Three explanations for the link between language style matching and liking

People who match each other’s language styles in dialogue tend to have more positive interactions. A person’s language style is defined by his or her use of function words (e.g., pronouns, articles), a class of short, commonly used words that make up the grammatical structure of language. The language style matching (LSM) metric indexes the degree of similarity between two individual’s patterns of function word usage. Previous research assumes that function word similarity and its positive social correlates, such as liking, result from convergence that occurs within an interaction. However, the link between language style similarity and liking may alternately be explained by two kinds of preexisting similarity. First, people tend to like each other more to the degree that they are similar in terms of attitudes, backgrounds, and personality, and these kinds of interpersonal similarity tend to manifest themselves in similar function word use. Second, processing fluency research suggests that people will process typical language styles—which are by definition similar to most other language styles in a normal population—more fluently and thus will like typical speakers more than less typical speakers. Two studies compared the relationship between liking and three measures of function word similarity (convergence, baseline similarity, and typicality) during brief conversations. Each language similarity variable was hypothesized to positively predict measures of liking individually. However, consistent with the behavior coordination literature, only LSM, a measure of within-conversation language convergence, was expected to predict liking above and beyond the other predictors. Study 1 revealed that both men and women in mixed-sex dyads were more interested in contacting their partners the more that their language styles converged during 4-minute face-to-face conversations. Men were also more interested in contacting their female partners to the degree that women’s baseline language styles matched their own. Study 2 found that men, but not women, were more interested in contacting their partners the more that they matched each other’s language styles during 8-minute online chats. Results support the hypothesis that language convergence, theoretically an index of interpersonal engagement, positively predicts quasi-behavioral measures of liking.