The students were seminarians, which means they had already decided to spend their lives on questions of right and wrong. On this particular day each of them was asked to walk across a Princeton courtyard to give a short recorded talk, and half of them had been handed the most on-the-nose topic imaginable: the parable of the Good Samaritan, the stranger who stops for a wounded man in the road. Between the students and the recording booth, the researchers had placed a man slumped in a doorway, head down, dressed in shabby clothes, apparently in trouble. The interesting question was never whether seminarians approve of helping a stranger. Everyone approves. The question was what they would do with their feet while holding a moral deadline in their heads.
🧠 The Experiment
In 1973, John Darley and C. Daniel Batson at Princeton ran a study they titled “From Jerusalem to Jericho.” The subjects were students at Princeton Theological Seminary, preparing to be ministers. The experimenters varied two things. The first was the topic of the talk: some students were assigned the parable of the Good Samaritan, others a neutral topic about seminary life. The second was timing. As each student was sent on his way, the experimenter said either that he had plenty of time, or that he was already running late and should hurry.
The route passed a doorway where a confederate sat crumpled on the ground. The results came out almost embarrassingly clean. Of the students who were told they had time to spare, roughly six in ten stopped to help. Of those told they were late, roughly one in ten did. The talk topic made no significant difference at all. Neither did how religious a student was. The variable that decided whether a person stopped for another person in distress was whether he was in a hurry.
⚙️ How It Works
Helping a stranger feels like a single decision, the kind of thing you either care enough to do or don’t. It isn’t. It is a chain of smaller judgments, and the man in the doorway has to survive every link before anyone bends down to him.
First you have to notice him. Then you have to read the situation as an emergency rather than as some ordinary piece of the street. Only after that does the question “should I do something?” even get a chance to be asked. Break any link and the helping never happens, no matter how kind the person is.
The chain is fragile because attention is a narrow beam, not a floodlight. When a goal takes hold of you — don’t be late — the beam tightens around whatever serves that goal, and everything else drops to background. A man slumped in a doorway on the way to a deadline is, quite literally, easy not to see.
And when he is seen, the hurry keeps working. A mind under time pressure reaches for the cheapest reading that lets it keep moving. He’s probably just drunk. He’s asleep. He’ll be fine. That reframing doesn’t feel like a decision to ignore someone; it feels like an ordinary judgment about a non-event. The emergency gets quietly downgraded, and the feet carry on.
This is what the data isolated: the situation did the deciding, and the person barely registered. But there was a twist in the fine print. Among the students who did stop, how they helped tracked their type of religiosity — some out of principle, some out of gut-level compassion. So the situation decided whether anyone acted. Character only got to decide what the act looked like once it began.
🤔 The Twist
Here is the detail that makes the study famous. The students were on their way to deliver a talk about the Good Samaritan — a story whose entire point is that one person stopped for a stranger in the road. Some of them walked past a man in a doorway to go and explain that story. The topic they were carrying made no measurable difference to their behavior.
We tend to treat morality as a trait, something a person has a store of. Surveys back this up: the large majority of people rate themselves at least above average in character, and almost nobody expects to be the one who walks past. The study suggests that when you’re already late, your self-image doesn’t get a vote. It isn’t that the seminarians were secretly cold. It’s that a person and a situation are two different variables, and the situation was the strong one.
🔗 Why It Matters To You
This reframes almost any design where you’re hoping a person will actually do something, and not just agree with it.
Think about a digital companion, the kind of product built to be present for someone. What makes a moment land is rarely the depth of the persona or the cleverness of the words. It’s whether the moment arrives when the person has any attention left to spend. The same message at 1 a.m. in a quiet room and at 6 p.m. in a taxi queue are not the same message, even if every word is identical. Presence is a timing problem as much as a personality one.
Advertising learned this years ago in the harshest way. A perfectly targeted message that arrives while someone is mid-task simply doesn’t get processed; the beam is pointed elsewhere. Placement and moment beat copy.
Writing has a version too. A kind character who walks past someone in need because she’s late reads as instantly true, because we have all been her. A kind character who always stops starts to feel like a label instead of a person. Let the pressure of the moment win a round here and there. That’s the part readers believe.
🎲 A Small Trick
Try to recall the last time you only learned afterward that someone near you had needed something. There’s a decent chance you were running late when it happened.
If that’s true, it changes where the lever sits. If you want more kindness to occur in your own day, the useful move isn’t a larger private resolution to be a better person. It’s leaving a little slack in the schedule — because the version of you with time to spare is a measurably different creature from the version who is late.