Overcoming Fiction

We love fiction, in novels, movies, plays, TV, and so on.  But fiction is made up; it is not true.  Worse, storytellers have lots of standard tricks to get us to draw the conclusions they want about their fictional worlds.  Tricks use character beauty and wit, camera angles, particular consequences emphasized, and much more. 

We do try somewhat to not let fiction overly influence our beliefs about the real world.  But surely we fail in many ways; the task is just too hard.  How can we do better?

One approach is to prefer apparently true stories, such as biographies, news articles, reality TV, or grampa's war stories.   Of course lots of fiction slips in, but at least these can be fact-checked. 

Another approach is alternate versions of standard stories.  Check out The True Story of the Three Little Pigs, Wicked, and The Case for the Empire.   I'd love to see bias-reversed movies portray the same events as a famous movie, but change the standard storyteller tricks to make us draw a very different conclusion.  Imagine It's A Wonderful Life from the view of Mr. Potter, or if E.T. the Extra-Terrestrial was really a scout for an alien invasion.  The lack of interest in bias-reversed stories suggests we aren't that interested in overcoming fiction biases. 

Many people seem to think our ancestors' fiction was less realistic than ours, suggesting we are making progress toward reducing fiction biases.  I am suspicious of this claim; the fact that no one seems to have bothered to try to measure it suggests to me people don't really believe it.

Comments

[Stuart Armstrong]

There's another problem - those factural stories we here the most often are those that make a good story (Napoleon and Howard Hughes rather Ivana, the Russian peasant's wife). Even then, we only get to hear the bits that make a good story (Napoleon's culinary tastes or language skills are rarely touched upon).

[Stuart Armstrong]

Sorry, previous comment should read "rather than Ivana, the Russian peasant's wife". And no, I don't have her story either :-)

One approach is to prefer apparently true stories, such as biographies, news articles, reality TV, or grampa's war stories. Of course lots of fiction slips in, but at least these can be fact-checked.

But even before we fact-check, we need to get some vague idea as to whether the story is credible or not. We have two tools here: our experience of reality, and our experience of pure fiction.

I think that seeing entriely unreal shows like Star Wars, ET, Sherk, Pirates of the Carib, etc... allow us to get a feel for the tricks that fiction uses. And when we see those tricks in more "factual" works, we can become more suspicions. This is the sort of experience that allows us to say that "Enigma" and "U-571" were suspicious before even looking them up, and that "based on a true story" means nothing more than "no eleves or ewoks".

But maybe a better guide for the truth of a story: stories are meant to be interesting and entertaining. If a supposed factual report is also interesting and entertaining, be suspicious. But if it's dull as dirt, then it's more likely to be true, and representative.

[Zubon]

Another approach is alternate versions of standard stories. [...] The lack of interest in bias-reversed stories suggests we aren't that interested in overcoming fiction biases.

To what extent is this the case? We have three examples cited that are rather popular, although the originals are all better known (which makes sense, since the derivative work loses something without the original). We could think of a great many more, and then there are stories that are non-standard takes on traditional tales, non-explicit re-tellings, etc.

All the popular Disney films, for example, are alternate versions of standard stories, often radically different. The author of Mary Poppins supposedly left the theater crying, the original of Pinocchio was so dark I thought I was reading a parody (notably when the Talking Cricket gets smashed against a wall), and the Little Mermaid gets to choose between dying or murdering her prince in the Hans Christian Andersen version. The best known versions of many stories are the alternate versions.

To some extent, that makes it the same problem, if we are interested in only the bias-reversed version. It also makes the original claim of bias practically unfalsifiable, since there will be few cases where multiple versions of the same story are equally popular. One is better known at any given time.

[sa]

but fiction is entertainment, why would anyone try to see a reverse biased account. scary movie sereis spoof is a step in the direction you suggest.

[Robin Hanson]

sa, yes, scary movie spoofs do indeed show and therefore counter the tricks scary movies use.

Zubon, not every variation on a story counts as bias-reversed - the point is to reverse the main sympathy the audience is supposed to feel. Thousands of stories may be variations on Romeo and Juliet, but how many of those take the side of the kids' parents?

Stuart, sure without some exposure to fiction it would be harder to see the fiction in supposedly true stories. But few people haven't seen enough fiction for this purpose.

[Gwern]

"Imagine It's A Wonderful Life from the view of Mr. Potter, or if E.T. the Extra-Terrestrial was really a scout for an alien invasion. The lack of interest in bias-reversed stories suggests we aren't that interested in overcoming fiction biases."

