Beyond Bechdel

In a recent AP article about a new Swedish film initiative based on the Bechdel test was the following section:

"The entire Lord of the Rings trilogy, all Star Wars movies, The Social Network, Pulp Fiction and all but one of the Harry Potter movies fail this test," said Ellen Tejle, the director of Bio Rio, an art-house cinema in Stockholm's trendy Södermalm district. 

Bio Rio is one of four Swedish cinemas that launched the new rating last month to draw attention to how few movies pass the Bechdel test. Most filmgoers have reacted positively to the initiative. "For some people it has been an eye-opener," said Tejle. 

Beliefs about women's roles in society are influenced by the fact that movie watchers rarely see "a female superhero or a female professor or person who makes it through exciting challenges and masters them", Tejle said, noting that the rating doesn't say anything about the quality of the film. "The goal is to see more female stories and perspectives on cinema screens," he added.

Actually Pulp Fiction and all of the Harry Potter films pass the Bechdel test, and the test is not related to whether a film has female superheroes or tell "female stories" or not. Ellen Telje (who is a she, not a he) might have said, more honestly, "a number a key feminist films fail this test." but she does not, presumably because that would have complicated the sales pitch.

In a follow-up article, Guardian's chief arts writer Charlotte Higgins writes that "an alarming number of films showing in cinemas fail to reach it" and adds that "[o]ddly enough, Thor (in which Chris Hemsworth plays the Nordic god, come to save us all from Christopher Ecclestone) does pass, since it features a scene in which Natalie Portman and Kat Dennings discuss nuclear physics."

Higgins's claim that it is alarming how few films pass the test is very common and yet the vast majority of the thousands of films listed on the website bechdeltest.com (primarily popular mainstream films) do pass it, between 65% and 70% of the films released over the last two decades. If 70% is alarmingly few, how many are needed for it to be acceptable? Or maybe Higgins has her own statistics. But it is not in the least bit odd, if you understand the test, and films, that Thor passes the test. Any film can pass or fail, regardless of genre, country, year or feminist intentions.

Although the test has some uses it says nothing about the way a given film represents gender, or whether the film is progressive, reactionary, feminist or misogynistic. The test demands that a film has scenes with named female characters talking about something other than a man, but whether the lack of such scenes is an issue or not depends on the film. If a film for example has only two characters, a man and a woman, the test is pointless, and if there are several characters it is still not in itself sexist if the women only speak of men. That becomes an issue if the dialogue between the women is different from the dialogue between the men. If in a romcom or a drama about a failing marriage, with all the men talking about women, it does not matter if the women talk only about men. Context and content matters but the test ignores all that. Using it on all films regardless of their content is like a test that rates vehicles with wheels on whether they have catalytic converters, which diminishes the pollution from internal combustion engines, and where it is then considered shocking that so many vehicles lack the converters (even though the majority of those vehicles are bicycles that have no need for them).

But if tests are wanted, may I suggest a different one, maybe something like this: In the film under consideration, do the women characters have sufficient agency and are they in charge of their own well-being and if not, why? (I say sufficient because we are all more or less constrained by our surroundings and circumstances.) If they are dependent upon the men and lack the capacity to take their own initiative the film would fail the test, unless there is a perfectly good reason for this dependency such as being hospitalised after a car accident and taken care of by a male doctor. Let's call it "the agency test". Unlike the Bechdel test it means that you have to intellectually engage with the film, but that surely is a good thing.

The Bechdel Test

It has been around since 1985, the so-called Bechdel Test, named after Alison Bechdel who mentioned it when making a comic strip called Dykes to Watch Out For. (Although Liz Wallace is the person who came up with it.) It is about the representation of women in cinema or other forms of narrative art, and consists of three criteria, or questions:

1) Are there at least two female characters in the film? (Sometimes they have to be named, sometimes not.)
2) Do they talk to each other?
3) Do they talk about something other than men?

