summarise it below – it was great to see how the instant back and forth of face-to-face discussion allowed the focus to move and develop into a project that could begin immediately.
Chris Hurley suggested several methods for achieving a “wonderland” of archival description that showed the complex interrelationships, alternative viewpoints and changes over time of the real world instead of being a 2-dimensional reflection. He suggested a wiki or similar format which would provide a gateway to native descriptions, stored outside the wiki but linked to a federated structure. He suggested the wiki could be an assemblage of initiatives, many of which have already begun, such as the Australian Pictorial Thesaurus, Geoscience Australia (for identifying places), perhaps Trove for many existing items, and so on. Wiki contributions would be output as data in native systems – it would just be necessary for the IT people in each contributing organisation to ensure there was one point of contact between the wiki view and their own. After that, he saw the rest of the work as being done by “citizen archivists” according to their needs.
Chris proposed that it would be necessary to set out functional requirements before any structure was developed, and he suggested seven as a starting point:
- Protean (open to differing viewpoints and constant change)
- Inclusive, easily attainable for “barefoot archivists”
- “Wholeistic” – not only records in collections – also e.g. research datasets, meteorological data etc.
- Collaborative – discovery trails shaped and enriched by a whole range of contributors
- Authentic – must use native descriptions, not just a rendition
- Multifaceted – opening up descriptions to different ways of viewing resources, e.g. parallel, multiple simultaneous provenance
Chris said he would provide a revision of his modest proposal on his website soon.
Luke Bacon responded, describing himself as approaching this from the context of the UNESCO Universal Declaration on Archives , and noted that by giving people tools to improve their access to archives we enhance their quality of life.
Luke said the proposal interested him because a lot of people in web design are talking about the same thing right now. He liked that it is decentralised, and includes people who can’t achieve the “gold standard” of archiving. Citizen archivists aren’t just providing descriptions, they are also providing access to archives themselves. Archival institutions need to stop being so “jargony”, should talk to people on their own terms.
Luke referred to the “indie-web”, a federated social web developed as a fight-back against the siloing of commercial social platform providers such as Facebook, Twitter. The indie-web developers have come up with some principles that could be relevant for federated archives:
- Everything must be open-source (crucial for decentralised system)
- “Eat your own dog-food” – build and implement on your own website, not just for others
- Simple formats
- Multiple points of access – users can make their own interfaces for using it
Luke suggested the project be set up as a GIT rather than a wiki – and then immediately set up a starting point, here . He said that a GIT allows parallel stories, different versions of documents can exist simultaneously, and any could be seen as authoritative depending on the viewer’s point of view.
Ailie Smith responded next, referring to a recent blog post, Kill the Finding Aid, which said 99.8% of people don’t know how to use finding aids and would rather search item descriptions and find digitised stuff. Most people who need archives aren’t professional researchers, and trying to control the pathways people use to get to information isn’t going to work. Instead, we need to decide what needs to travel with each record we make available so people can interpret what they are seeing. Ailie reiterated the points that small archives might not have the resources to do detailed description, and that existing descriptive standards can be either very flexible, such as EAD, where people don’t necessarily use the same elements for the same thing, or very inflexible, such as RIF-CS (used in Australia for research data sets). Ailie suggested that collaboration, engagement and empowerment were needed to help small archives with little support, and that Chris’s proposal could be empowering for under-resourced archives, if it didn’t require too much IT knowledge. Ailie also commented that it was important for archivists to claim our space in this area.
Kirsten Thorpe spoke next, saying that what interests her most is connecting people with knowledge, and that sometimes description is what isolates people from gaining access. She also noted that many records about vulnerable communities can’t be put in a public space. Instead, it needs to be put somewhere that communities can give consent to access. People need to have a say about rights management – which leads to questions about whose records they are, whose knowledge, and who owns the interpretation. Kirsten pointed out that Chris’s proposal raised the question of what to do with these other types and sets of information, where communities might say “This material shouldn’t belong to anybody but us.”
At the same time, we have to find a way to connect the dispersed information, or people won’t know these rich, vast collections exist. Kirsten talked about returning material to indigenous communities, telling them “this is your stuff” and ensuring the right protocols are used so that the communities aren’t harmed again by our actions. She asked “What does “the looking glass” (Chris’s “wonderland”) look like to indigenous communities?”
After this there was general discussion, about what risks we could take, and how we could progress the discussion.
