The reverse-Bechdel test


Categories: Personal Politics

The “Bechdel test” is a test which categorizes movies based on a simple criterion: At some point in the movie, is there a conversation between two women which is not about a man?

This is obviously not a flawless heuristic, but intuitively, it seems like a movie in which this doesn’t happen even once might well be underrepresenting women in roles other than “prospective love interest”.

Lots of people have written about this, and periodically someone says “but what about applying this to other groups”, and gets yelled at for distracting from the point. Well. For me, a number isn’t a point or a support for a claim without context. I likes me some context.

So I’m going to do a simple experiment: I’m going to pick twenty movies I’ve seen and check whether they have a scene in which two men discuss something other than a woman. (Prediction: 100% pass rate.)

Batman: Under the Red Hood: Pass.
Black Snake Moan: Pass.
Sherlock Holmes: Pass.
Kick-Ass: Pass.
Yojimbo: Pass.

Seven Samurai: Pass.
Raiders of the Lost Ark: Pass.
Scott Pilgrim vs. The World: Pass.
12 Monkeys: Pass.
Unforgiven: Pass.

Cyrano de Bergerac: Pass.
Iron Man 2: Pass.
The Hudsucker Proxy: Pass.
Cowboy Bebop, the Movie: Pass.
Time Bandits: Pass.

Snatch: Pass.
The 13th Warrior: Pass.
Crouching Tiger, Hidden Dragon: Pass.
Iron Monkey: Pass.
The Long Kiss Goodnight: Pass.

Chocolate: Not sure, actually. I’d have to rewatch it. Very light on conversations.
Harold and Maude: Not sure.

Okay, without going back and rewatching, I had to skip two movies. But that’s 2020 on things I’ve seen. Pretty sure Harold and Maude is a pass, but not quite sure.

Conclusion: Just from this quick informal review, I think the test is reporting something significant when it observes that many, many, movies either don’t have two women, never show those women talking, or they talk but only about a man.

Comments [archived]

From: Chris
Date: 2012-02-27 20:12:29 -0600

Seems you are skipping something here, or misstated something. Your quick informal review did not show that many movies either don’t have two women, never show those women talking, or they talk but only about a man. You were looking for “whether they have a scene in which two men discuss something other than a woman.” They all passed, they all showed two men talking about something other than women.

All of those movies you looked at could also have had two women talking about something other than a man, but you didn’t look for that. So we don’t know if they passed the two woman test.

From: seebs
Date: 2012-02-28 01:47:12 -0600

That’s right. I felt the original test had been adequately explored. What struck me as an open question was: Are there explanations other than a systematic difference? For instance, imagine a parallel universe in which very few movies feature any two characters of the same sex discussing something other than a member of the oppposite sex.

The low rate at which movies pass the original test has been covered; what I wanted to do was frame that number with another number, because that struck me as a quick way to get a feel for what the existing number means.

I picked my Netflix queue to avoid or at least mitigate availability heuristic bias.

My conclusion is that nearly all movies (90% or more?) appear to include at least one conversation in which two men discuss something other than a woman. By contrast, other people have shown that not all that many movies (I think less than half) have a scene in which two women discuss something other than a man.

(Just off the top of my head, at least one of those movies passes the original Bechdel test, and at least a few fail.)

From: Chris
Date: 2012-02-28 06:51:53 -0600

I think you may be missing my point. Just because the movies pass one sort of Bechdel-n, test, does not mean that they fail another sort of Bechdel-n test. In the original cartoon test, it’s only indicative of a single movie. Sure, two men may meet and talk about (x-women), but in the same movie, two women may meet and talk about (x-men). Your test actually doesn’t demonstrate what you concluded.

The test applied broadly, does only one thing, it looks at movies like this: (all movies)/(all Bechdel-n movies). It’s only looking at the ratio, and comparing it, not to movies or media, but to a concept of “appropriate” representation of whatever Bechdel-n is in the society’s population. If Bechdel-n is women, and women make up 50% of the population, is it appropriate that they make up 5% of the movie or media offerings?

From: Daniel Schulz-Jackson
Date: 2012-04-30 09:20:56 -0500


The Bechdel (and Reverse Bechdel) are macro. The point is to look at the entire industry. When half of movies fail the Bechdel and <10% of movies fail the Reverse Bechdel, you know you’ve got a society-wide (or at least Hollywood-wide) issue of female underrepresentation. Plain and simple. The point isn’t to evaluate the helpfulness or fairness of individual movies.

Of course, the best rigor and detail for the Bechdel would be to sort movies into four groups:

Bechdel/Reverse Bechdel
1. Pass / Pass
2. Pass / Fail
3. Fail / Pass
4. Fail / Fail

That would shed better light on the whole thing. I really do think that people reporting Bechdel results should also report Reverse-Bechdel results as well, to set a baseline to show JUST HOW SIGNIFICANT the Bechdel results are.

So I’m really glad to have found this article, and would love to see more study with more rigor in this area.

From: seebs
Date: 2012-04-30 11:45:30 -0500

Half of movies fail the Bechdel? I thought it was more like 80-90%.

And yes, Daniel’s expressed what I was aiming for; we already know that movies tend to fail the Bechdel test, but without a comparator to something else, that’s not usable information.

From: Xuncu
Date: 2012-07-25 15:56:45 -0500

If you’re gonna do a Reverse Bechdel, shouldn’t the movie also be reversed?
ie: a “chick flick” that’s almost all female cast, where if there’s a guy—- hm, thinking what’s more effective if he DOES talk about a relationship/girl, or if it’s one of thsoe cynical movies where all he talks about are sports and beer or something.

Like one scene, the gils are on their night out, the guy is at home, channel surfing, then suddenly, Discovery Channel and Carl Sagan telling everyone why Creationists are full of shit.
BAM. Instant Cult Classic.

From: Raziya
Date: 2012-11-08 17:30:20 -0600

If you want to find examples of films that fail the reverse Bechtel test, perhaps you should look at a woman’s Netflix queue?

From: seebs
Date: 2012-11-08 22:39:32 -0600

That’s an interesting thought; I don’t know that I know many women who have netflix, but it’d be an interesting test case.

From: Rae
Date: 2014-04-08 01:23:43 -0500

I recently learned about the Bechdel Test and it helps explain why I have trouble relating to a lot of the films that fail this test. This includes a lot of sci-fi films like Lord of the Rings, as well as Romantic “chick flicks” where the only conversation between women is about the man they have fallen in love with. One-dimensional female characters are boring, difficult to relate to, and not a realistic representation of how women actually are.
Here are a couple interesting sources that grabbed my attention on this topic, including some stats: