A doctor's judgment about whether you have a brain tumour, diabetes or heart disease shouldn't depend on other people's medical conditions. But some of our most important judgments turn out to depend on what's prevalent.
The researchers' central experiments were deceptively simple.
In one test, they showed participants 1000 dots on a continuum from very purple to very blue, and they asked them to decide whether each dot was blue or purple. After 200 trials, they decreased the number of blue dots for about half of the participants.
As the number of blue dots was reduced, those participants became a lot more likely to categorise dots as blue. They started to "see" dots as blue that had previously looked purple.
When Levari and his co-authors increased the number of blue dots, exactly the opposite happened. Participants became more likely to see dots as purple - even dots that they had earlier categorised as blue.
At this point, you might be thinking: So what? Whether dots are blue or purple is not exactly the most pressing question facing the world today. But the dot experiment uncovers a significant social phenomenon - and it can be found in many places.
The researchers support this bold conclusion with an experiment addressing the question: What's ethical?
Levari and his co-authors asked people to say whether proposals for scientific research were ethical. When people saw a large number of obviously unethical proposals, they found less egregious proposals to be just fine. No surprise there.
But when the researchers reduced the number of obviously unethical proposals, proposals that once seemed ethical started to look, well, terrible. The upshot: Our judgments about whether behaviour violates ethical standards turn out to depend on what other behaviour we have in mind.
In the most vivid test, the researchers showed people a series of 800 computer-generated human faces, and asked them to say whether they looked threatening or not. You probably won't be surprised by what happened.
Whether people see human faces as threatening depends on what other faces they are seeing. When the number of very threatening faces was reduced, people started to see faces as threatening that they did not previously see that way at all.
Levari and his co-authors draw a powerful lesson from their experiments: Society can be a victim of its own successes. We fail to see the progress we've made.
If a nation has made serious dents in major social problems, we might not recognise what we have done, because we view existing problems in the new and improved context that we have brought about.
In the context of poverty, crime and racial equality, for example, we might end up thinking that things have got a lot worse - even if they are immeasurably better.
But there is a corollary. If things are actually getting worse, we might fail to appreciate that, or we might not see it clearly. That point helps to explain public acquiescence in the face of cruel and vicious political acts, even horrors.
The rise of Nazism is an extreme case. But consider that a searing account, quoted by Milton Mayer in his classic They Thought They Were Free, from a philologist who was in Germany at the time, emphasises "the gradual habituation of the people, little by little, to being governed by surprise."
Interviewed in the 1950s, the philologist reports that "each step was so small, so inconsequential, so well explained or, on occasion, 'regretted,'" that people could no more see it "developing from day to day than a farmer in his field sees the corn growing. One day it is over his head."
Something analogous is now happening in many places - most prominently Turkey, but also Poland, Hungary and the Philippines.
US President Donald Trump is no dictator, and thus far he has been far more bark than bite.
But there is no question that he is putting serious pressure on longstanding democratic norms. The number of blue dots (so to speak) is increasing, and people are starting to see a lot more purple.
Trump's critics like to proclaim, "This is not normal."
But what if it is?
- Bloomberg