Sorting the past
Its publication week for our new book Social Media and the Automatic Production of Memory: Classification, Ranking and the Sorting of the Past
Sometime around the late spring of 2019 I started working on a journal article with Ben Jacobsen. At the time Ben was working on a bigger project about algorithms and memory, the article we began to develop together was intended to be a small spin-off from that bigger project. At least we thought it was going to be a small side-project, it turned into something quite different.
The aim was to look at how classification and ranking processes facilitated the delivery of social media memories. We wanted to think about the archiving processes that lead to those notifications telling the social media user that they have a “memory". We also wanted to focus on the way those classified and ranked memories are received. It seemed to us that this was really important for understanding how social media might intervene in what and how we individually and collectively remember. This, of course, has implications for identity, social connections, shared experience and social ordering more broadly. We set out to think about the algorithmic archiving structures of social media memory-making and how people responded to them.
As we began working on the article it started to mutate. We just about managed to contain it in an original 10,000 word article version (which was already longer than standard journal artucles). Inevitably, we were then faced with a need to split the material up in order for it to work as a single article. We even considered splitting it in two. The problem with this was that we knew that it was the relation between classification and ranking that was the really important thing, and so we wanted to elaborate on those relations rather than carve them apart. We also wanted to keep the processes alongside the social media users’ responses to them. More space was needed to manage these relations and to explore them. This was the dilemma we faced when we worked again on the article in 2020. Ideas don't always lend themselves to typical article-sized packages, however happy you might be to edit and tighten the writing.
When we returned to the article in 2020 we could see that there was too much there for an article and we remained reluctant to chop it down. During a Zoom meeting we decided that rather than split it or try to contain it we would gamble on letting the ideas breath by turning it into a book. We didn't know if it would work as it hadn't been planned as a book, but the opportunity to give it a go was appealing. So, collaborating on a Google Doc (sometimes with us writing synchronously) and occasionally meeting to discuss our progress on Zoom, we began to develop the ideas and expand the existing sections that we had put in place. Rather than limit the data and the concepts we just took off the brakes. It quickly expanded and we realised that making it into a book had created space for us to explore the issues and concepts in the detail they needed. This shift allowed us to open things up rather than trying to close them down. It also meant that we didn't need to try to separate-out these related aspects of algorithmic memory-making. Instead we could explore the relations themselves.
And so, after some revisions following reviewers comments, we ended up with the book Social Media and the Automatic Production of Memory: Classification, Ranking and the Sorting of the Past, which is published on the 1st of April by Bristol University Press.
Following a contextual introduction about how the logic of social media becomes part of the circulation of memories, the book itself has three main sections. The first looks at the way that past social media content is categorised and classified. We look here at the way that content comes to be labelled as a memory once placed into a taxonomy of memory types. We develop this through a notion of partitioning. The second part then looks at how these classified memories come to be ranked. This section looks at what we call memory promotion. It is through this ranking that the memories are sorted in terms of embedded notions of value. This is also how the masses of past content are selected to become visible as packaged “memories". We thought it was interesting that not even something as intimate as a memory can escape the logic of ranking. Finally, in a third step, we then looked at how people respond and react to these classification and ranking processes. This revealed not just how people felt about their past being sorted, but also indicated how people shape their memory practices in response to their circulations in social media. This is the longest section of the book, where we track four particular sets of issues that we found as the automated classification and ranking of memory is experienced in everyday life. The book closes by reflecting on what this means for our relationship with the past. In addition, it also provides insights into the more general experiences of the politics of archiving, classification and ranking.
We use a short fragment from Walter Benjamin to frame the book. One point that Benjamin makes in that fragment is that memory-making requires active digging. This active digging and marking-out is, Benjamin suggests, where memories get their meaning and legitimacy from. This led us to wonder where this meaning and legitimacy might arise from when memories are dug-up and sorted on our behalf by automated systems.
If you are interested in the book, it is available to order in hardback, with the Kindle and ebook available on the 1st of April. More information can be found on the book's page at Bristol University Press, where it can be purchased at a discount. That is the best place to order it from. It's also available from the usual outlets. The Kindle and ebook versions are particularly well priced.
Here is a Spotify music playlist I've created to accompany the book.
If you think you know of anyone who might be interested in the book, please do share this with them.
Here is the book blurb:
Social media platforms hold vast amounts of biographical data about our lives. They repackage our past content as ‘memories’ and deliver them back to us. But how does that change the way we remember?
Drawing on original qualitative research as well as industry documents and reports, this book critically explores the process behind this new form of memory making. In asking how social media are beginning to change the way we remember, it will be essential reading for scholars and students who are interested in understanding the algorithmically defined spaces of our lives.
And here are the endorsements from the back cover:
"This book illuminates how our memories are formatted by the interfaces of social media. Jacobsen and Beer have written an insightful analysis of what ‘technologies of memory’ look like in data-driven societies." José van Dijck, Utrecht University
"Jacobsen and Beer’s book is essential reading for anyone wanting to understand the automatic production of memory on social media. This book is one to be remembered and revisited time and again." Taina Bucher, University of Oslo.