Is Horror Dead?
Film data researcher Stephen Follows (whose work I admire—seriously, go subscribe to his Substack and get a copy of The Horror Movie Report) recently argued that the label “horror” might no longer be useful:
The word horror no longer tells us what a film is trying to do, how it feels, or who it’s for, meaning that the term alone has stopped being useful.
Stephen is nothing if not data-driven, so his claims tend to be pretty rock solid.
But I think he’s wrong about this one.
And it’s not entirely his fault. Horror is poorly defined and even more poorly understood.
But a poor understanding of horror doesn’t make the term obsolete, nor does it mean horror lacks a common core that tells us what the film is trying to, how it feels, or who it’s for.
It does all of those when understood properly.
What is Horror trying to do?
I believe what Stephen means here is, “what is horror?” It’s an historically contentious question with no clear answer — at least, until now.
I’ve written some initial thoughts about what horror is as well as a more refined, evolutionary-based theory of horror.
The TL;DR of my theory is that horror is, at its core, about a power asymmetry between the protagonist and the antagonist. In horror, an malicious antagonist is always much more powerful than the protagonist. I’ve developed a formula you can use to calculate the Protagonist Vulnerability Index (PVI) to see how “horror-centric” a film is.
My colleague Edgar Dubourg and I put my theory to the test recently, so I even have some data to back this up. I was going to wait until we had a preprint up to talk about this project, but now seems like a good time to showcase some of the results.
First, we took a set of over 600 films from IMDb that have viewer personality data associated with them (from Nave et al., 2021: one of my favorite big data studies on film and personality).
Study 1
We then fed detailed prompts to ChatGPT so that it would annotate each film for every variable of interest, a method that Edgar has pioneered. After a few iterations of the prompts, we had detailed data and ratings for the formidability of the protagonist, formidability of the antagonist, hostility of the environment, and overall presence of the threat to the protagonist for over 600 films. It’s the kind of thing that, less than 5 years ago, would have taken 3 research assistants over 3000 hours to do. We did it in about a day.
The first question was simple: when we calculate the horror formula for a film, do films that IMDb categorize as horror have higher horror index scores (PVI) from my horror formula?
Yes! Films categorized as horror on IMDb have a higher PVI as calculated with the horror formula. In particular, protagonists are very weak and antagonists are very strong. Threats are present fairly often (though a bit less than thrillers and mystery) and the environment is pretty hostile to the protagonist.
This suggests that the horror formula is tapping into something at the core of the horror genre.
How does Horror feel?
If you ask most people what defines the horror genre, something about fear will come up. It’s certainly the emotion most associated with the genre. While I don’t think “causing fear” is a necessary feature of horror, there is a good reason it’s a common one. My claim about horror is that it’s the genre of story where a malicious antagonist is far more powerful than the protagonist.
In nature, that scenario is terrifying.
To the extent that we empathize with the protagonist (which depends on a lot of things, from narrative structure to personality of the viewer), we should feel afraid when we experience a horror story.
We’ve all felt afraid before, and we all know what the feeling is like: sweaty palms, a racing heart, quickened breath. These are physiological changes that occur in response to increased sympathetic nervous system activity, often referred to as a fight or flight response.
Because these physiological changes are part of the universal nature of fear, we can pretty easily measure fear using something like heart rate. There are of course caveats here, but heart rate is one of the best, and simplest, measures of fear.
The Science of Scare Project uses this fact to put together a list of the scariest horror films each year. The logic is pretty straightforward: films that elicit a higher average heart rate in viewers are probably scarier.
To take our horror formula test to the next level, we decided to see if horror films that have higher scores (higher PVI) also scare viewers more, as measured by heart rate (BPM).
And they do!
Taken together, these two studies provide a pretty compelling argument that we can identify what horror is trying to do (tell a story about a protagonist who is vulnerable to a formidable antagonist) and how it makes us feel (it scares us). The more horror centric a film is, the more it does both of these things.
Who is Horror for?
My instinct here is to say “everyone,” but that’s not helpful.
We have decent data on the personality of horror fans. For example, your level of morbid curiosity is the single strongest predictor of how much you enjoy horror films (at least among traits that have been tested). Horror fans also tend to be younger, and somewhat higher in intellect/imagination and sensation seeking.
Edgar and I also found some really interesting data on personality using the horror formula, but I’ll keep that a surprise until the paper comes out.
All this to say, we have as good of an idea of “who horror is for” as we do of who comedy is for or who action is for.
So is Horror dead?
If we continue to use the term horror without a good understanding of the underlying psychology, then yes, horror might be dead. But there’s no reason this has to be the case. New research like the kind Edgar and I are working on can help redefine the genre based on its core elements.
We need to understand what the psychology underlying the horror story if we want to identify what horror is trying to do, how it makes us feel, and who it’s for.
Horror doesn’t need an obituary, it needs science.
Maybe my next book should be on this.