Saturday, March 10, 2012

My Beef With The Bechtel Test

Kyle had a post up a little while back linking to a video from Feminist Frequency about how the 2011 Oscar contenders measure up on the "Bechdel Test". I forget where I ran into the test before, but I definitely remember it rubbing me the wrong way. As summarized in the video (if, like me, you hate videoblogging, you can read the transcript here):
The Bechdel Test is a very basic gauge to measure women’s relevance to a film’s plot and generally to assess female presence in Hollywood movies. It was popularized by Allison Bechdel in her comic Dykes to Watch Out For back in 1985. In order to pass the test a film just needs to fulfill these three, very simple, criteria: A movie has to have at least two women in it who have names, who talk to each other, about something besides a man. Pretty simple right? I mean this is really the absolute lowest that we could possibly set the bar for women’s meaningful presence in movies.
Now, I gotta say, this just makes me climb the walls. I have a strong dislike for simplistic litmus tests that allegedly determine the quality (or even more generally the qualities) of fiction, and I also dislike "gotcha" tests that are supposed to determine hidden bias. These seems at the confluence of these two, and as a result it takes a certain amount of effort to separate my annoyance from the topic enough to try to come up with some less gut-level reaction to it. Given this aim, I'll start with working through the things I can't accuse the test (in this particular use) of, than then work back to why I have an issue with it as a means of evaluating fiction.

First off, the way that the test is being discussed here is not necessarily that advocated by the comic strip characters who originated it -- where "the rule" was to not watch any movie that didn't fit the criteria. Feminist Frequency (and Kyle) on the other hand, admit that a movie might be quite good while not passing the test. Here's Feminist Frequency again:
Again, to be clear this test does not gauge the quality of a film, it doesn’t determine whether a film is feminist or not, and it doesn’t even determine whether a film is woman centered. Some pretty awful movies including ones that have stereotypical and/or sexist representations of women might pass the test with flying colours. Where really well made films that I would highly recommend might not.
Obviously, there's some level of tension here, because while she insists that some really good and recommendable films might not pass the test, she also seems to see it as a problem when individual films (such as 7 out of 9 Best Picture nominees) do not clearly pass. But still, at least formally the concession is clearly made that passing the test is not necessary to a movie's being good or indeed to having strong and essential female characters.

Also, I feel I should concede that passing the test will correlate fairly decently with films that have more female representation in them. Yes, you can have a woman-centered story with a female main character but structured in such a way that the film doesn't pass, or a story which is evenly balanced between one male and one female character, but in general movies in which a woman or women have a lot of screen time will pass.

This leads to the beginning of my criticisms, however, because although most stores that have a lot of screen time for women will pass, they will often do so rather incidentally. So, for instance, I'm finishing up reading Jane Eyre -- certainly generally considered a "girl book", and one that has such a strong female main character, who repeatedly stands up against the expectations of a male dominated world, that it is considered by some critics to be an example of proto-feminist fiction. Thinking of the recent movie of Jane Eyre, it does indeed pass, but even so probably 90% of the screen time involves either Jane interacting with men or Jane talking with other women about men. Even with the more spacious book, if you cut it down to only the scenes which involve Jane talking to another woman about something that does not involve any men, you get some fairly pedestrian scenes in which she talks to her female cousins (and listens to them argue), talks to Helen Burns about books and life, talks to her teachers and to the kitchen maid and her aunt's house, and talks to her pupil Adèle. You miss virtually all of what's interesting in the plot: the conflict.

Now, someone will be pointing out to me that the claim isn't that the scenes that meet the Bechtel Test are the most interesting in the story, just that if the story has a substantial part for women that it is likely to pass the test. And I get that. But it frustrates me that someone would try to analyze fiction based on a criteria that is clearly and admittedly incidental to what it is they're trying to measure. Measuring something like the percentage of the time a named woman character is on screen, or the percentage of lines spoken by a woman, would, however, but a lot less fun than the gimmicky three criteria Test. And something like "are female characters realistic and important to the plot" is hard to quantify, and thus wouldn't produce a fun site full of people rating movies according to one's criteria.

