Parity Test
Considerations for Analyzing Representation in Media
Litmus is a substance made from lichens, I guess, which is most commonly infused into little strips of paper that you dip into a solution to check the general alkalinity. Litmus paper changes color depending on whether the solution is acidic or basic. I don’t know if anyone uses it anymore give you can buy a digital pH meter from Amazon for like $12 that uses an electrode and absolutely no litmus. Maybe high school chemistry classes are still using paper.
I had to look up the information on how litmus paper works because I slept through high school chemistry. I mean, literally, I Rip-Van-Winkled it. I only passed—and, on the shoulders of that sturdy D-minus, graduated—because Ms Bednarz decreed that anyone who passed the final would pass the class and then Gizachew slipped me the answer key the day before. I owe that man a lot. At the very least, the summer after my senior year of high school.
Anyway the thing with litmus is it doesn’t give you the precise alkalinity of a solution. It’s a piece of paper, Jan, what do you want from it? It only tells you, by changing color, whether the solution is basic or acidic; but not exactly how much. Again, my $12 digital meter is much better. But sometimes in life you can’t get an exact answer to a question and you have to settle for a general answer, as my non-sleeping peers had to do in high school chemistry back in the ancient time before Amazon dot com.
That’s how the term litmus test gets applied to politics, where you might look at someone’s record of voting on legislation or their decisions from the bench to project how they might vote or rule in the future. You won’t know for sure how they’ll vote until the time comes but the litmus test gives you a general idea of what’s likely. This is a way better solution than simply listening to what politicians say about what they’re going to do in the future since all of them are lying at all times.
Today’s Shelf Life is about how to apply various existing media litmus tests to the things you read—and, ideally, the things you write—to gauge whether representation in a work (or body of works), if representation is present at all, is tokenistic or meaningful. I like calling this specific type of media test a “parity test” because it’s testing for parity of representation in the media, but also because “parity checking” in binary code refers to using a parity bit as a check digit to validate the integrity of the rest of the byte.
That’s what we’re doing: We’re testing parity and we’re checking integrity (albeit in a different sense of the word integrity).
Good morning: It’s a beautiful day in the neighborhood, single men in possession of good fortunes are in want of wives, and the clocks are striking thirteen.
I hear a lot of people say that we don’t need stuff like this—media parity tests—anymore, because we’re aware of discrimination and bias and we’re thinking critically about those things and so we can tell, when we look at a piece of media, if there’s good representation or not. We don’t need to apply a test. If something is skewed to the point of being completely male-driven or white-driven, we’ll notice.
Everybody, I am here today to tell you that well actually that’s not true.
Because we—meaning consumers of media—are so used to seeing certain groups underrepresented on page, stage, and screen, we’ve come to see that underrepresentation as the norm and interpret it as parity when we see it. Sometimes we even interpret it too far in the other direction, as Geena Davis (of Beetlejuice and the Geena Davis Institute for Gender in Media fame) explained to NPR:
“We just heard a fascinating and disturbing study, where they looked at the ratio of men and women in groups. And they found that if there's 17 percent women, the men in the group think it’s 50-50. And if there's 33 percent women, the men perceive that as there being more women in the room than men.”
What she is getting at is this: We’re so used to seeing women in the minority in media that we perceive the usual minority of women as parity and perceive a larger minority of women than usual as a majority of women. We think we’re at or near parity in a lot of places, but it’s only because adjusted our understanding of parity to align with the underrepresentation of certain groups that has become the norm. Dr Virginia Valian, who has written extensively on gender parity and people’s perceptions of it, has described this well-documented phenomenon as a “cognitive error.”
Listen, there’s references today just in case anybody feels contentious. In 1990, Cutler and Scott1 found that in dialogue between a woman and a man, wherein each party speaks an equal number of words, the woman is perceived to speak more than the man. The Geena Davis Institute for Gender In Media found that, in a sample of 122 films with a total of 5,554 speaking roles, only 29.2 percent of those speaking roles were held by women (or female characters).2 Another study found that women comprise about 17 percent of the people in any given crowd scene3 when, logically speaking, crowd scenes should approach an equal split among men and women (with space set aside for, you know, other people who are neither men nor women).
