On the Subject of Narratives · 9:36pm Jan 5th, 2021
Narratives are, generally, a good thing. They are a heuristic that we use to take a disparate collection of information, and assemble it into a meaningful distillation. Or, if you prefer a less pretentious phrasing, they're a way for people to turn a bunch of data points into something meaningful.
Humans are really, really good at creating narratives. We do it all the time! I mean, you and I do it all the time, anyway: writing fanfiction is creating a narrative, and reading fanfiction is the practice of sussing the narrative out of someone else's writing. We're very narrative-obsessed here on FiMFiction, which is why this felt like an appropriate place to share a few insights about narratives.
But it's not just us fanfic writers and readers. Humanity as a species appears to be hardwired to create narratives, even where no meaningful ones exist. Case in point? one of my all-time favorite comics:
source, if you're weird enough to find kind of this thing funny
In many games, you can be "winning" or "losing," in the sense that you are more or less likely to win the game than you were at the start. But in a game like Candyland, where all the moves are pre-set? That's just a false narrative we project onto the process of counting.
I was thinking of this because of the Georgia senate elections today (go vote, if you're a Georgian, eligible, and haven't already!). Don't worry, there's absolutely nothing political in this blog, unless you consider "go vote, if you're Georgian, eligible, and haven't already!" to be political. But I do want to talk about narratives. Because at seven o'clock Eastern, the election is over.* And from that point on, any narrative about candidates "making up ground" or "falling behind" will be a false one.
*technically, there will probably be at least some people still waiting in line at seven o'clock, and they'll be permitted to vote as long as they got in line before seven. But seven is the enforced cutoff, which is what I'm getting at here.
We saw some of these false narratives in the last election. For example, in North Carolina, absentee ballots are counted as they come in, and since absentee ballots skewed heavily democratic this year, that meant that early returns in North Carolina showed Joe Biden with a significant lead. In-person ballots weren't counted until the polls closed, though, and they heavily favored Trump. This lead some people to say things like "Trump clawed his way back in North Carolina," but that's not really the case. Trump was always ahead in North Carolina; any appearance otherwise was an artifact of the order the ballots were counted it. He had as many votes on November 3rd as he did on November 14th, when the result was called, and any suggestion that he "caught up" with Biden is a misunderstanding of how voting works. Trump won NC and he won it on election day, full stop.
Or, for an example of things going the other way, take Pennsylvania, where absentee ballots cannot be counted before election day. This meant that the first votes to come in from Pennsylvania were the (disproportionately republican) in-person votes, and the last votes to come in were the (disproportionately democratic) absentee ones. As in North Carolina, but reversed, this lead to a narrative that Biden was "coming back from behind," but that's another false narrative. Biden had more votes than Trump in PA on November 3rd, just like Trump had more votes than Biden in NC on November 3rd; past that point, any attempt to project a horse race onto those was just, well, projecting.
It's true that it took us voters a little while to find out who won on the 3rd, just like it takes a little while to find out who won at Candyland. But don't be fooled by the way information comes out piecemeal; once the voting stops at seven/once the deck is shuffled and put in place, there's no more "coming from behind" or "holding onto a lead." There's just "who won?" And even if we don't find out the answer right away, it doesn't mean the narrative we want to project has any meaning.
Narratives in fiction can be a lot of fun. I enjoy writing and reading, and if you're following me, it surely means you like doing at least one of those things. Narratives in games can be a lot of fun; I'm not saying Candyland is particularly enjoyable, but can you imagine how much less enjoyable it'd be if you had to end every turn by intoning "Of course, the outcome was predetermined before this move, and this move reflects no fundamental change in the nature of the game, which is, to repeat, entirely predetermined," all with utmost solemnity?
Narratives can be useful learning tools in many nonfiction settings, too. History books often focus on specific narratives to help readers understand the causes and results of complex events. Heck, it might be using the word "narrative" loosely, but this stupid anime cartoon is the first thing that got me to remember the difference between amps, volts, and ohms:
Look, I'm not proud of the fact that this is how I learned what they mean, but I can't deny it worked
But a false narrative based on incomplete information leads to misunderstandings. So when you hear about the election results in Georgia over the next few days, just remember that the vote is done, and all that's left is the count. And just like in Candyland, there's no "ahead" or "behind;" there's just who got the most votes, and even if that number comes in piecemeal, do not be fooled by anyone telling you that a partial result means, well, anything.
At seven o'clock eastern tonight, two candidates from Georgia will collect enough votes to become US senators. We probably won't know which two for a few days, if the Nov. 3rd election is any indication, but they'll still have won at seven o'clock tonight. And if you hear any newscasters, politicians, or various internet denizens suggest otherwise, well...
That's the power of a narrative.
A friend of mine adores Existential Comics. I don't know enough philosophers to appreciate the jokes half the time, sadly. In any case, excellent reminder of the power of the brain to tell itself stories, for good and for ill.
Aw, I was expecting some Grand Unified Narrative Theory of Everything, but this is just a coded way of flipping off political pundits. That's almost not as good!
I also have to declare myself too philosophically handicapped for a discussion of absurdism and existentialism, to say nothing of the confusion minefield that is free will.
I will overreach and say the fundamental unit of a narrative is the concept of causality, and our use (and abuse) thereof.
More generally, one reason people are such story-hungry monsters is the general trigger-finger on the button that says, "Causality Detected!" It is very hard to step back and remind oneself of all the hidden variables, confounding factors, preconditions, prior history, tangled causal networks, and so on. It is very, very easy to point and say, "That's why it happened! That! That is the cause!" People in psychology tests will see causality in meaningless moving shapes on a screen, for goodness' sake.