I would argue that this is contingent: those aren't made and sold commercially on any scale because the copyright owners would *sue the everloving snot* out of you if you tried. For comparison, George Lucas leaves fans alone mostly, and we can see all sorts of reversed stories in fan-made videos or material. More subjectively, I don't remember seeing any glaring lack among the Japanese doujinshi community (who generally have even less to fear legally).

[Hal Finney]

An online example of this kind of thing is the Darth Vader blog from a couple of years ago, The Darth Side, Memoirs of a Monster.

"It is a period of civil war. Guerrilla spaceships striking from a hidden base have won a narrow victory against our loyal Imperial forces, inflicting heavy losses in terms of casualties and, worse, the galactic peace. During the battle insurgent spies managed to steal secret plans to our ultimate weapon and the brainchild of Tarkin's dream, the DEATH STAR, an armoured space station with enough firepower to destroy an entire planet.

"The whole thing has given me a massive headache."

It's entertaining, funny, and surprisingly moving to see the events of the first Star Wars trilogy from Darth Vader's perspective. I'm not sure the moral is really all that different in the end, but we do get to see Vader protecting galactic order from a band of terrorists and anarchists. A very well written story.

[Carl Shulman]

"Many people seem to think our ancestors' fiction was less realistic than ours, suggesting we are making progress toward reducing fiction biases."
This may be an issue of semantics. It might be that the average difference between fiction and human models of the actual world remains fairly steady, but fiction becomes more realistic over time as our models of the world become more accurate.

[Zubon]

Zubon, not every variation on a story counts as bias-reversed - the point is to reverse the main sympathy the audience is supposed to feel. Thousands of stories may be variations on Romeo and Juliet, but how many of those take the side of the kids' parents?

Granted. I want to ask if you have an argument about how rarely/often this happens, but that is not terribly easy to quantify. I infer that you are suggesting that stories with reversed sympathies are exceedingly rare, while I would suggest there are a great many, especially if you accept imperfect parallels (not exactly the same story but an intentional reversal).

Also, sympathies in stories are not necessarily binary. Wicked may tell the Witch's side of the story, but Elphaba is still a paranoid former terrorist in a state of manic rage, while Dorothy is even more virtuous than she is in the original. What would it mean to reverse sympathies in Pinocchio, to root for the whale? In the original, he is a layabout and cad, while the Disney version has him as an innocent who is occasionally led astray. Disney's Little Mermaid suggests that being headstrong and lovestruck will ultimately lead to happiness, as opposed to suffering and death in the original. That is the common reversal of sympathies in modern versions of fairy tales: the protagonist is shown as an admirable victim of circumstance, rather than being an object lesson in the consequences of foolishness. Fractured fairy tales that shift or reverse the perspective are quite common, although as you say, many are re-tellings rather than reversals. Is Anouilh's Antigone a reversal of sympathies, a shift, or what?

The reception of such reversed and re-told stories may strengthen the original point. They are usually considered oddities and amusing side-notes rather than competitors to the originals (although Wicked is doing quite well on the stage). I am reminded of one of the chapter openings from Speaker for the Dead that twice retells "let he who is without sin cast the first stone" and puts both retellings in positive and negative lights; it then uses them to reflect upon the original.

[Nic "RedWord" Smith]

I'm a bit curious to know how pen and paper role playing games fit into this idea. On one hand, they're very much fiction - the rules that apply to a game are different from reality, sometimes radically so, and not usually a mere abstraction. On the other hand, some level of moral ambiguity is not uncommon, in order to allow players to make decisions and run into the (probabilistic) consequences. To some extent, it seems that game systems are fiction with a quantifiable bias - Paranoia may be biased toward killing the character in his everyday life, while Exalted may be biased toward keeping the character alive in situations well beyond what she would normally encounter if a real person. On the other hand, the moral bias of a game system is not quantified by its creator, even though it's clear that different game systems portray morality differently.

[Robin Hanson]

Gwern, lawsuits explains lack of satires on recent fiction better than on older fiction.

Hal, Darth Side does far less than it could to make us sympathize with Darth.

Carl, I'm not sure today's fiction presents a more realistic social world than older fiction.

Zubon, yes, I'm pointing to how rare it is to have revisions that reverse sympathies.

Nic, role playing games do reduce some biases by losing plot control.