In short, that’s why we still need to use these litmus tests to determine the amount and quality of representation in media. First of all, having an equal number of women and men characters does not mean that a piece of media has achieved gender parity. But even if that was what it meant, as consumers of media we are not good at doing that and should not rely on our subjective perception of what we’re seeing to make a determination.
So: Along comes Alison Bechdel and her comic strip, Dykes to Watch Out For. In a 1985 strip titled “The Rule,” two characters discuss whether they should go see a movie. One character states that she will only see a movie when it has:
At least two women in it,
Who talk to each other,
About something other than a man.
When her companion tells her that’s a pretty high bar, she agrees and says she hasn’t been able to see a movie since Alien (1979). I won’t reproduce the original because I don’t think it’s in the public domain, but you can see the full strip here. These three criteria are known as the Bechdel Test (or the Bechdel-Wallace Test). Hopefully those lesbians got to see Aliens (1986) when it came out the year following the strip’s publication. Aliens passes the Bechdel Test.
The test has some problems. There are many examples of films that have (what I consider) good or excellent representation of women, but which do not pass—like Gravity (2013) and Arrival (2016). There are also plenty of examples of movies with poor representation of women that do pass, like Goodfellas (1990) or the 1992 Sir Mix-a-Lot anthem “Baby Got Back,” which famously passes with its opening vignette of two women discussing the obvious aesthetic merits of a third woman’s butt.
The Bechdel Test was one of the first media parity litmus tests we got to gauge whether a movie has roles for women as more than set dressing—but it’s not perfect. When Pacific Rim came out in 2013, many people pointed out that it failed the Bechdel Test in spite of the well-developed character of Jaeger Pilot Mako Mori (played by Rinko Kikuchi). Pacific Rim has the distinction of being the film I have most often been told I would love if I bothered to watch it, which I assume is because it’s a rare movie with a good woman character in it and people (understandably) have simplified my taste in movies to “woman fight monsters good.”
Anyway, that whole situation gave rise to the Mako Mori Test, which is passed when a piece of media has:
At least one woman in it,
Who gets her own narrative arc,
That is not about supporting a man’s story.
By this metric, films like Arrival and Gravity pass while Goodfellas does not. Steps 2 and 3 of the Mako Mori test are excellent guidance for assessing any character in a piece of media who is not a man (women, nonbinary people, and so on) to determine whether the character is well-realized or is only in the story to serve as an accessory to a man.
Before I depart the subject of gender parity, I will also briefly mention comic book writer Kelly Sue DeConnick’s Sexy Lamp Test, which she characterizes thus:
“If you can remove a female character from your plot and replace her with a sexy lamp and your story still works, you’re a hack.”
I think she meant this in a tongue-in-cheek way, but there are plenty of women characters who don’t pass: Lois Lane from Man of Steel (2012), Daisy from The Great Gatsby by F Scott Fitzgerald, or anybody who has the misfortune to be taken from Liam Neeson.
I trust that has already given everybody a lot to think about on the topic of gauging representation, generally, even though the tests mentioned all explicitly address women, which many of us inherently take to mean cisgender, heterosexual white women, because, after all, wouldn’t transgender representation, or Queer representation, or BIPOC representation be their own tests?
That brings me, in a roundabout kind of way, to the topic of tokenism. As I discussed in My Big Inclusion Agenda, many writers feel they’re doing their due diligence vis a vis parity of representation by throwing in some characters who are BIPOC, who are Queer, who are women, who are living with disabilities, who fall on the neurodiversity spectrum, and so on.
I don’t want to name names but I’m thinking of a title I read a few years ago that had a character I presume was nonbinary, as they used they/them pronouns. When the character was introduced, the character making the introduction said something to the effect of, “this is Sam, they go by they and them.” There was no mention of Sam’s gender identity there nor in the rest of the book, the character did not have a meaty role, and none of the other characters’ pronouns were specifically mentioned or addressed. At the time, and in reflection since reading, this felt like box-ticking to me: A writer going down a list headed “Inclusion” and checking nonbinary off the list by inserting one line referencing a minor character.
This is a variation on the non-representation representation I think of as the “just happens to be” character. This is a character who just happens to be a woman/a person of color/a queer person/a trans person, but that identity does not affect the character or the story in any way. Isn’t that good, even ideal representation? People just being who they are, whether the story “calls” for an examination of their identity or not?
