After several heavy weeks, I’m writing to you in a much lighter state of mind.
I’m on the train from Ottawa, admiring the frozen scenery (frosty grey swamps are maybe my favourite backdrop for a train trip?) and setting no expectations for myself other than to enjoy some books. (Last night, to ease into this state of relaxation, I finished Elena Ferrante’s Neapolitan novels and wrote a brief reflection on them.) I’m approaching a true vacation for the first time in a long while, and I’m ready for it.
But before I throw off completely the last vestiges of this term, I want to share a project I worked on for my course in digital history. (A course about which I’ve previously written.)
Given the space to choose both a subject that interested me and a tool that I wanted to learn, I decided to investigate what topic modelling (a tool for textual analysis) could indicate about the changing public priorities of Canadian federal governments, using speeches from the throne as my corpus (the body of text on which I ran the tool).
Because I’m a digital dork, I made a website to share my project findings, where you can read the full writeup if you’d like. (If you want just the interesting analytical bits, see the list of topics and discussion of results.)
To save you from that, though, I’ll highlight two top findings that interested me:
- Mentions of “economy”, “women”, and “environment” bunched together under one topic that has gradually increased since the 1980s. Digging in a little further, though, shows that not only has the usage of words like “women” and “environment” increased, the meanings attributed to them have changed, too. For example, early mentions of words related to “women” referred to women generally, while more recent mentions have referred to specific groups of women. (Fuller description of that observation in the linked topic annotation.)
- Over time, the tone of speeches from the throne seems to have dropped in formality. (This was the most exciting observation, for me, because it wasn’t at all the kind of thing I even thought to look for.)
This project isn’t the most rigorous or exhaustive—there’s definitely deeper and more thorough analysis I could’ve done. But I think it’s an interesting proof of concept demonstrating how we might use digital tools to inspire new interpretations of familiar texts. This potential isn’t limited to academic applications: I can imagine using this in various professional contexts to quickly get a sense of the content of a text. Digital tools are “thinking tools” (to borrow a phrase from Graham, Milligan, and Weingart) that can suggest new ideas we hadn’t thought to consider.
Anyhow, I’m away from school for a few months, so my reflections will probably take on a different hue. I hope you can find some moments of calm in this topsy-turvy month of December. All the best for the week ahead!