Aside from the fact that the Bechtel Test is clearly measuring something incidental to what it's actually looking to find, it seems to me that by design it isn't really set up "to measure women’s relevance to a film’s plot" as to determine the extent to which women are a separate and self sufficient group within the plot. So, for instance, despite it's female protagonist the movie True Grit fails the Bechtel Test as applied by Feminist Frequency:
Interestingly, even though True Grit is a female centered story, following the adventures of Mattie Ross struggling to get by in a man’s world, when we apply the 60 second rule the film doesn’t pass. In fact the only exchange she has with any other woman is with Mrs. Floyd the innkeeper and those incidental interactions total less than a minute. This style of film where the female lead inhabits an almost entirely male world, brings to mind the Smurfette Principle which I’ve discussed in my Tropes vs Women video series.
So clearly, what's being tested for (and what Feminist Frequency desires to see in a movie) is not simply that women in the movie be well rounded character or that they be essential to the plot, but rather that they be in a world which is at least moderately separate from men. This seemed particularly driven home to me by her complaint against a movie I haven't actually seen (and thus can't provide my own evaluation of the characterization) but the analysis itself is telling:
Tree of Life is a more experimental film about a boy and his family. It fails the test because the only brief scene where two women talk, the conversation is about the death of the family’s son. While it’s true there’s very little dialogue in the film as a whole, the father and the son do speak to each other on multiple occasions.
So, a conversation between to women isn't real if it's about how they feel about a male who has died. If it was a woman, that would be fine, since they're talking about the death of a male, it fails. (Heck, if they'd talked about designer shoes instead of about the death of a family member, it would have passed.)

Similarly, any film which is tightly focused on a relationship (between a man and a woman) is going to at least flirt with failing.

Now, given that this test originates on a comic stripe entitled "Dykes to Watch Out For" perhaps that's reasonable enough. But if one's interest really is in characters and stories well told, I don't see why one would allow oneself to get caught up in this particular analysis gimmick. I don't think that woman characters are less real when they talk to men than when they talk to women, or when they talk about male characters than when they talk about female characters. Yes, perhaps that fact that many Hollywood films don't pass it says something about the extent to which women are represented as independent characters in mainstream films, but if so only in a very incidental fashion. And given that the majority of films lack any well rounded characters of either sex, I find this hard to get particularly excited about.

20 comments:

  1. I've never understood either this nonsense, or its deeper implication: that women need strong female characters in order to be engaged by a work of fiction. I'm sorry, but I'm more well rounded than that.

    ReplyDelete
  2. Ditto the above.

    Also, you made me really happy; someone else hates videoblogging!

    ReplyDelete
  3. I think that the Bechdel test can be useful in pointing out when the female characters in a story don't have any identity or life outside their relationship to the male characters. (The Girlfriend, The Mom, etc.)

    But yeah. You can have well-rounded female characters who fail the Bechdel test, and sexist stereotypes who pass, and stories that have completely valid reasons for skipping female characters entirely--so it really annoys me when people treat it as some kind of Holy Last Word. It starts turning into the Gotcha school of feminist criticism, which is just ridiculous. (Woman sacrifices herself for her boyfriend--oh, that's so sexist, she has no identity outside of a man. Man sacrifices himself for his girlfriend--oh, that's so sexist, the girl can't take care of herself. FEMALE CHARACTERS SHOULD NOT BE GUILTY UNTIL PROVEN INNOCENT.)

    ReplyDelete
  4. This is the first I've heard of the Bechtel test, and I actually like it a lot.

    I am perfectly happy to watch movies that have a completely male cast, such as 12 Angry Men or Saving Private Ryan. I don't mind movies about boys who have a Mom and a Dad in the background.

    What I AM tired of is tokenism. "OK! We'll put in a girl." Yay, now there's one girl for some reason in a group of guys. Oh, can't let the feminists complain that she's too passive. "OK! She will not just be quiet and feminine -- she gets a sword too! That should shut up those annoying complainers."