[JWR]

Every day in my job as a lawyer I am startled by the fiction bias: the participants in the legal system rely on it very heavily, and (most distressingly) don't even seem to recognize it as a bias as opposed to an appropriate "truth-seeking" device.

Lawyers obviously work very hard to "tell a story" to a jury that seems believable --- as opposed to simply introducing the evidence and then explaining what it shows. What's most troublesome, though, is that they don't feel any need to hide this, but instead boast about how they will tell the jury the "best story." Lawyers often start opening argument with, "I'm going to tell you a story," or "This case is a story about . . .," or words to that effect. There's no self-consciousness about this at all. Juries WANT to hear "a story," as opposed to "just the facts."

[Robin Hanson]

JWR, that is an interesting observation; it would be great to see that issue explored in more detail.

[Bob Unwin]

"The lack of interest in bias-reversed stories suggests we aren't that interested in overcoming fiction biases."

I think it suggests very little. The people I know who want to avoid influence by fiction do not read bias-reversed fiction. Instead, they talk about what they've read critically ("That super-power violates the conservation of energy", etc.) or they just forgo reading the sort of literature that might bias them. Reading lots of science fiction might bias one's view of future technologies. Ditto reading political novels or plays. But there are literary works that try to make philosophical arguments of some sort (Montaigne's essays for example, or Plato's Symposium, or Beckett's Waiting For Godot). I don't see the events in Waiting for Godot or The Symposium being biasing in any way. There is literature that is mainly formal (e.g. Stein's poetry, some of Lewis Carroll's poetry, some of Ezra Pound's stuff, some of Auden). And there is literature that tells mundane tales in illuminating ways (Joyce's Dubliners). I don't see how these works could be biasing.

[Brian]

I can think of a good example: An Inconventient Truth.

[Eliezer Yudkowsky]

In Lawrence Watt-Evans's novels, you can never tell what is going to happen next - halfway through the novel, the dreaded evil villain can drop dead of a heart attack, and the hero can leave the emperiled kingdom and decide to become an innkeeper instead.

Bias-reversed novels abound on my bookshelves; it's one of the main thematic sources of modern speculative literature. What would happen if the Antichrist were raised as an ordinary boy? (Good Omens.) What if Prince Charming behaved like any other hormone-charged teenager when confronted with a scantily clad wicked stepmother? (Slay and Rescue.) Etc. etc.

[Robin Hanson]

Many people in essence say, "Sure most stories are bad and biased, but the good stories, the ones I like, are not biased." But every literature has to avoid making its biases too obvious to its median reader. Your literature may be more elite, because its readers notice biases more easily. But that doesn't mean it doesn't have biases; it just means that your authors had to work a bit harder to hide them.

[Eliezer Yudkowsky]

The fiction that holds us most deeply is the one that most strongly exemplifies human psychology; and if there is a bias in us, our fiction will amplify it along with the rest of us. It is imagination cut loose from the constraints of reality; what else would you expect to happen? One interesting bit of advice I read on writing fiction is that you must always shout - emphasize all the qualities of events and characters much more intensely than real life - because otherwise they won't come through to the reader at all through quiet words on paper.

To do better, read reliable history books, and try to make the events in them as available as your own memory - fiction trains us to discount stories of strange people and distant lands, but history books are not like novels, they are retellings of events that really happened to people as real as you. And if histories are not always retold accurately, well, neither is your own memory.

[Bob Unwin]

"But every literature has to avoid making its biases too obvious to its median reader. Your literature may be more elite, because its readers notice biases more easily. But that doesn't mean it doesn't have biases;"

Where are the biases in Jabberwocky, or in this poem by Pound:

In a Station of the Metro

The apparition of these faces in the crowd;
Petals on a wet, black bough.

As I said above, I think something like Waiting for Godot is similar. And there are many more examples. Of course, the authors of these works have all manner of cognitive biases. But all mathematicians are replete with cognitive biases, and yet I don't think it makes much sense to talk of biases inherent in all math papers.

More pertinently, there are many philosophy papers in which the main argument can be seen as resulting mainly from some cognitive bias. Yet since the argument is laid out explicitly in a paper, you can scrutinize its steps and decide for yourself whether it is cogent. In some cases, an argument that can be seen as resulting from a cognitive bias will turn out to be correct. For instance, someone might initially construct an argument in ethics or political philosophy with the motivation of status quo bias. But sometimes such arguments will be correct.

