Why do we sit in silence next to a stranger on the train?

What an experiment on Chicago trains says about the first move nobody makes, and what we took from it for an AI

Our position is that the silence between two strangers on a train is often a mistake, not a wish. We base that on a 2014 study by Nicholas Epley and Juliana Schroeder in Chicago. Commuters who talked to the stranger next to them had a better ride than commuters who kept to themselves, and other commuters guessed the opposite. The step from that finding to our position is ours, not the researchers'.

What did the train study find?

The researchers recruited 118 people on Chicago's Metra commuter trains and used the answers of 97. Each person got one of three instructions for the ride to work. One group had to talk to the stranger next to them. A second group had to keep to themselves. A third group rode the way they always did. After the ride, everyone rated how it went.

The people who talked had a more positive ride than the people who stayed quiet. The paper shows the average scores only in a chart, so we give the direction here and not the numbers.

Then the researchers asked 66 other commuters to imagine the same three rides and guess how each would feel. These people did not ride. In the authors' words, they predicted "precisely the opposite pattern". They expected talking to feel worse than sitting alone.

A second train study asked the talkers how the stranger had responded. Among the talkers who sent the survey back, not a single person said the stranger refused to talk. That is what people reported, and only the people who answered. It does not prove that nobody ever gets a cold reply.

So why does almost everyone stay quiet? The authors' answer, which they say explains the pattern at least partly, is that each person thinks the stranger does not want to talk. The stranger thinks the same thing about them. Both wait, and nobody starts.

What the study leaves open

These were commuters in one city, on their way to work, and the mood results come from the riders' own ratings. The study shows what happened on one ride. It does not show that talking to strangers always makes people happier.

A 2022 paper by Schroeder, Lyons and Epley reports the same wrong guess among commuters. It comes from the same research group, and we have read only its summary. We found no independent replication that we could check.

What should an AI take from this?

The study was about strangers, not software. Our read is that the hardest part of a new conversation is the first line, and each side waits for the other to say it. Take away that first line and much of the wait goes with it.

People who use Eudaio talk with AIs named Daimons. When you start a chat with a Daimon, you meet somewhere, for a reason, and your first message is already written for you, so you never face a blank box. You can also write your own. That solves the first-line problem inside an app. On the train, the first line is still yours to say.

Our answer

Our answer is that the silence on the train is two people making the same wrong guess. In this study, the commuters who talked rated their ride better than the ones who stayed quiet. In the second train study, the talkers who answered reported no refusals. The next time a stranger sits down next to you, the first line may cost less than you think.

If you want a first conversation where the opening is already there, a Daimon starts it with you. Join Eudaio and see how a chat begins when the first message is set up for you.

Source

Epley, N. & Schroeder, J. (2014). Mistakenly seeking solitude. Journal of Experimental Psychology: General, 143(5), 1980 to 1999.

November 18, 2026