Edition note: Independent English rewriting for English-language readers; mechanics and development claims have been checked.

Early versions of Journey allowed players to collide. They could push one another, block one another, and move another body across the environment.

Then the designers built cliffs. Playtesters began pushing their companions over the edge. Jenova Chen has recalled that he felt the impulse himself.

The team removed collision.

Designing around one’s own temptation

The decision attracts an obvious criticism. If a game deletes stealing, fighting, deception, verbal abuse, and physical interference, can the remaining civility count as virtue? People appear kind because unkindness has been made impossible.

The development story changes the force of that objection. The designer did not stand outside the players, diagnose their bad nature, and build a cage for them. He encountered the temptation in himself and removed a capacity from the system he was making.

This is not proof of human goodness. It is an admission that the desire to push need not arise from hatred or gain. Sometimes an action becomes attractive because the button exists and the result will be vivid.

Removing collision also removes innocent possibilities. Players cannot nudge a companion onto a platform, physically shield someone, or play with shared weight. Safety is not free; it narrows the range of relation along with the range of harm.

Why there are only two

The team also tested larger groups. With four players, informal factions appeared: two against two, or three together while one was excluded. No resource shortage or explicit competition was required. “Us” and “them” emerged from the group itself.

The final game reduced the encounter to two people.

This solution is honest and limited. The design does not discover how a crowd can remain open. It avoids the crowd. The tenderness of Journey depends partly on scale.

The freedom that remains

The game removes many ways to injure a companion, but it does not remove departure. You can hurry ahead, ignore a lost player, choose another route, or leave someone behind in the snow. No score falls. The ending remains available.

That leaves one action with unusual weight: waiting.

You rarely possess a tool the other person lacks. You cannot donate equipment, solve a conversation, or revive a permanent death. Often the most meaningful thing you can do is stop moving until the stranger catches up.

There is no public record of this decision. No kindness badge appears. The seconds spent waiting are inefficient by every visible measure.

The same system that removes rewards for cruelty also refuses to reward care. Waiting remains valuable only through what it does to the relation.

The problem of pace

Not everyone experiences this as liberation. Chen’s former teacher, game designer Tracy Fullerton, criticized the pressure to remain with a companion when she wanted to explore slowly and the matched player wanted to advance.

Without language, the pair cannot negotiate: “Go ahead,” “Give me five minutes,” or “I want to search this ruin.” The game has a default direction, and the faster player can follow it without experiencing that movement as a demand.

A relationship without negotiation does not eliminate power. It can hide power inside whatever pace the environment treats as normal.

Journey protects strangers from many forms of domination. It does not solve the question of whose rhythm governs two people who cannot explain what they want.

Based on thatgamecompany’s 2012 Journey, directed by Jenova Chen with music by Austin Wintory. Development claims draw on public talks and interviews. Full spoilers throughout.