Imagine that tomorrow I were to show you a newspaper article describing a deadly wildfire. Do you think you’d be more upset upon reading that 10,000 people died than if you read that five people died?
This scenario makes people engage in affective forecasting—predicting their future emotional states. We expect that hearing about 10,000 deaths would make us sadder than hearing about five deaths.
But that’s not what happens.
Social psychologists Elizabeth W. Dunn and Claire Ashton-James ran a study in which half of the participants received short briefs about longer newspaper articles. Some got briefs saying that five people died, while the others got briefs saying that 10,000 people died. These “forecasters” were then asked how sad they would feel if they were to read the whole article, on a scale of one to nine. Not surprisingly, people in the 10,000-victim group predicted sadder responses than people in the five-victim group.
The other half of the participants were randomly assigned as “experiencers,” who read the full newspaper articles and then simply reported how sad they felt afterward. That is, they were not predicting their future emotional states, but merely reporting their current states. And a startling thing happened—the 10,000-victim group felt no different from the five-victim group. The researchers call this effect “emotional innumeracy.”
This finding supports the anecdotal observation that our emotions at least sometimes seem to ignore numerical information. In the famous quote probably misattributed to Joseph Stalin, “The death of one man is a tragedy, the death of millions is a statistic.” Indeed, there is some evidence to suggest that hearing about small numbers of people suffering is more emotionally powerful than grand tragedies!
The question is, then, why did the forecasters (incorrectly) predict that they’d be sadder upon learning that more people died?
The “dual-systems theory” of the mind and brain holds that the mind can be divided, roughly, into two systems. The “old brain” is evolutionarily older, meaning that its basic structures evolved a long time ago, and we share many of those structures with other animals. This is why, for instance, we have been able to learn so much about the human visual system by studying cat brains. The old brain is mostly concerned with perception, action, and emotion, and is primarily located near the back of the brain. The new brain is in the frontal areas, far forward in the cerebrum, and is thought to specialize in deliberative, step-by-step thinking. Evolutionarily, it took shape more recently. (This dual-systems theory was popularized by Daniel Kahneman’s 2011 book, Thinking, Fast and Slow.)
One explanation for these findings is that forecasting uses the rational new brain and actual feeling happens in the old. When asked to forecast, it’s reasonable to think that more people dead would make us sadder. The new brain has a theory about the workings of the emotional brain, and this theory is limited, and sometimes erroneous. The new brain is (relatively) good at numbers and numerical information. The old brain is not. But what if we could present the tragedy of many deaths in a language that the old brain could understand?
To test this idea, Dunn and Ashton-James ran another experiment that showed people pictures of the dead. That is, among these “experiencers,” some people simply read that 15 or 500 people died, while others saw actual photographs of the 15 or 500 people who died. (The account was fictionalized; the people in the photographs were not actually dead, but the participants believed they were.) Looking at pictures of lots of people is a different kind of information from thinking of the numbers 15 and 500, and, indeed, people reacted differently. In the picture group, the emotional innumeracy went away! Looking at pictures of 500 dead people makes people sadder than looking at pictures of 15 dead people, because the visual information is something that the old brain can process really well.
Is there a maximum number of people we can emotionally process through techniques like showing photographs? Dunn and Ashton-James’ research suggests there’s a qualitative difference between looking at 15 versus photos versus 500. But what about 9,000 photos versus 90,000? Human minds are pattern detectors, and as you add dots to a wall, eventually you stop seeing dots and just see patterned wallpaper. A texture of minuscule pictures of dead people would probably not be as emotionally moving, because you’re no longer thinking about people, as such, but about a more abstract pattern.
Our minds are set up to notice change, and to compress information into patterns when it can. It would be difficult to make a concrete image of, say, the suffering of thousands, without habituation kicking in and making people numb. Humans spent most of our evolutionary history in small groups of a couple hundred individuals (pdf), which is thought to impose a social and cognitive limit sometimes called Dunbar’s number. We simply were not built to empathize with thousands of people, let alone millions or billions.
At the same time, we know, even if we can’t feel, that less suffering is preferable to more. We can’t trust our feelings when numbers are involved—we should use our rationality instead.
For the death of one person, you can trust your old brain; for the death of millions, you should use the new one.
Cognitive scientist Jim Davies is the author of Riveted and a frequent contributor to Nautilus.