# Who’s watching?

You can sing when you’re alone. In front of colleagues? Your boss? A study of 25,384 people finds that what we consider acceptable depends strongly on who is there.

4 September 2026 · 5 minutes

In samples from 90 societies, 14 of 15 everyday behaviours were rated less appropriate in front of peers than with nobody around. In front of managers or professors, all 15 received more restrictive ratings than in front of peers.

The study measures judgments about described situations. It does not observe how people actually behave.

## Same act. Different audience.

### When nobody else is around

15 of 15 behaviours with positive average ratings. All 15 behaviours received positive average ratings in this setting. That includes singing, crying and cursing. These are estimated averages across the samples, which does not mean that every participant approved of every act.

### When peers are present

14 of 15 behaviours rated more restrictively than with no audience. Ratings fall for 14 behaviours compared with having no audience. Talking is the exception. Kissing has the largest drop; bargaining has the smallest among those 14.

### When managers or professors are present

15 of 15 behaviours rated more restrictively than among peers. All 15 behaviours receive more restrictive ratings than in front of peers. The additional drop is largest for listening to headphones and smallest for shouting in anger.

Kissing: −2.15

Bargaining: −0.56

Listening to headphones: −1.25

Shouting in anger: −0.30

What the presence of peers changes: Kissing, Bargaining. What hierarchy adds: Listening to headphones, Shouting in anger. Drop in rating points on the same scale spanning 5 points. These are the extremes reported in the text for each comparison, not all behaviours. They are not percentages.

## Talking breaks the pattern. That is what makes it interesting.

If the explanation were simply that everything becomes inappropriate in public, talking should move in the same direction. In the study’s aggregate result, it is the sole exception to the drop associated with peers being present. Conversation usually needs a listener. That is a plausible reading of the exception, not a mechanism separately demonstrated by the experiment.

With managers or professors as the audience, even talking receives a more restrictive rating than it does among peers. The question “is it appropriate?” leaves something essential unsaid: in front of whom?

## Headphones can mean something different in front of the boss

A higher-ranking observer changes the evaluation even of acts that need not make noise. Listening to headphones has the largest additional drop compared with the peer setting: 1.25 points.

One possible interpretation is that being available to pay attention matters around someone with authority. But the study does not separate prestige from the power to evaluate or sanction you. Professors and managers can have all of these. The data cannot establish each respondent’s motive.

## The surprising cultural difference appears when nobody is watching

In societies whose participants gave more weight to avoiding harm and unfairness than to loyalty, authority and purity, the gap between being alone and being observed was larger. The authors show that this came primarily from broader approval of acts when others were absent.

That distinction matters. A larger private-to-public drop can start with greater freedom in private. By itself, it does not establish greater severity in public. The relationship between societies’ values and their ratings is an association, not proof that one causes the other.

## Something to notice tomorrow

At your next meeting, consider an ordinary act: eating, resting, laughing. Who would have to enter the room for it to seem less appropriate? What, exactly, has changed for the people present?

The study offers a useful way to ask the question. Before calling someone rude, we can name the setting and the expectation we think they violated. These results describe participants’ judgments; they do not require us to adopt the same rules.

## The study and its limits

Brent Simpson, Kimmo Eriksson and colleagues analysed the observer module of the Global Study of Everyday Norms. Data were collected between July 2023 and May 2024. The paper appeared in American Sociological Review on 27 August 2026.

The analysis includes 25,384 adults who rated at least one scenario in this module. Participants rated appropriateness from 1 to 6. The authors recentered the scale from −2.5 to +2.5. Differences in points are unchanged.

These were convenience samples, not nationally representative population samples. The mean age was 27.3 and 54.9% of participants were students. One society sample had only 16 participants. “90 societies” describes the study’s reach, without guaranteeing equal precision in every country.

The comparisons come from statistical models that account for differences between societies. The numbers here are those reported in the results section. We have not re-estimated the models and do not present a country ranking. The behaviours studied are everyday acts; the finding does not automatically extend to theft, violence or every moral rule.

Arguing, laughing, cursing, kissing, crying, singing, talking, flirting, listening to headphones, reading a newspaper, bargaining, eating, resting, shouting in anger and using a mobile phone.

[Read the research](https://journals.sagepub.com/doi/10.1177/00031224261464586)

Editorial interpretation and graphics adapted from Simpson and colleagues. Article available under CC BY 4.0.

A project by [Marius Comper](https://mariuscomper.uk/en/)