Let Me Make the Songs

Epsilon Theory may have gotten its formal start in 2013, but the soft launch took place in 1991. That’s when my partner Ben finished his doctoral work at <lifts nose slightly> a school up in Cambridge. His work there (and his 1997 book, Getting to War) focused on how news could be used to predict the likelihood of war. It also analyzed how news was used by institutions and individuals to foment an appetite for and belief in the necessity of war, which is related to but not the same as the ability to use that news for predictions.

So yes, we were interested to read about Project Cassandra, a collaboration between German academics and military leaders to quantify the risk of conflicts using…literature. The Guardian covers it here in what I think is a very worthwhile read.

The project is really interesting. It confirms, or at least shares, many of the core principles of our Narrative Machine project. Not least among them is the recognition that oft-mined sentiment possesses a fraction of the influence of memes with subjective power that can only be identified objectively through words and phrases of meaning. Or the observation of identifiable, recurring story archetypes.

Leaving aside that we focus on different universes of creative output, our projects DO differ in one respect. The Cassandra Project appears to place great emphasis on the ability to quantify the popular and critical acclaim of a fictional literary work. For example, more popular, more viral, more well-regarded work would have different influence, which is a perfectly sensible hypothesis.

Because we focus on a larger and more frequent dataset of news, blogs, press releases and transcripts, we get to do two different things: first, instead of estimating how much of an audience has heard an idea and how much that idea has changed their minds, we can more easily observe how much a specific memetic idea has reproduced within the dataset over time.

Second, because our dataset includes items which are nominally news / non-fiction content, we can observe the extent to which explicit and implicit opinion language is being deployed, which also provides a window into the narrative-shaping efforts of Common Knowledge missionaries.

Two different approaches to a similar problem. I prefer ours, because I feel more confident saying that an idea has reproductive legs by observing how it changes the language other people use to discuss it, and because knowing some of the intentions of powerful institutions and influential individuals is an indispensable part of any effort to quantify the potential effects of narrative.

But there are absolutely features of the zeitgeist which will only manifest in the arts, in literature, and in the groanings and yearnings of those who make the songs. For that reason alone, we think the Germans have the right of it here in the way that the Scots did before. It’s a very cool project.

Let me make the songs of a nation, and I care not who makes its laws.

Andrew Fletcher, Commissioner of the Old Parliament of Scotland (1703)

— Rusty Guinn | July 8, 2021 | 12:32 pm