How We Describe Friends Reveals Cultural Differences in Reading Relationships

When asked to describe a close friend, most people assume that the words they choose are simply a portrait of that individual. But new research suggests that the way we talk about the people in our lives is shaped just as much by culture as by the person being described. A study published in BMC Psychology by Cherie Zhou, Eun Ju Son, and colleagues at Queen’s University, working with collaborators in China, compared how Chinese and Euro-Canadian participants described and judged close friends and acquaintances. The findings reveal a striking divergence: the two cultural groups not only describe others differently depending on relationship closeness, they also draw opposite conclusions when reading descriptions written by someone else.

The research team set out to answer a deceptively simple question: what do our words say about our relationships? Person description is one of the most common social activities humans engage in. We introduce friends to others, write recommendations, gossip, and evaluate the people around us in everyday conversation. Because such descriptions are so routine, psychologists have often treated them as a transparent window onto social perception. Yet the new study argues that the window itself is tinted by cultural norms about what a close relationship is and how closeness should be expressed in language.

Across three studies, the researchers compared Chinese and Euro-Canadian participants on two related tasks. The first task was production: participants wrote descriptions of a close friend and of an acquaintance. The second task was inference: participants read descriptions of a person and judged whether the description had been written by a close friend or by an acquaintance. This production-and-inference design allowed the team to examine both how people encode relationships into language and how they decode relational cues from the words of others, two processes that must be culturally coordinated for communication about relationships to work smoothly.

The results from the description task showed a pattern unique to the Chinese participants. When Chinese participants described a close friend, their descriptions were more semantically dissimilar from one another than when they described an acquaintance. In other words, the language they used for close friends diverged more sharply from the language they used for other targets. Euro-Canadian participants, by contrast, showed no such difference: their descriptions of close friends and acquaintances did not differ systematically in semantic similarity. This asymmetry suggests that for Chinese participants, relational closeness leaves a measurable fingerprint on the vocabulary and content of person descriptions, whereas for Euro-Canadians the act of describing a person appears comparatively independent of the type of relationship involved.

One way to interpret this pattern is through long-standing theoretical accounts of cultural differences in self-construal and social orientation. East Asian cultural contexts, shaped by Confucian traditions, tend to emphasize interconnectedness, relational roles, and the embedding of individuals within social networks. In such contexts, describing a person may naturally involve situating that person relative to oneself and to the relationship, so that close relationships are represented in qualitatively different semantic territory than distant ones. North American cultural contexts, which emphasize individual traits, personal attributes, and independence, may encourage a more trait-focused style of person description that applies similarly regardless of who the target is. The semantic dissimilarity observed among Chinese participants is consistent with the idea that closeness reorganizes the conceptual space of person perception in cultures where relationships are central to identity.

The inference studies, which formed the second and third parts of the research, produced an even more provocative result. Participants were shown descriptions of a person and asked to judge whether the writer was a close friend or an acquaintance of the person described. Here the two cultural groups again diverged, and in mirror-image fashion. Chinese participants were more likely to infer relational closeness when the description was semantically dissimilar from a reference description, whereas Euro-Canadian participants were more likely to infer closeness when the description was semantically similar. The same linguistic signal carried opposite social meanings depending on the cultural background of the reader.

This reversal has important implications for how we think about cross-cultural communication. If a Chinese speaker and a Euro-Canadian speaker each wrote about a mutual friend, an observer from the other culture might misread the relationship. A description that signals deep intimacy to a Chinese reader might signal mere acquaintance to a Euro-Canadian reader, and vice versa. The study suggests that the implicit code linking language to relationship type is culturally specific, and that people are generally unaware of their own code when interpreting the words of others. Misunderstandings arising from such mismatched inferences could affect first impressions, hiring decisions, reference letters, and even the interpretation of social media posts, all contexts in which third-party descriptions of people carry social weight.

The methodological approach also highlights the growing role of computational text analysis in cultural psychology. By quantifying the semantic similarity of descriptions, the researchers could move beyond intuitive judgments about how people talk and instead measure the structure of language use directly. Semantic similarity metrics capture how much overlap exists in the meaning of two pieces of text, allowing researchers to detect subtle differences in the conceptual content of descriptions that participants themselves might not notice. Applying such measures to open-ended person descriptions offers a way to study social relationships that does not rely on self-report questionnaires, which can be vulnerable to cultural differences in response styles and in the meaning of scale points.

The findings add to a substantial body of work showing that basic social cognitive processes, long assumed to be universal, vary systematically across cultures. Person perception has traditionally been studied within Western samples, and the resulting models have emphasized the detection of traits and dispositions. The new study demonstrates that even the seemingly straightforward act of describing another person is filtered through culturally learned assumptions about relationships. What counts as evidence of friendship, in language as in behavior, is not a human universal but a cultural convention that must be learned and that can differ profoundly between societies.

The authors suggest that these results underscore the importance of considering cultural context in understanding interpersonal relationships and social perception. For a field that increasingly operates on a global scale, from international collaborations to cross-cultural teams and multicultural societies, the message is practical as well as theoretical: the words people use about one another encode relational information in culture-specific ways, and reading those words correctly requires more than fluency in the language. It requires fluency in the cultural logic that connects words to relationships. The research, which was supported by grants from the Social Sciences and Humanities Research Council of Canada, was approved by the institutional Research Ethics Board at Queen’s University, and all participants provided informed consent. The article is published open access, allowing readers worldwide to examine the full details of the three studies and the analyses behind these findings.

Subject of Research: Cultural differences in person descriptions and social inferences about relationship closeness

Article Title: What do words say about us: cultural differences in person descriptions and inferences

Article References: What do words say about us: cultural differences in person descriptions and inferences. (n.d.). https://doi.org/10.1186/s40359-026-05769-8

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05769-8

Keywords: cultural psychology, person perception, social inference, cross-cultural studies, social relationships, semantic similarity, close friends, acquaintances, self-construal, language and cognition, BMC Psychology, social perception