Brussels, Monday 22 May 1967: a spring day marred by a chilly drizzle. Two days ago, my best friend had got a new pair of shoes, but there was a problem. After he tried them on, the sales assistant had mixed up two sizes, and when he wanted to put them on the next day, the box turned out to contain a size-34 left shoe and a size-35 right shoe. So my friend’s mum had intended to get the wrong shoe swapped for the correct one that Monday. But because of the rain, and because the forecast was better for the next day, she decided to wait until Tuesday to return to the store. That store was A L’Innovation, in the Nieuwstraat, Brussels’ busiest shopping street.
And that afternoon of 22 May, more than three hundred people died in one of the largest peacetime calamities in Belgium, when shortly after lunchtime the store caught fire. When it became clear how fortunate my friend’s mum had been, the term ‘guardian angel’ popped up regularly in conversation, at school as well as in both our (at the time fairly religious) families. What other explanation could there be for the fact that she had been spared? Surely that could not have been coincidence?
This event came back to mind a little while ago, when the media reported that one of the people injured during the bombing of Brussels Airport was a teenager who had previously “survived” the terror attacks in Boston in April 2013, and in Paris in November 2015. A friend of the family said he thinks divine interventions helped the young Mormon missionary survive in the face of such extraordinary circumstances.
Both anecdotes illustrate how strong our inbuilt mechanism is for coming up with explanations of what we observe, using our prior beliefs.
I recognize that pattern
Our capacity to recognize patterns is an essential part of our intelligence, but sometimes it does make us see what is not really there. We see horses in the clouds, monsters in the wallpaper, Jesus in our food, and faces everywhere (this is known as pareidolia). And sometimes we also see causality where there is none.
Our understanding of the world is based on understanding causality, as it allows us to predict what will happen next. So when we get up to go to the toilet, and someone scores at that very moment, or when we’re in the shower and the phone rings, how do we react? We — sometimes only half-jokingly — say that it is the very act of getting up to go to the loo, or to jump in the shower, that makes a football player score or the phone ring.
And if we do so, we are in eminent company: the great scientist and skeptic Carl Sagan relates an anecdote in which he dreams that his father had died — he wakes up in a sweat, calls his dad, and of course he is fine. As Tom Gilovich relates here, Sagan was pretty clear that, had his father really died there, is no way that anyone could have convinced him that he hadn’t had a prophetic dream. Of course he knew that our dreams are full of people we know, and that these familiar people die. Hundreds of millions of people have such dreams every night, so it is inevitable that fairly regularly someone, somewhere will dream of a relative or a friend who turns out to have died the next morning.
This kind of one-sided event misshapes our perception. We don’t notice when we dream of someone who didn’t die, and we don’t remember when someone dies that we didn’t dream of the night before — just like we don’t particularly recall the times we shower without being interrupted by the phone, or when the phone rings when we’re not in the bath.
The reason is that there is nothing about these events that is salient, and it is the salient events that form an impression strong enough to kickstart our pattern recognition engine. If we have a prior belief in certain connections or laws, it’s tempting to embrace events that appear to confirm them, and to ignore everything that goes against them — that is our confirmation bias. But for every story that can be explained by a guardian angel, there is another one that challenges that belief — not least the stories of the victims of a disaster. Did the 323 dead from the department store fire not have such a protector?
It is just as tempting to let dubious causalities influence our future choices. In 2014, Malaysia Airlines suffered two calamities: in March MH 370 disappeared without a trace with 239 people on board, and a few months later MH17, carrying 298 people, was shot down above Ukraine. Bookings on the arline dropped by a third — seemingly because too many people felt this was not coincidence: surely there was something wrong with this airline! Not so: statistician Nate Silver, who after these two disasters investigated the 56 top-100 airlines that operated continuously since 1985, and concluded that air accident fatalities are exceedingly unpredictable, and that the risk of being involved in a crash barely increases after a recent disaster.
Unfortunately this tendency to see patterns that are not there is hard to suppress. Complete coincidences appear to be the result of a mysterious law or supernatural interference. But superstition is not the worst that can happen as a result of our runaway pattern-seeking.
When something goes wrong, we often feel a strong need to hold someone to account. Sometimes we take the blame ourselves — a tyre puncture just when we decided to take the scenic route home is of course the consequence of that choice. Often, though, we seek to shift it to someone else: we step on a piece of Lego barefoot, and it’s junior’s fault for not picking it up (rather than our own for not watching where we’re going, or for not wearing shoes); in the middle of cooking a complicated dish we realize there’s no nutmeg, and it’s our spouse’s fault for not replenishing it (rather than our own for not checking the ingredients before starting to cook).
Such accusations are hardly fair, but we seem almost to be hardwired to seek causation and responsibility for every event. But there is worse still: while our explanation for the kind of event just is carelessness or stupidity, sometimes we go further and we assume someone else’s actions are explained by malice.
When the boss gives us a crappy task we don’t like at all a couple of times, it is because she has got it in for us. Never mind that it happens rarely and that often we get really interesting assignments, or never mind that we actually never told her that we hate this task. Or the neighbour listening to heavy metal music on full blast, just when we are tired and want to get to sleep early: of course that is not coincidence, it’s deliberate.
As soon as we have convinced ourselves of the ill intentions of someone else, our old friend the confirmation bias, that most pernicious of all biases, makes sure that we will only notice whatever supports our mistaken assumption, and will cleverly conceal any evidence to the contrary. Some of UK Labour party leader Jeremy Corbyn’s supporters see such evil motives all over the place, as illustrated by this snippet:
The fact that the polls suggest Labour would not do well is sufficient proof for the hypothesis that the survey organizations are corrupt. In a similar vein, Donald Trump is already providing the explanation in case he loses the presidential elections: “The election is going to be rigged.”
This kind of thinking is unlikely to make us happy and to be in our interest. Few of us will go completely paranoid, but assuming agency where there is none can lead to unnecessary guilt and inappropriate blame, and to superstitious beliefs and poor choices.
We may not be able to overcome our strong pattern-recognizing and explanation-seeking processes. But we can try to control our confirmation bias, and be more circumspect before we come to a conclusion, especially if it would imply that someone, somewhere was behind the outcome of an event. We would then recognize coincidence for what it is. It was coincidence, not providence that safeguarded the young guy wounded in the Brussels Airport terror attack, and that made my friend walk around in odd shoes for nearly a year.
If we can keep tabs on our confirmation bias, we will make better decisions, and better understand the world.
At least, that is my strong belief.
Originally published at koenfucius.wordpress.com on August 5, 2016.