The problem with fiction, as I see it, is quite different from the general problem that most people we communicate with are biased. The problem is as follows. In our evolutionary past, basic inductive reasoning from observation was very useful. If every observation of a snake was followed by a dangerous snake-attack, then it was an advantage to expect attacks after future observations of snakes. But in our evolutionary past, there were no realistic, life-like movies. Hence, we probably don't have a strong ability to intuitively discount observations we make when watching movies or TV.

For instance, I grew up outside the US, but watched US teen sitcoms on cable (Saved by The Bell, Hang Time, etc.). In these sitcoms, the group of friends, which was mostly pretty white people would always include a black or hispanic person. When I came to the US at quite a young age (before I became more critical in my thought) I expected to find whites and blacks to be socially integrated, and for race to not be much of an issue anymore. Sadly this was false: there is significant social segregation in the US even among the most affluent and wealthy classes. And I had no good reason to believe otherwise, for all my evidence was drawn from fictional TV shows (with PC production teams). If I had read in an article about the amazing racial integration in the US, then I think I would have been more critical about it. But I acted as if I had *seen* the social integration, when I had only seen it on TV.

Science fiction movies and comics are also bad. You *see* Cyclops shoot energy from his eyes without there being any recoil (contra Newtonian mechanics). You see Superman lifting up buildings (old comics) without the buildings falling apart (again, contra basic physics). In all sorts of movies, you see people fighting on after numerous bullet wounds from very powerful guns.

The problem with these movies is that they lead us to subconscious inferences from 'observations' of things that didn't even happen. This is what makes them really pernicious. There are forms of literature which are about linguistic virtuosity, or about putting across some philosophical idea. These forms are more like math or philosophy papers. But movies (and some fiction) function more like the people who do spoon-bending or the quack doctors who claim to do surgery without any needles and pretend to pull bits of body tissue out of their patients.

[nilefever]

Please point to the medical/psychological/behavior studies that indicate the pathologies you're attempting to describe actually exist, i.e. having a difficult time separating faction from fiction, when you know that the fiction is fiction ...

I would argue the opposite, that blatant fiction is far better for you than documentaries (non-fiction) of dubious quality. Who facts checks? Very few people. Fiction at its worst gives you a license to consider morally inconceivable alternatives -- this may be a good thing, because it affords explicit opportunities to test and refine ones thoughts on a given subject.

A "bad" documentary might actually take the truth from you and replace it with a convenient, white lie. Believing in an un-truth can be lethal.

In fiction you bring the bias; with documentaries the author does. Fiction requires a suspension of disbelief to be effective - you know its false going in. For the fictive narrative to be effective, the audience must explicitly participate by allowing the author to establish false premises. On the other hand, except in regard to them most cynical amongst us, documentaries benefit from implicit trust; why would you watch a documentary authored by someone you didn't trust?

There are plenty of valid scientific studies concerning the ease with which humans are manipulated by those they perceive to be in authority; humans are biased to trust individuals who provide the proper credentials, wearing the right cloths, and speak the proper language.

[Stuart Armstrong]

I feel that a lot of fiction actually helps at de-biasing, or at uncovering biases we didn't suspect we have. Specifically things like historical films or science fiction - they both portray a world that is similar to ours, but with some of the background assumptions out of kilter. They do often return to a standard morality in the end, but before that you are left with questions like "could I justitifiably argue against an Inquisitor in the middle of a religious war" and "how do slaves really live their lives? Could they even be happy?" Even a manichean film like Star Wars has such inquiries - when I first saw it, I was wondering how the court of Jabba lived day to day, and what they would do after his death, and whether such a set-up was possible in the real world.

I've become aware of a lot of my own prejudices while watching fiction. Escapism brings you away from yourself - if you cling to the often standard endings, and reassure yourself that "our good and our bad are shared everywhere", then it's a waste.

But if instead you take advantage of the trip to question your assumptions and wonder how many of them depend on your current situation, then it's worth it.

[Aaron Davies]

Are you familiar with The Wind Done Gone? It's a reinterpretation of Gone with the Wind from the point of view of a slave working at Tara.

[Mike Blume]

Zubon, not every variation on a story counts as bias-reversed - the point is to reverse the main sympathy the audience is supposed to feel. Thousands of stories may be variations on Romeo and Juliet, but how many of those take the side of the kids' parents?

Well, the source material for starters - Romeo and Juliet was a cautionary tale to wayward teens before Shakespeare got a hold of it.