The Princess Problem

Like many other kids growing up in the 90’s, linguist (and recent addition to dscout’s research team) Karen Eisenhauer was a big fan of the movies of the “Disney Renaissance”—the films, beginning with 1989’s “The Little Mermaid,” that marked Walt Disney Studios’ return to animation after a thirty year absence.

“I have a special affinity for those films,” Eisenhauer says. “I know the songs by heart. Aesthetically, you can’t beat them. There’s a reason why Disney created so many blockbusters. They really know what they’re doing.”

Some of the most profitable blockbusters in Disney’s portfolio were created during this Renaissance period, which is also credited with bringing the “princess” archetype that had been a mainstay of Disney’s earlier films into the modern day. In 2001, Disney formally created the Disney Princess Line, bringing many of its most popular film characters together under a single franchise umbrella. The line is now worth over $4 billion dollars, and has become the highest-selling entertainment franchise in the world. And, as they say, with great influence comes… well, you know.

Scrutiny of the Princess Line is not new; both the mainstream media and academia have examined the line’s influence from numerous angles, including their influence on feminism, diversity, and the message they communicate to kids, particularly young girls. But, as Eisenhauer points out, most of the analysis that had been done had been qualitative or literary. Triangulating that qualitative research with quantitative data, she says, helps paint a more complete picture of the films’ influence.

Alongside fellow linguist Carmen Fought, Eisenhauer embarked on a language-based study of the twelve films in the Disney Princess canon. Examining both the frequency and duration of speech across genders, as well as discourse markers related to directive and complimentary language, the researchers produced a more rigorous understanding of just what these films are really saying. dscout sat down with Eisenhauer to chat about the project.

Notes from Karen Eisenhauer....

What Linguistics Has to Do with Princesses...

Disney is clearly one of the biggest media and entertainment forces in the world, and especially since their big animation renaissance in the 1990’s that started with The Little Mermaid, they’ve been at the forefront of pop culture in a way no other company has been. Their animated films have only continued to grow in terms of their influence over the zeitgeist, and because they’re marketed and targeted specifically toward children, and in the case of the Disney princesses, to girls, the films are a huge buzz topic amongst parents. They’re scrutinized for the messages and values they’re communicating, but also for their gender and ethnic diversity. And we thought bringing linguistic methodologies into that conversation would be really valuable, because linguistics can actually help quantify and contextualize concepts that don’t intuitively feel quantifiable, like gender and ethnicity representation. Applying these methodologies to the Disney films, which are so well-known, felt like an opportunity to bring linguistics into a conversation where it could be really beneficial, especially for people who are trying to improve the media being served to kids.

On gender parity...

The first thing that we did was just count speech. There are a lot of different ways to do this. You can record time. You can record how long people are holding the floor or how many turns they’re taking being the primary speaker. For automation purposes, we decided just to count words. We totaled up character by character, and aggregated the data by gender, based on their representation in the film. Songs were excluded from the word count because they aren’t designed to represent “real life” speech in the same way dialogue is. What we found was a massive discrepancy in the screen presence of male characters and female characters.

What was surprising was that in the early films, things were actually more even. But with the Renaissance films, the cast of the movies started to expand pretty dramatically, and the expansion was largely male. Generally, the films have one really exceptional woman at the forefront, the princess. She doesn’t speak less than her corresponding prince, they usually have pretty equivalent speech paths. But the whole world that’s filled in around this central woman character tends to be male, unless there’s a really specific reason for the character to be female, like she’s a prostitute or a mom. It’s a sort of default male world with one exceptional woman standing in the middle. If you look at the numbers, Brave (which was actually produced by Pixar and released in 2012) is the only film to have a female speech majority since 1959.

Diving into a discourse analysis

Although the lack of gender parity turned out to be really shocking, dividing the dialogue by gender was really more about understanding the sample size, so we could start to build a more detailed picture of the specific behavior. The main goal of the study was to take a close granular lens to specific speech patterns: compliments, insults, apologies and commands.

We hypothesized that the Disney princesses would have a hyper-feminine style of speech, in the same way that they’re drawn with really small waists and delicate wrists and have a hyper-feminine appearance. That they’re going to tend pretty unwaveringly towards feminine style of speech and perhaps even in a less nuanced or more constant way than women in real life do. We did a pretty broad background study of previous literature, and came away with a few big generalizations: women tend to pick more indirect and “polite” approaches to giving directives, to “soften the blow.” In linguistics, we call them mitigated approaches. Things like, “Can you just do this?” or “Would you please do this.” Men have a tendency to mitigate less and to use more straightforward commands. And when you take all three of these factors into consideration, they start to build a pretty cohesive picture, which is that women aren't nearly as well represented in these films as you would expect them to be.

On the evolution of dialogue…

In the early films, the dialogue was very stilted and formal, almost operatic. A lot of times, they spoke in verse. It was almost a more formal and much shorter script. By the ’90s the expectation was that movies embraced a more realistic speech style. There was a big spike in directives, and mitigation and politeness because that’s how people actually speak from day to day. You have to be polite in order to not seem like a jerk. So you started to see characters do that more.

On how things have improved since the 1930’s...

Early on, in the 30’s films especially, women were complimented the most on their appearance, nearly twice as much as anything else. And men were complimented most on their skill. But that’s changed really dramatically over the last 70 years. The number of compliments women receive based on their appearance has gone way down, to the point where it almost feels like they’ve banned it in the “new era” movies. Whereas compliments to women about their skill in the new era movies has gone way up. It’s not super surprising because there’s been such a push for the Princess movies to be more feminist. And compliments are an obvious speech marker that shows that, because the compliments are a really explicit way to emphasize the kinds of values you want to emphasize. So in that sense, they have really improved in a lot of ways.

But still have a long way to go...

The movies that came out before 1989 are just garbage in terms of gender representation. That’s not a surprise, of course. But what was surprising to us was that things may not have actually improved as much as it seems like they would have. If you dig into the actual speech behaviors of the characters, there actually isn’t quite as much progress. Because it’s not just what they’re saying, it’s how they’re saying it. With the Renaissance movies, there was a sort of superficial push toward girl power, but then they had really distressing numbers when it came to the actual representation or the linguistic patterns of the genders. Even in the newer films, when women give bald directives, it tends to be much more noticed by the other characters around them. It’s a plot device that’s supposed to be noticed by the viewer. Whereas with the male characters, it seems more taken for granted that they will say things and that people are going to listen to them. So on the surface, the newer films look like progress. But what we found is that underneath, we’re still internalizing that women being proactive is unusual.