A month ago, I attended a masterclass with author Candace Savage. It was a primer on historical research and how to use fragments of data – old photos, genealogical records, land titles deeds, newspaper clippings and family keepsakes – to construct a rich narrative history. It was built around Savage’s experience researching her book Strangers in the House which was inspired by photographs, drawings, newspaper clippings and memorabilia that she found inside the wall of her early 1900s home.
It was fascinating to look at the physical documents and see how clues encoded in these artifacts led towards a story. At one point, a friend who was also attending the session deduced that someone left out of a will might have been an illegitimate child (she was right!). I’ve never done historical research, but I do remember the days when writing an essay meant that you had to go to the actual library and scroll through microfiche or pour over a stack of periodicals to find your source material. It’s a far cry from how I do research now - sitting at the kitchen table entering key words into a database.
It was during one of these kitchen table sessions that I landed on Colin Koopman. The title of his latest book, How We Became Our Data: A Genealogy of the Informational Person, caught my eye. I added it to my list of books to read but in the meantime, I went down the internet rabbit-hole into Colin Koopman’s work, including a talk that is an overview of the book, a blog post about the book, another talk, an article, and a paper. This is how my research tends to unfold. I get obsessed with a person or an idea.
The interplay between our self and our data is a central theme I’m trying to unpack as I explore artificial intelligence and ethics which is why I was so excited to find Koopman. His work explores the underpinnings for the “datafication” of our lives that date back to the turn of the 20th century. This was a time when documentation started to develop standardized formats, to be stored in ways that made it easily accessible and it was during this time that we started to systemically perform data audits, Koopman explains. The idea of counting or auditing information was formalized in Canada in 1918 with the Statistics Act which created the Dominion Bureau of Statistics, which became StatsCan. The systems and processes that we take for granted today that form the structure for the data necessary to enable computing, the internet (and now AI) were being put in place roughly a 100 years ago.
Koopman’s talk on algorithmic culture traces much of our personal data identity to the system of developing standardized birth certificates. It’s hard to imagine a time when people were born without this corresponding piece of paper being issued. A birth certificate is now a ubiquitous piece of data that defines us. It forms a central piece of our documentation and is used to then establish other core documents, like a social security number, passport or driver’s license. It’s necessary to be documented to participate in our society. Being undocumented is a problem, which, sadly, many people in the United States are trying to navigate. Personal data becomes a piece of both who we believe ourselves to be as well as who others believe us to be forming what Koopman calls “the informational person”.
There were many inconsistencies in the historical documents that Candace Savage shared with us. Some of the documents looked similar but names were misspelled, the dates for key events didn’t always align and categories weren’t consistent. Personal information at that time was in the process of becoming standardized but not yet made standard. While it’s certainly possible for documents today to contain errors or inconsistencies, this can cause huge problems. You can’t get on an airplane if the name on your ticket doesn’t EXACTLY match what’s on your ID (I speak from experience) and if you lose your ID while travelling, it can cause a world of panic (I've been there too).
I find all of this fascinating. In trying to understand the ethical considerations for AI, I think it’s important to understand the historical and socio-political context surrounding data. It provides perspective on the systems we’ve built so far, systems that seem natural but are actually constructed and built over time. I have a feeling I'll be posting more from Colin Koopman once I've had a chance to read his book.