To be fair, some of it's down to intuitive communication: if something was supposed to be relevant to a story, for instance, we generally assume the storyteller will bring it to our attention at some point. "The queen killed the king" is very different from "The queen killed the king after ten years of abuse", even though logically the former doesn't preclude the latter.
Perhaps it's enough to know the king is now dead, and we can move on with another story (e.g. about the subsequent fight for the crown). Perhaps we do need to know why the king is dead, and we can instead move on to the queen's story (e.g. about being brought up to respect fellow royals come what may, and about the years of facing a king who sees her as a pure political tool). Changing the spotlight on the causal web. And changing focus changes the story.
And thanks to that combination of "itchy causality trigger-finger" and intuitive assumption that what you see is what you need to see, that's why people see election results slowly come in and, instead of holding back and thinking "Well, all the votes were already cast, so this is just the arbitrary way the information is coming to us, not an actual race", leap onto it and deliver a purely blinkered causal interpretation, saying: "X is progressing! X is gonna win! And... wait! Y has just overtaken X and is shooting towards the finish line!"
Good old Camus! I just reread The Plague... for some strange reason.
I think that the tendency to impose a narrative on a series of events, whether or not there is any causal links, is what drives a lot of conspiracy theories. Narrative adds meaning and comfort in a world where random chance is just as disturbing as predestination.
5429266
Check out the Existential Comics Twitter!
Its author is a bonafide communist like me!
Hey Chris,
Thank you for this blog post. I always like discussion of narratives and their place in human experience, but what I like even more here is that this gives me two great examples I can use when I'm teaching statistics to undergraduates.
To wit, I (and most people) have a go-to example for explaining the difference between two interpretations of probability: the (frequentist) long-run frequency interpretation and the (Bayesian) subjective/personal judgment interpretation. That example is so trivial that I've long wondered whether students really understand the depth of the issue it tries to capture. In the example, one flips a coin, catches it, looks at it, and asks the class for the probability of heads. Under the frequentist interpretation, the probability is either 0 or 1, depending on which way the coin has landed—and I know that probability because I've looked at it, but the students don't know it yet. Under the subjective interpretation, the probability is still validly 0.5 for (most) of the students, because that's an accurate statement of their belief about the uncertainty that still exists for them. I think students understand this, but I think it's too artificial to really help them build an intuitive understanding.
Candy Land is much, much better for the same purpose. If you know the number of players, then the winner is determined as soon as the deck is done being shuffled. From a frequentist perspective, again, there's a deterministic answer to the question of who wins and no objective uncertainty exists. From a subjectivist perspective, though, it's still entirely valid to attach probabilities to each player winning—and those subjective probabilities do meaningfully update as the game is played and some players get ahead of others on the board based on the cards that have been seen.
And, of course, as you're describing, elections are basically the same thing as Candy Land. Watching Georgia senate results come in tonight, and the NY Times system for expressing "uncertainty" as the count comes in, is an excellent example of the same point in a very real-world context.
If you want credit when I use this example, in class or if I record a video on in for my (very minimal distribution) stats YouTube channel, shoot me a message letting me know how you'd like to be identified. And again, thank you very much for this blog post and for pointing me at some great new examples I can use in discussing the main interpretations of probability and helping students to understand them.
Whether something is unpredictable because of intrinsic uncertainty, or just because the information is not available to the observer, and what difference that makes, is something philosophers and physicists can have long discussions about.
While journalists just recognise that elections make gripping narratives (provided there is enough unpredictability). How could they not be? We have characters, conflict, the potential to make history, and the promise that one way or another it will be decided very soon. This has us repeatedly checking our phones to get the latests results. And the media outlets will do all they can to get each update on their channel first. I find it far more exciting than sports events, although I'm sure plenty of sports fans would disagree. Although the cause of uncertainty is different, the narrative is very similar.
In science writing we are always looking for ways to turn an abstract bit of research into a compelling narrative, so I find it very interesting how something as simple as counting voting papers can make such a gripping story.
derpicdn.net/img/view/2020/10/8/2461820.png
Here is your anime girl Ohm’s Law poster but ponified.
5429266
Meanwhile, only like 40% of the philosopher humor goes over my head! I find his stuff is like Kate Beaton's, in that the stuff I understand is funny, and the rest inspires me to go look up stuff I'm missing often enough to count reading the comic as a learning experience.
5429276
It really doesn't matter how many times you tell yourself "correlation is not causation," it still feels like they're the same thing. We just have to keep reminding ourselves that feelings, while they can be just as important as facts, are not the same as facts.
...Aaaand that's enough italics abuse for today. BTW, I love your example of how changing focus can change the narrative!
5429425
As the old saw goes, conspiracy theorists are the world's greatest optimists: they never accept that bad things can happen unless a powerful cabal goes out of its way to make them happen.
I lied, I will never stop abusing italics.
5429642
That twitter is the strangest combination of political theory, philosophy joshing, and nerd-bating that I've seen in a while!
5429696
Hey, glad that's useful to you! But it sounds like the comic author's the one who you should credit, if anyone; for my part, feel free to use anything I've inspired, free of charge. We educators gotta stick together, and all that
5429790
One need go no further than the bingo hall to see how something as simple as reading off numbers can be incredibly exciting. Of course, there is chance built into that, unlike the Candyland and vote counting examples, but we can still see the same kinds of narrative-building around a process that's nothing more and nothing less than reading off a set of digits.
5429947
That is adorable!
5430409
I, for one, do not feel as if there is anything wrong with the proper exercise of an underused typographical emphasis!