    The Bechtel test won't let people throw a female character or two in the mix with no thought given to developing them. They should have ideas that don't center on men. Most women spend a lot of time in conversation, and most of that conversation is not about men. In a movie are we getting a man's view of women, or are we getting a truer picture of what a woman's life is like? I actually hate the male characters in "chick flicks" that never converse with male characters about anything other than their womenfolk. It's clearly a women's view of men, and a self-centered one at that. It's like kids who think their teachers would never go to the grocery store or have conversations that aren't about school. I also think Jane Eyre would not have been as vivid as she is without the wonderful chapters about her cruel aunt and her dear friend in boarding school. It made her more true to life. The plot centers on romance, but the protagonist has a life before and outside of Mr Rochester.

    What I like about the Bechtel test is that it can be used to call out movies that feign feminism but do it shoddily. I don't really care if great movies pass it or not, because I don't need every movie to have women in it to enjoy it. I am so sick of the warrior chick shoehorned into movies, though, that it would actually be kind of nice to see movies that have to do a bit more.

    ReplyDelete
  5. Yeah, the "Instant Feminism, Just Add Sword!" trope really needs to die.

    ReplyDelete
  6. I think you may think the test is about more than it is. It's not helpful for determining how strong or compelling the female characters are, how woman- centered a movie is, how good a movie it is.

    It really deals with the women in the background of a movie, and how women are assumed portrayed to be the exception to the rule. In movies, men can be supporting characters, comic characters, villains, ensemble members. They are shown as just part of the story, part of life--and perhaps it's better to think about the collective of films in totality that make up Western cinema than any movie individually. When a woman makes it into this ouvre, she is not just a part of life, she is an Event.

    I'm not sure this is making sense. But as an above commenter pointed out, it's mainly a critique of the Smurfette syndrome. With that in mind, I think it is most useful for ensemble films, and also as a broad metric of the way Hollywood tends to write women than an artistic critique of a particular film.

    ReplyDelete
  7. Well, mostly it's biased against movies that don't have long scenes of conversation. Which, given the demands of modern pacing, is pretty much all movies; it's rare that a conversation lasts thirty seconds.

    ReplyDelete
  8. Clare,

    It really deals with the women in the background of a movie, and how women are assumed portrayed to be the exception to the rule.

    I guess the thing is: that doesn't really jibe that well with my experience of watching movies. Though admittedly, I watch very few movies these days and my tastes don't run to big budget ensemble movies, so that may simply put me out of the demographic.

    Though now I'm thinking about it, I pulled an old list of favorite movies I posted to see how it lines up:

    Gosford Park -- Pass
    Princess Bride -- Fail or near fail
    Spirited Away -- Pass
    Henry V -- Pass
    Brazil -- Barely passes
    Babette's Feast -- Pass
    Gattica -- Fail or near fail
    Wag the Dog -- Pass
    My Neighbor Totoro -- Pass
    Gladiator -- Fail
    A Christmas Carol (George C. Scott) -- Pass
    O Brother Where Art Thou -- Fail or near fail
    Bladerunner -- Fail or near fail
    Apocalypse Now -- Fail
    Persuasion -- Pass
    Ran -- Pass

    Also, I guess it seems to me like once people get talking about the test they almost invariably end up taking it farther than it really stretches -- as the Feminist Frequency video blogger does a few times. What the test shows is sufficiently trivial it's almost hard not to make it into more than it is.

    ReplyDelete
  9. Suburbanbanshee,

    Yeah, I have to wonder if she sat down with a stop watch to see how long most movie conversations really are. 60 seconds is a very, very long time in a modern movie.

    Now if the Girls on Film folks would redo Glengarry Glen Ross and My Dinner With Andre, we'd be all set.

    ReplyDelete
  10. Fair enough. I've always seen it as a specific and limited way of organizing information about Hollywood's norms that a litmus test of value signifier.

    ReplyDelete
  11. sorry--that's "than a litmus test" and "or a value signifier."

    geesh.

