Speaker familiarity and affective involvement: Insights from the Santa Barbara Corpus

This study investigates how affective involvement in conversation is shaped by speaker familiarity,using five systematically annotated linguistic cues: laughter, co-operative overlaps, affiliativeresponses, back-channels, and filled pauses. Drawing on the Santa Barbara Corpus of SpokenAmerican English (SBCSAE), it examines whether conversational style differs between highlyfamiliar speakers, such as close friends and family, and less familiar speakers, such as colleagues orparticipants in service encounters. Multivariate tests confirm that affective involvement differssignificantly between groups, with laughter occurring more frequently in familiar interactions andback-channels and filled pauses more common among less familiar speakers. Co-operative overlapsand affiliative responses show little variation across contexts, suggesting they are less influenced bysocial closeness. In addition, a logistic regression model reveals that laughter is the strongestpositive indicator of familiarity, while back-channels are the strongest negative indicator. Filledpauses also contribute to distinguishing speaker relationships, though less prominently. Thesefindings suggest that affective involvement is best explained by the joint influence of multiplelinguistic features rather than isolated markers. The study offers a replicable framework for corpusbased research on relational talk and highlights implications for language education and the analysisof interactional behavior in socially diverse settings.