MayaWelcome to The Big Question on Slow English FM, where we explore ideas that shape everyday life. Today: Why do we trust strangers? Trust can help people cooperate, but it also means accepting some risk. So what can we learn from the signals people notice, the expectations they bring, and the ways trust may be repaired? The question matters because strangers often meet before they know much about one another. They may still need to decide whether to cooperate. That decision is not a promise that the other person will act well. It is a choice made with limited information. Research can help us examine what people notice and what may happen after trust is damaged, but each study looks at particular people and situations. We will look at those limits as well as the findings. First, what can signals tell us when we meet a new partner?

AlexStart with a simple situation: you meet someone new and need to decide whether to work with them. You do not know their history. You have to make a judgment with limited information. Researchers Frank DeSteno and his colleagues studied this kind of problem in two experiments: one with pairs of people, and one involving a person and a robot. In both settings, participants encountered a set of nonverbal cues, such as signals given without words. The researchers found that exposure to this set helped participants judge whether a new partner was trustworthy. The human-robot study used a person and a robot as partners, while the other study involved two people. The researchers also changed the cues in their experiments and found that this affected people’s trust judgments and what they did in later exchanges. These results show that cues can matter in those settings. They do not show that one gesture or expression can reveal whether any person is trustworthy.

MayaWhat did they find?

AlexIn those experiments, participants judged new partners as more trustworthy after exposure to a set of nonverbal cues. The researchers also found that changing the cues affected people’s perceptions and later exchange behavior. That is evidence about those experimental settings—not proof that one gesture, expression, or movement can tell us whether any person is trustworthy.

MayaThat distinction matters. A signal can be useful in a particular study without becoming a universal rule. And in daily life, trust is not only about reading another person. It can also depend on what people expect from one another and from the community around them.

AlexA study by sociologists Molina, Nee, and Holm looked at cooperation with strangers among manufacturing-firm CEOs in China’s Yangzi River Delta. The researchers also ran a later online experiment with managers and people who were not managers. Their results were not the same everywhere or across every group.

MayaWhat varied?

AlexIn the CEO analysis, stronger support for retaliatory norms was associated with more cooperation with strangers in Nanjing and Wenzhou. A retaliatory norm is an expectation that people may respond to harmful behavior in kind. This finding came from the study of manufacturing-firm CEOs. In a later online experiment with managers and people who were not managers, the researchers found no statistically significant overall relationship across the region. In other words, they did not find clear evidence of that overall link in the online experiment. The researchers also noted limits in the samples and uncertainty in some estimates. The two results describe different groups and settings. They do not show that retaliation causes cooperation.

MayaSo it would be too simple to say that retaliation always creates cooperation. The findings suggest that relationships between community expectations and cooperation can depend on the place and on who is studied. They are associations in a specific research setting, not a general causal rule for every community.

AlexNow consider what happens after trust is broken. An apology is one possible response, but it does not guarantee that trust will return. Yuan, Dong, and Li reviewed 18 papers, covering 36 effect sizes and 4,731 participants. Their meta-analysis found a positive average effect of apology on trust repair. This average combines results across the studies they reviewed; it does not mean each apology repaired trust. The researchers found that the average effect varied with the kind of violation and with the comparison group used in a study. So the overall result was positive, but its size depended in part on what had gone wrong and how the apology was compared.

MayaA meta-analysis combines results from multiple studies. Here, the average effect was positive, but the researchers also found that it varied with the type of violation and with the control group used for comparison. So an apology can help on average without working equally well in every situation.

AlexA different set of studies looked at the form of an apology. Gao and Yan conducted three experiments using hypothetical workplace hiring scenarios with student participants from institutions in eastern China. Across those scenarios, verbal apologies were more effective after competence-related violations, while written apologies were more effective after integrity-related violations.

MayaCompetence concerns someone’s ability to do a job; integrity concerns whether they act honestly. Those findings may sound like practical advice, but the experiments used imagined hiring situations and student participants. They do not establish a universal formula for apologies in friendships, families, or workplaces.

AlexTogether, these studies offer a careful way to think about trust. Signals may inform judgments in particular settings. Community expectations may relate to cooperation differently from one place to another. And apologies may support trust repair, though their effects vary.

MayaTrusting a stranger always involves some uncertainty. Research can show patterns in specific experiments and groups, but it cannot turn one cue or one response into a guarantee. That’s The Big Question on Slow English FM: ideas for understanding the world, one question at a time. Thanks for listening.