Thanks for this response! I think another metric would be great, and would certainly get at the finer details of this type of work better than looking at dialogue alone. I actually just got back from about a 2 mile walk where I really tried to think about how to measure something as ephemeral as “plot line importance”. I came up with a few ideas, but by the end of my walk, I had already decided against every one.
Regardless, here were my thoughts:
- You could look at character connectivity through some sort of social network analysis. The more important characters to the plot are arguably the ones that are connected to the most other characters. However, I shot this down when I thought of Pocahontas. I’m not sure why that was the first example that came to mind, but if memory serves, Grandmother Willow, who was incredibly important to the plot, only ever interacted with Pocahontas and, eventually, John Smith. She would be considered not important using social network analysis, which is inaccurate.
- In reading up on the Bechdel test before doing this project I stumbled upon a (semi-satirical) “Sexy-lamp test”. The author suggests “If you can replace your female character with a sexy lamp and the story still basically works, maybe you need another draft”. So, I wondered if that could be implemented (with some variations), but it’s too subjective. If a character only spoke two words but those words were “Look out!” so that the main character didn’t get hit by a bus, you couldn’t replace him/her with a sexy lamp, but that doesn’t make them an overly important character to the overall story.
- Another derivative of the Bechdel test is the “Mako Mori” test, which basically asks whether a female has a story arc that isn’t about supporting a man’s story. But what counts as “supporting a man’s story?” Would Ava from Ex Machina (great reference, by the way) pass that test? She did support a man’s story but had her own story arc too.
Concept testing is a great idea, but it’s so hard to do in practice. We try to quantify things so that we can directly compare them and end up using less-than-perfect metrics because it’s all we’ve got.
In the Reddit thread about this visualization, a user (nukacola) made an interesting comparison to body mass index (BMI) that I think articulates this point well. They said “The Bechdal test is a bit like BMI. There are people who are healthy despite having an overweight or obese BMI. There are movies that promote equality despite not passing the Bechdal test. There are people who are unhealthy despite having a normal BMI. There are movies that pass the Bechdal test but are incredibly sexist. A person having an overweight BMI is still a pretty good indicator of health issues despite the exceptions. A movie not passing the Bechdal test is still a pretty good indicator that it might not be completely on the up when it comes to equality, despite the exceptions. And most importantly, if the majority of the population has an overweight or obese BMI, it’s indicative of some pretty serious national health problems. If the majority of movies fail to pass the Bechdal test, it’s indicative of some pretty serious cultural issues.”
We do what we can with what we have :)
That being said, if you come up with or find any other metric, please let me know. I’d love to check it out!