    ReplyDelete
  12. The biases in the test are indeed rather peculiar; for instance, science fiction movies pass with a surprisingly high frequency simply because conversations in science fiction movies are virtually never about men. One of Bechdel's original examples of a movie that passes is Alien because Ripley and Lambert just happen to exchange some brief words about the alien. And that rather underlines the problem; the fact that this happens is so incidental to the plot, and depends so little on either of them being a woman, that the fact that it passes, and what is more passes on that rather than (e.g.) the fact that its main character is a woman in a role that usually went to men, is extraordinarily trivial.

    ReplyDelete
  13. I've listened to many conversations between women, both where the subject included a man and where it didn't. Believe me, none of them are worthy of putting on film and expecting somebody to pay money to sit through it for 90 minutes. That said, as a man I can assure you that 99.999% of our conversations aren't movie worthy either.

    ReplyDelete
  14. +JMJ+

    Brandon: I believe Ripley was originally written as a man.

    ReplyDelete
  15. Enbrethiliel -- I'm very sure you're right; and as I recall not only did the script call for a man, the original actor they were hoping to get was a man -- but the studio was pushing to have more women cast, and early on they had recognized that for practical purposes all the parts were unisex, so they eventually decided to cast a woman in the lead role to make it stand out. Which I find interesting; looking back it's impossible to imagine anyone filling the role better than Weaver did, and, of course, after the sequels, Ripley as a man just became unthinkable....

    ReplyDelete
  16. A Philosopher3/14/2012 2:32 PM

    Try running the Converse Bechdel Test (at least two men with names who talk to each other about something other than a woman), and then compare the "pass" rates for the Bechdel and the Converse Bechdel. I haven't actually done the empirical legwork, but I'm pretty confident there's going to be a substantial difference between the two.

    I think that fact is really all people are trying to get at with the Bechdel test. It's meant to be a crude statistical instrument for revealing the differential treatment of male and female characters in Hollywood movies. As such, it seems not bad.

    ReplyDelete
  17. When you actually compare Bechdel test and reverse Bechdel test it becomes very noticeable that the three major factors in whether a movie passes one or the other is (1) the sex of the main character(s); (2) whether the movie is structured so that most character interaction is dominated by the main character; (3) whether the plot is focused on relationships or some external plot. They aren't the only factors, but they clearly are what usually does the work, and the combination of (1) and (2) seems to be especially significant. Thus the real question is what the Bechdel test provides that could not be provided simply by tracking whether women are getting major roles in a wide range of movies.

    ReplyDelete
  18. I don't really deny that the Bechtel test would loosely correlate with the number of medium to large female roles in movies -- though it seems like there are measures that would correlate better with that such as the number of female roles with more than ten minutes of screen time. I kind of suspect the reason that the Bechtel Test is more popular than such a measure is that it's more fun (and less work) to administer, and it bundles in an implicit desire that female characters operate in a semi-separate sphere from male characters.

    I suppose a more general question is to what extent one wants to see movie characters be representative of the population as a whole. Over-represented groups that immediately occur to me are criminals, police, soldiers, aliens and supernatural creatures. (Among the first three of these, men form a majority in the real world as well as on film, though I have yet to see diversity figures on the latter two groups.)

    I'm not necessarily sure to what extent these disproportions are a problem, though movies that break the mold can often be more interesting than those at follow it, all other things being equal.

    ReplyDelete
  19. I would agree that the Bechtel test is not a great way to evaluate individual movies, but seeing the results of the test over a wide range of movies can still be informative.

    Consider, for example, a "reverse" Bechtel test. For a movie to pass, it has to have two named male characters who talk to each other about something other than a woman.

    How many movies would fail this test? Not many. Yet lots and lots of films fail the Bechtel test, including plenty of movies that are considered high quality (i.e. it's not just "guy movies" that fail).

    ReplyDelete
  20. How many romantic comedies could pass the test? They're usually one man, one woman, interacting with each other, and only trivially interacting with anyone else.

    ReplyDelete