The narrative fallacy

People like narratives, but people also like heroes, straight lines, unidimensional developments, events that happen for a reason and destiny. If we tie all these things together and construct it into a straightforward narration of cause and effect, and omit the details that don’t fit into the main line, we have a success story, or just the opposite, a story in which the hero is doomed to go down. But beware, if you have done this, you may have committed the narrative fallacy. It’s a natural tendency of humans to make such a “summary” story, for we like clear and simple lines that make the world around us understandable. However, often the world is not that way. It’s full of confusion, disorder, coincidences, and complexity, but it is difficult to live with that and therefore we look for order, intentions and simplicity. Didn’t Hume already say that cause and effect are in the mind?
The term “narrative fallacy” was coined by Nassim Taleb – a Lebanese-American essayist, mathematical statistician, former option trader, risk analyst, and aphorist – in his book The Black Swan, and further developed by Daniel Kahneman (pp. 199-200) and others. The essence of the narrative fallacy is that we prematurely create a coherent story from disconnected bits and pieces of information and leave out other pieces of the story that are relevant, either because we don’t know them or because we think that they aren’t relevant. The point is, however, that the pieces of information that we connect may have no relation at all, and that pieces that are left out may be essential. Such a misleadingly constructed story may lead to false conclusions about success or failure. Descriptions of successful life stories are often made this way: The successful businessperson had a good education, a stimulating father, took the right decisions at the right moment and knew where to invest. However, the moments that the person had luck, the contributions of others and the like are often not mentioned. Then we might think that someone with the same education, stimulating parent and drive simply must be able to get the same results. But look around, many people are like this businessperson but luck was against them because of an unexpected drop in the stock market, others were working against them, etc. Your personal contribution is only one factor in your success (or failure).
Today, it is difficult to ignore AI, and after I had typed “narrative fallacy” in my Google search engine, unrequested (for I was looking for articles) I got a summary of its most important elements. Here they are:
- Oversimplification: We reduce complex realities (like market trends, personal successes, historical events) into tidy, memorable narratives, ignoring contradictory information.
- Illusion of understanding: The stories give us a false sense of comprehension, making us believe we grasp cause and effect when we’re just connecting dots.
- Underestimating luck: We attribute outcomes to talent or intent (hard work, intelligence) rather than to chance; success seems inevitable for those who follow the “story”.
- Impact on decisions: In finance, this leads investors to chase “good stories” over data; in life, it makes us believe we can predict the future based on past narratives. However, as I want to add, “bad stories” and failures are left out, though these are (together with the good stories) important to estimate how the chances of success are.
Examples of narrative fallacies are:
- Success stories in which the success is stressed but the moments where it might have gone wrong are left out. Chance is often described as just taking the right decision.
- Financial markets: Explaining a stock’s rise with a simple story, even if multiple unknown factors are at play.
- Conspiracy theories: Crafting elaborate stories to connect unrelated events, blaming a single entity, like a person, a group or a country, in a simple manner, while leaving out many other factors and “mechanisms”.
However, as Alexey Tolchinsky says in Psychology Today: “The examples of narrative fallacy need not be extreme; we are all prone to this phenomenon. It’s a human proclivity to connect the dots quickly, which is likely related to the limitations of our minds—it is effortful for us to hold on to multiple unconnected details.”
Avoiding the narrative fallacy is above all a matter of being aware of the problem, especially, so AI:
- Focus on data and probabilities rather than just stories.
- Acknowledge the role of randomness and luck.
- Be sceptical of simplistic explanations for complex events.
Nevertheless, being aware is often not the solution but the problem itself: People often think that they are aware of the problem and next interpret the facts just so that they fit into their prejudiced stories, while leaving out, just because of their prejudices, relevant facts as irrelevant. Is it surprising? As Kahneman says (ibid.): “Narrative fallacies arise inevitably from our continuous attempt to make sense of the world.” However, so he continues: “The explanatory stories that people find compelling are simple; are concrete rather than abstract; assign a larger role to talent, stupidity, and intentions than to luck; and focus on a few striking events that happened rather than on the countless events that failed to happen.” Could it be else? “[N]o story can include the myriad of events that would ever have caused a different outcome. The human mind does not deal well with nonevents.” We have come full circle.