Wednesday 22 May 2019

Lies, damned lies, and statistics


I don't often do refutations. Most of the time, they're a waste of effort and rapidly devolve into people throwing rocks at each other. But this recent BBC piece on speaking roles in Game of Thrones.. well, you might want to read it for yourself before continuing. One thing you can say for it is that since the BBC isn't ad-supported, it's not click-bait.

Anyway, here's the gist- the piece suggests that there's something wrong (I won't use the 'P-word' because they don't) with the fact that roughly 2/3 of the dialogue in the series is spoken by men. Here's one of the quotes:
"Overall, the actions and words women are participating in are still very attached to gender-related stereotypes".

Now this is a repetition of something I've mentioned before and call Raven's Cry Syndrome. If you're not familiar with what I'm referring to, Raven's Cry was a game that came out a few years ago set in the 17th century that was mostly about pirates. It wasn't very good, but simply pointing this out wasn't enough for certain critics, who instead devoted large chunks of their reviews to complaining that the characters were racist, sexist, and swore a lot. Well guess what kids, pirates were Not Very Nice People, and there's a reason no-one holds up Blackbeard as a paragon of virtue. In effect, Raven's Cry Syndrome is when a period-set piece of fiction or art is criticised simply for accurately reflecting its setting.

MINOR TO MAJOR GAME OF THRONES SPOILERS MAY FOLLOW
Coming back to Game of Thrones, we're looking at a low(ish) fantasy story set in a basically medieval world. Whilst there are plenty of noblewomen, queens and other women in positions of power, as a world where military tactics still revolve mostly around using muscle power to hit people with sharp bits of metal in various shapes, the majority of the military are male. To define our terms here a little, I say low fantasy because whilst GoT has plenty of magic in it, it's usually of the ritual or mystical variety. Even the most powerful magic-users, like Melisandre, can't cast Fireball and obliterate an entire regiment. By contrast, in a high fantasy setting, magical abilities make female military commanders or warriors far more feasible, as seen in my own books, for example. (Come on, had to get that in there). It's hard to tell Magister Thalia Daran to sit quietly on the sidelines when her personal offensive  firepower is up there with the average battle tank.

This means that when you get a big battle scene, like the Battle of the Bastards, the defence of Winterfell, etc, you're going to have a lot of blokes in beards standing about discussing tactics. It's just inevitable in such a setting. GoT, of course, makes a point of having characters like Arya and Brienne who break the mould, but they do it in a setting-appropriate way- Arya by gaining mystical powers and utilising stealth and deception, and Brienne by being simply an unusually tall and strong woman. Even so, both run up against patriarchal attitudes, as you'd expect, but they overcome them.

Even allowing for the setting, though, the analysis in the BBC piece is blisteringly unfair. The quote I use as my title has been variously appropriated to Mark Twain, Benjamin Disraeli and others:
"There are lies, there are damned lies, and then there are statistics."

The point of this isn't to suggest that all statistics are worthless, but to remind us that you can massage them all sorts of ways to make a point if you'd like. To use the language of Sherlock Holmes, statistics aren't facts, they're data. A fact is data with the context to interpret it. For example, the statistics tell us that the character who has the most lines in GoT is Tyrion. This should hardly be a surprise, given that talking is Tyrion's primary talent (especially TV Tyrion) and he's in more episodes than anyone else. If we look at the chart, we see that in the top four speaking characters we have two men, and two women (and second place looks to be a male-female tie).

Here's where trying to use statistics to analyse fiction falls to bits completely. The two top-scoring women are both, for most of their time in the show, queens or rulers. The thing with being in charge is that often you do more listening than talking. Consider the example of a board meeting- the Chairman is the most important person, but spends most of their time asking their subordinates for reports and getting comments on them. As the saying goes, there's no point having a dog and barking yourself. Possibly the best example of this, though, comes from the very last episode, where the remaining lords of Westeros are appointing a new King. Tyrion does a lot of talking, but it's abundantly clear that as far as those listening are concerned it's Sansa wearing the proverbial trousers. "Shut up, Uncle." may be one of the simplest and most effective burns in GoT to not be delivered by an actual Dragon.

Ultimately, trying to determine who has the most impact or agency in a story based on how much time they spend 'on camera' or how many words they speak is pointless. It's like trying to analyse cooking with economics- cold, hard numbers have nothing to do with the feel of a story. If everyone watching thinks a story has strong, powerful women who take control of their own destiny in it, they are by definition correct. Using statistics to try to argue against that is like watching an unconscious boxer being carried out of the ring on a stretcher and saying that because he landed 100 punches and his opponent only landed three, he actually won. It's not quantity that counts in communication, but quality. To go back to the chart, we see Lord Varys sitting very near the bottom. I defy anyone to claim his was a weak or ineffectual character, he just doesn't make a big noise about his influence. He's called 'Master of Whispers' for a reason.

I'm not sure what the motive of the BBC piece or the research behind it was. I'm all for strong female characters in fiction (I hope my own writing reflects that) but Game of Thrones is completely the wrong target for this sort of piece and nothing hurts a good cause worse than a terribly flawed argument.