Alaya Dawn Johnson wrote in an article for The Angry Black Woman about applying a Bechdel-like test to determine degree of racial representation by evaluating a piece of media as follows:
It has to have two POC in it,
Who talk to each other,
About something other than a white person.
That makes sense, but how often do we see a piece of mainstream media with more than one BIPOC character? And in light of the cognitive perception bias I discussed above, consider: How many Black characters can a movie have before it’s “a Black movie” and white people lose interest in seeing it? Andre Seewood covered this topic in depth in an article for IndieWire that I highly recommend.
Whatever group of people you place in criterion one of a Bechdel-style test, there have to be two of them before they can speak to each other (let alone about something other than a man/white person/straight person/cisgender person). For me, a character’s identity need not be the focus of the story for the representation to be “good,” but neither is my desire for inclusion satisfied with representation at the level of “this type of person exists in the story.” What does it mean for a character who is not a cisgender, heterosexual, white man to be represented well?
I’ve come around to this thinking on it: If a character belongs to a community that is marginalized in our real world and that in no way affects the course of the story nor the way the character acts and reacts to the things happening around them, there could be two reasons:
You didn’t fully realize this character in your mind before you wrote, or
The character inhabits a reality in which this aspect of their identity is not marginalized.
The latter is what I saw in Pet. Without getting into great detail, the inhabitants of Lucille live in a society where people are not marginalized based on characteristics like race, sexual orientation, gender identity, relationship preferences, and disabilities. Therefore, it is not noteworthy if someone is Black, or queer, or has a disability—because this is something everyone sees, everywhere, all the time.
What you can’t do is pull a “this type of person is not marginalized in my fantasy world!” but then there’s only one person of that type in the story or that anyone’s ever met. “In my epic fantasy world, queer people are not marginalized! You can tell because there’s one queer character out of the 500 characters who are seen or mentioned but no one discriminates against her!” If queer people aren’t marginalized, and therefore they have no need to conceal their identities to get by, then why would you only encounter one person like that in hundreds of thousands of words?
I mean, look, if it’s speculative fiction you can do whatever you want, and nobody can stop you. Just maybe think about why any given type of person is so vanishingly rare in a fictional society that there’s only one of them in your whole story.
I’ll wrap up with some thoughts about how I evaluate representation in my reading and my own writing when I am developing characters.
If two queer characters (or two characters from any specific race or ethnicity) are present in a story, is it obviously for the sole purpose of pairing up romantically?
Is the lack of any romantic plot or subplot for a given character fobbed off as meaningful asexual/aromantic representation?
Is the representation all coming from the villain/antagonist side, or does representation manifest alongside stereotypes of any kind? For instance, a villainous woman portrayed as a ballbuster or a queer person who weaponizes sexual advances on characters who do not share their orientation.
Are characters with disabilities portrayed as overcoming their limits or as being good or successful in spite of their disability? Do a character’s villainous origins stem from bitterness about having a disability?
If a character is the only person representative of a marginalized group in the entire story, why is that? What is it about the setting that allows for one, and only one, of a specific type of person? Reading a story about going to Yale in the 1940s and there’s one Black person in your pre-med track? Believable. Big wide fantasy world and there’s one non-white-skinned person in it? Not so much.
Okay that’s it, that’s all I got. And it was already too much. You’re all very patient, that’s why I like you so much. Here are the aforementioned references. I hope you like bibliographies more than I do.
If you have questions that you'd like to see answered in Shelf Life, ideas for topics that you'd like to explore, or feedback on the newsletter, please feel free to contact me. I would love to hear from you.
For more information about who I am, what I do, and, most important, what my dog looks like, please visit my website.
After you have read a few posts, if you find that you're enjoying Shelf Life, please recommend it to your word-oriented friends.
Cutler A and Scott DR. 1990. Speaker sex and perceived apportionment of talk. Applied Psycholinguistics 11: 253-272.
Smith SL and Choueiti M. 2010. Gender disparity on screen and behind the camera in family films: The executive report. The Geena Davis Institute on Gender in Media.
Smith SL and Cook CA. 2008. Gender stereotypes: An analysis of popular films and TV. The Geena Davis Institute on Gender in Media.