Ok – here’s the test. It’s in two parts.
PART ONE: You are given the following rule: “If a card shows an even number on one side, then it is red on the other.” These four cards are spread before you:
1. 3
2. 8
3. RED
4. BROWN
Which cards do you need to turn over to check if the rule has been broken?
PART TWO: (Don’t cheat on Part One – commit to an answer. Remember, in this, you can cheat only yourself. And even that only for so long.) You are given a new rule: “If you borrow the car, then you have to fill the tank with gas.” A further four cards are spread before you:
1. Dave did not borrow the car
2. Helen borrowed the car
3. Bryony filled up the tank with gas
4. Kirk did not fill up the tank with gas
Again, which cards do you need to turn over to check if the rule has been broken?
ANALYSIS: The two tests, logically, are identical. However, the socially normal human performs generally poorly on questions of the first kind (scoring around 20%), while much better on the second kind (around 70%). So basically, you should be aiming to get Part One wrong and Part Two right (the answer to both is the same – turn over cards 2 and 4).
The difference is that the rule of Part Two is formatted as a social contract, and as social creatures, we are programmed to recognise and understand the rule-architecture of such contracts very well. Pure logic on the other hand, as demanded in the first part, we’re not very good at. In these specific examples, people can recognise quite easily that Bryony filling up the car with gas doesn’t mean she necessarily borrowed it, while it’s very easy to mistake “even on one side, red on the other” for a reciprocal “EVEN=RED” rule, and so erroneously turn card 3. We interpret the rule less well because it doesn’t come built into a social situation which is immediately graspable. (Logic questions using “human” terms but without social contracts – e.g. a rule like “People from California are patient; John is patient” – is answered generally as poorly as rules concerning numbers and colours).
As for psychopaths – well it turns out that like non-psychopaths they’re mostly bad at logic, but they’re bad at understanding social contracts too. By this, typical psychopaths get Part One wrong and follow this up by getting Part Two wrong too.
There’s the test. Here’s the full article.
I am not sure what this says about me. I initially got the first part right, but then got so wrapped up in different scenarios where Bryony filled somebody’s tank of gas for free – (he has a diesel truck, mistakenly buys a jerrycan of unleaded 95, drops by your house, finds you’re not home, fills up the tank as a good will gesture, leaves) – that I decided all four human scenarios needed verification, forgetting all about the non-reciprocity of the rule. Then, all muddled about diesel and petrol and who the fuck is named Bryony, I retroactively went back and changed my answer to the first question to say that I needed to look at all four cards because I was no longer sure of anything and figured I should play it safe.
Damn.
I become instantly suspicious at the first question and decided the entire experiment was rigged, skipped right over the second question to visit the article at the Economist to see who the author was, and finding none, breathed a deep sigh of relief that my paranoia was validated.
I had thought the test was a binary psychopath / non-psychopath indicator, but clearly it’s able to flag up a whole range of deformed psychologies. Here is a third reformatting of the question. This time the rule is: “If you massacre the world, then you get a black mark from God.” Which of the following cards do you need to turn:
1. Jim didn’t massacre the world.
2. Janey massacred the world!
3. Jewel got a black mark from God.
4. Joe didn’t get a black mark from God.
Hum. And if you get both problems right? (I wrote down the answers I had to both before seeing that there were only two questions. Silly me didn’t notice how long the scroll bar on the side was).