You reach an empty crossing and wait for the green signal, although the road is clear. You lower your voice in an almost deserted library. There may be no guard, fine or witness. Yet breaking the rule can still feel oddly uncomfortable.

That feeling is one sign of a social norm: a shared expectation about what people do and what they think people ought to do. It differs from a law because its main machinery is social rather than legal.¹,²

Expectations travel with us

Imagine a queue. Its exact geometry is conventional: people could, in theory, use numbered tickets or a circle. But once a community expects a line, joining at the back helps strangers coordinate fairly. If you cut in, you risk disapproval. Even before anyone says a word, you know how others are likely to read the act.

Social norms involve beliefs about behaviour and about approval. We may follow a rule because we think others follow it, because we think they expect it of us, or because we ourselves now endorse it. These motives overlap. A person may begin by avoiding embarrassment and later regard the rule as the right way to behave.¹

This helps explain why the rule persists when the audience disappears. Habits do not switch off whenever a room empties. We have rehearsed the behaviour so often that it becomes an easy default. We may also imagine how we would explain our conduct to ourselves later. The outside judge has partly become an inner one.

Rules do useful work, even when arbitrary

Some norms solve real coordination problems. Keeping to one side of a path is useful, though either side might have worked if everyone agreed. Taking turns in conversation makes exchange possible, though the precise pauses vary by culture. Calling a rule arbitrary does not mean it has no function. It may mean only that another convention could have served a similar purpose.²

Other norms are less benign. They can exclude outsiders, enforce status or outlive the problem they once solved. We should not assume that a rule deserves obedience merely because many people follow it. The same social force that keeps a queue orderly can also make an unfair expectation feel natural.

What experiments reveal

Researchers study norm compliance by comparing behaviour under different expectations and levels of observation. Some experiments find that people continue to comply even when direct monitoring is reduced, suggesting that immediate punishment is not the only driver.³ But no single experiment reveals a universal reason for every rule. Context, group identity and the meaning of the act matter.

There is also a practical wrinkle: we are rarely certain nobody will notice. A person might appear around the corner, a neighbour might remember, or a rule-breaking act might affect others later. Caution can coexist with genuine internal commitment.

When a norm changes

Norms can seem permanent until enough people behave differently and discover that others are willing to accept it. A new practice becomes easier as expectations shift. That is why visible examples matter: they change what people think others do and approve of. Yet change is uneven. A rule may weaken in one setting and remain powerful in another.

The power of a visible example

Suppose everyone in a shared office leaves their mugs in the sink. A newcomer may do the same, even if they would prefer a cleaner kitchen, because the visible pattern suggests that this is what people do. If several colleagues begin washing mugs immediately, the expectation can shift without a manager issuing a rule. Norms are often reproduced through small observations of one another.

This can create a gap between private preference and public behaviour. Many people may dislike a convention yet assume everyone else supports it. Each person goes along, and the pattern continues. Discovering that others share the discomfort can make change suddenly easier. Social expectations are powerful precisely because they are beliefs about a group, and those beliefs can be mistaken.¹,²

Enforcement is also broader than punishment. A raised eyebrow, an awkward silence or a compliment for doing the expected thing can teach the rule. People learn what counts as considerate through thousands of such signals. By adulthood, the signal may no longer be needed; the behaviour feels like common sense.

That is why different communities can disagree sincerely about what politeness requires. One person may find direct eye contact respectful, another intrusive. Neither can infer the entire social logic from an individual act. Understanding a norm means asking whose expectations are in play, what purpose the convention serves and who pays the cost of following it.

A rule may also survive because people overestimate how strongly others support it. An individual who dislikes a convention may still follow it to avoid standing out; observers see compliance and infer approval. The loop is self-reinforcing. Sometimes a small, visible change breaks it. Other times the rule is tied to power or identity and resists easy reform. Either way, the empty room is not quite empty: past interactions are present in the habit.

The interesting question at an empty crossing is not simply “Who is enforcing this?” It is “What expectations have I learned to carry with me?” Society often works through those portable expectations. We follow them because they coordinate life with others, because we care how others see us and sometimes because, after enough repetition, they feel like part of who we are.