Luke Bacon suggested that perhaps the interface itself wasn’t the best contribution that archivists could make, as others have more skills in this area. The key points archivists should work on would be providing access to the huge amount of data they hold. There could be many different interfaces, and archivists could engage with communities (and people with IT skills) to develop these.
There was discussion about “ownership” of data – once it is available on the internet it is open to everyone. However, privacy and copyright legislation permits government archives to have privileged positions, as do libraries and museums, but not archives as a whole, which leads to issues with data sharing and reuse.
It was suggested that it is important if setting up a federated structure that it not be just creating another, virtual institutional archive – records that belongs to individuals and communities should not be corralled into an institutional space, perpetuating the imbalance of power that already exists. There was general agreement that the discussion of who has power over data once it is released into the internet is an important one: people are already using and reusing it as soon as it is available. Many questions remain, how to protect people’s rights, and how do the new stories get told? Perhaps by decentralising the power, treating it more seriously, and involving the communities affected in the discussion.
A significant issue with a federated space is who controls the structure: the specification for the structure must allow for multiple and contested views. Similarly, there was discussion over the difficulty of dealing with authority records: to be able to view from multiple perspectives, “authority” needs to be contestable. To deal with multiplicity, it is necessary to support the contested territory, and honour the different narratives. In many cases, consensus cannot exist. There will be some collections that are not acceptable for some institutions, and some descriptions that are not acceptable to some groups – our system needs to be able to transcend political boundaries. What we need to build is not a repository model, but an infrastructure to allow multiple types of access and use – a network of networks. This may answer some of the questions we have about sustainability, authority and ownership. This would move the archival role from one of custodian to one of steward, whose description is not privileged above that of anyone else.
Currently, many archival interfaces are developed with the researcher in mind, but a more open archival system would cater to all people who need records.
It may be that some records are described only at a very high level, so that there is general information about their existence, even if the material itself is not available to the wider public.
It was agreed that any system we design must support multiplicity, and allow people to arrive at the records they need from any direction, any description that is meaningful to them. We need to remember that interpretations shift over time, and controversial views today may be mainstream tomorrow, and vice versa. The only way to deal with the constantly changing resources and interpretations available will be to outsource much of the description. This will also enrich the description, as we have only our perspectives, and need to reflect the reality of a messy, constantly-changing world. Records are not fixed or static, and they cannot be pinned down to a single authoritative form.
It was recommended that instead of beginning with working on functional requirements for such a system, it would be more important to start with the principles and rights of description. From these, the functional requirements should flow. The question that needs to be answered is “What do records do in people’s lives that nothing else does? We need to work this out before federating existing description, which already has a lot of problems. People need access to the records they need in the way that they need to help them in their lives, and we need to design systems that can help with this. We need to stop just thinking about scholarly access.
People are already doing this, without archival input, linking open data, and we need to get involved in the discussion. We can start now as an open-source project, and let it grow organically.
At the end of the discussion, the first step was taken, when Luke Bacon set up a GITHub where we can begin with a set of archival first principles, and then build from there – the link is repeated here. It is already beginning to grow.
Great summary Belinda 🙂
Here’s more info on the Indie Web community:
* HQ – https://indiewebcamp.com/
* Formation story: https://www.youtube.com/watch?v=Au4bHb5twF4
* Principles: https://indiewebcamp.com/Principles
* More videos: https://ind.ie/summit/videos/jeremy-keith/ , https://indiewebcamp.com/videos_about_the_IndieWeb
Just one clarification:
As Chris says, he is suggesting the Wiki as a conceptual model for a system the could be developed. I’m suggesting that Git as an alternative conceptual basis for consideration, because it provides a bit of a different take on ‘multifaceted’, ‘distributed’ and decentralised. I don’t think either of us were suggesting that the system would BE a wiki or a git repo. The repository I setup is really a place for continuing this discussion and an entry to git for people who want to dive in, not necessarily a starting point for the project itself. Interesting to note that the indie web community do all there comms through IRC chat and a wiki, regular events to share and help each other build stuff.
I caught up with Richard LeHane at the Global Legislative Openness Week Hackfest in Sydney this weekend and he highlighted that while its important to get down fundamental principles and rights, it’s also easy to eternally discuss them. New systems are being developed as we speak and if archivists don’t get involved. I think that was an important point at the forum as well—not to be forgotten.
If people would like to discuss these principles, there is the github repo for that discussion, but it would also be nice to focus on developing something sooner rather than later :).
It was a pleasure to be part of the event, very grateful to be present.