If all three are answered Yes then the film has passed the test. The argument is that many (perhaps most) films do not pass the test.

It is a quick and relaxed way of analysing a film and it is mentioned every now and then, the other day in a long article in the New York Times. I am however somewhat puzzled by the attention it has received, on blogs, in mainstream media and in Academia. I am puzzled because the test has so many problems.

The first, less important, point is the vagueness of the third question. What does it mean to talk about a man? If two lesbian women discuss their adult son, does that qualify as talking about a man, or if two sisters talk about their dying father? If two female physicists discuss Albert Einstein, is that disqualifying?

A second point is that it is based on memory, which is unreliable, and consequently a lot of films are disqualified even though they have named female characters talking about something besides men. Unless you are actually writing down exactly who says what to whom, you will not remember it all. On the website bechdeltest.com 3550 films are listed (2012-12-09), and for each film it is specified how they score, from 0 to 3. To take one example, John Ford's Stagecoach (1939) got a 0, despite having female characters (named) who speak to each other about other things then men, and hence should have a 3. According to the wikipedia-entry for the test, the TV-series Sex and the City  fails the test. How is that possible? During the six seasons the women talk about pretty much every thing known to woman. The only way SATC fails is if the rule stipulate that women are not allowed to talk about men at all, at no time, but that is not the case.

But the important point is that the test is completely devoid of context. The test on its own says almost nothing about the film. Let's say that we have on the one hand a film called Anal Sluts Tour America which begins with Chrystal and Tiffany discussing their breasts, and then they spend the rest of the film fulfilling the promise of the title. This film would pass the Bechdel Test without a glitch. Then let's say we had a film called Silent Suffering, about domestic violence. It is set in an apartment over a weekend, with a man tormenting his wife, until she manages to escape. In the last scene we see her going to a police station. This film would completely fail the test, and get a 0.

The importance of the three points depends upon the individual film. If it is a romcom and the male characters only talk about females, then it would not matter if the females only spoke about men. On the other hand, f it was a political drama where the men spoke about law, affirmative action and filibusters, but the females only spoke about their boyfriends' annoying habits, then it would be an issue and a problem. Another problem is that since the rules are so vague, a lot of films pass even without real merit. As an example, a female characters might buy a latte at a Starbucks from a female barista and they talk about coffee, and so get a 3, regardless of what happens in the rest of the film.

But there is something else that is a bit puzzling. It is almost always said that so shockingly few films pass the test (as A.O. Scott says in the above linked-to article). But is that the case? Surprisingly enough, on the bechdeltest-website most films actually pass the test. 54% of the films listed pass the test completely, and another 11% get 2 out of 3. But, as the case of Stagecoach shows, these figures are not trustworthy and it is more likely that films are "better" than it seems. (The number of films that pass the test is even higher, about 70%, if only films made the last two decades are counted.)

And finally, what is the goal? Let's say that the figures are correct, that 70% of all films get a 3. Is that good or bad? How many should get a 3 for us to be able to relax. After all it cannot be the rule that all films should get a 3. Even films that get a 0 are perfectly legitimate. If I wanted to make a film about two gay men who going sailing together I would get a 0, but that would be perfectly fine. Or films about monks, or soldiers in World War 2, or cellmates.

Some have argued that it is because it is so crude and simplistic that it is valuable, since it shows how despite the crudeness so few films actually pass. But I cannot agree with that. Partly because, as I said, it would depend on the film whether the score was meaningful or not, and partly because so many films actually do pass.

Misogynistic films are a real problem, and the lack of female filmmakers and good roles for females is shameful (particularly for older women). But the Bechdel Test will not help solve those problems, neither will it help finding those films that are problematic. It is not how a film is rated according to some test, it is how it deals with characters and events that is meaningful. No random numbers can tell you that. In the end, the problem with the Bechdel Test might perhaps be that so many films actually pass it.

------------------------
2013-12-10 (Slightly amended to include more recent percentages of films that pass.)