I spent most of today in a large meeting room, running a workshop for a few members of a large government department. We were talking about how to get control of your records, resources, documents, files – or, to use my default word while training: stuff.

A key focus was the importance of collections. When people are asked to document stuff – particularly digital stuff – to make it usable and searchable in future, many dive right down to the lowest level of detail and try to start at the item. They attempt to extract metadata, add keywords, tag things, or move things into different folders. Often overlooked is the importance of the collection.

By contrast, I’ve spent a lot of time working in and around collections of various sorts. I started out as an art historian and have spent a lot of time in galleries. I’m a fan of museums and museology. And, as an archivist working across many different contexts, I feel like collections are part of what I do. So, when working with people from different contexts, how do we describe their value?

Collections are greater than the sum of their parts. Or, turning that idea the other way around, individual things (records, files, artefacts, documents) are not as useful, discoverable or understandable when considered in isolation. They are contextualised by the things around them and the context(s) in which they are found.

And, from a practical resourcing perspective, getting control of a collection is (relatively) straightforward. Listing 2000 files individually takes time and effort and still potentially only tells you information about 2000 individual items. You don’t find out where this stuff came from, who brought it together, or when, why and how this stuff was collected. A good collection description tells you this, and will at least help you discover useful places to look, even if it won’t help you discover exactly what’s inside.

I know archivists know this – that’s why there are a bunch of ‘collection level descriptions’ in many archives, with resources for more granular description assigned based on demand, specific funding or perceived importance. But many people put in charge of ‘stuff wrangling’ (which is one of my plain-English job titles) in organisations, be they knowledge managers, CMS [content management system] developers, information asset co-ordinators or web publishers, seem to skip over this and go straight to the files.

There’s a broader issue here too. Organisations (and collecting institutions) deal with hundreds, thousands, hundreds of thousands or millions of records. Numbers of records beyond the capacity of people to wrap their heads around. Search and filter technology helps whittle this down, but only if you know what you are looking for. Otherwise, discoverability is seriously compromised by just bagging and tagging digital files for storage in a single big bucket.

Humans understand the world through collections (and through context entities, but I’m not going there in this post). We create collections at home – photographs, files, books, music, tax receipts, ceramic dolphins, art, ukuleles – and we create collections at work. We have filing cabinets, project files and libraries. Project teams have a space on the server. Divisions and departments have lists of resources. I have piles on my desk. These human-scale collections are relatively easy for us to understand and wrap our heads around. So, when trying to get control of stuff – physically, digitally and intellectually – collections are the place to start.

There isn’t any new thinking here. And there’s a lot missing. (For example, massive collections, well beyond a human’s capacity to understand, open up possibilities for algorithmic analysis and visualisation. But I would prefer to call these ‘aggregations of content’ or similar.) At the same time, we (from the GLAM sector) can’t assume these ideas are self-evident. As people and organisations move more into digital technologies and automated systems, we have a role to play in helping them understand and benefit from the power and utility of good collection descriptions written by humans, for humans.