Summary: Researchers at The University of Texas at Arlington found that when two strangers meet, mutual understanding develops mainly through verbal exchange—talking and asking questions—rather than through non-verbal cues like gestures or eye contact.
Source: UT Arlington
UTA psychologists report that first encounters between strangers build shared understanding largely through words, not non-verbal signals.
Psychologists at The University of Texas at Arlington used a specialized linguistic analysis to measure how much pairs of strangers “get in sync” with one another during their first, unstructured interactions. Their results indicate that the volume and content of verbal exchange—asking questions and exchanging words—predict how closely two people align in the meanings and ways they use language, a phenomenon known as latent semantic similarity.
“Beginning in the 1970s many studies emphasized non-verbal communication as crucial for first impressions and connection,” said William Ickes, co-author of the study and UTA Distinguished Professor of Psychology. “Our findings suggest that for establishing common ground in brief, initial interactions, spoken language—the exchange of words—is the primary mechanism.”
The research team, led by doctoral student Vivian Ta with co-author Meghan Babcock and faculty co-author William Ickes, published their findings in the Journal of Language and Social Psychology under the title “Developing Latent Semantic Similarity in Initial, Unstructured Interactions: The Words May Be All You Need.” The study examines how strangers come to use words similarly during short encounters, shedding light on the cognitive and social processes that underlie mutual understanding.
In their first study, volunteers from undergraduate psychology courses at UTA were paired as same-sex dyads: 26 female-female pairs and 20 male-male pairs. Each pair was placed in a “waiting room” setting and left alone for a six-minute period; their initial unstructured interaction was covertly audio- and videotaped while an experimenter stepped out. A follow-up study expanded the sample to include mixed-sex pairs as well as additional same-sex pairs, providing a broader data set for replication and comparison.
Researchers transcribed the conversations and used a linguistic program to compute each pair’s latent semantic similarity (LSS)—a measure of the degree to which partners use words in comparable ways and develop similar semantic patterns over the course of interaction. In addition to automated textual analysis, the videotapes were coded for specific verbal behaviors (amount spoken, number of questions) and non-verbal behaviors (eye contact, gestures, acknowledgements).
Analysis revealed a clear pattern: dyads that spoke more and asked more questions developed higher levels of semantic alignment. In other words, talking—especially interactive verbal engagement where partners ask and answer questions—was the strongest predictor of shared linguistic meaning between strangers.
By contrast, non-verbal behaviors commonly thought to foster connection did not reliably predict semantic similarity. Pairs who engaged in more eye contact, who acknowledged each other more frequently, or who used more gestures did not score significantly higher on measures of LSS than pairs who used these non-verbal cues less. This suggests that while non-verbal signals may influence feelings or impressions, they were not the primary drivers of semantic convergence in these initial, unstructured meetings.
“Establishing common ground is essential for effective interaction,” said lead author Vivian Ta. “Our data show that words drive that process in initial encounters—asking questions and exchanging information is what aligns people semantically.”
Perry Fuchs, Chair of the UTA Psychology Department, noted the broader relevance: “These basic findings about how strangers develop mutual understanding have implications across social and workplace contexts. It will also be important to explore whether similar patterns hold in online and telephone interactions, where only verbal exchange is possible.”
The study aligns with UTA’s strategic emphasis on Health and the Human Condition and highlights the psychology department’s expanding research profile in social psychology and interpersonal processes.
Vivian Ta is a doctoral student in the Department of Psychology at the University of Texas at Arlington. Her research focuses on semantic similarity in dyadic interactions, interpersonal relationships, and personality.
Meghan Babcock is a doctoral student in the Department of Psychology at the University of Texas at Arlington. Her interests include interpersonal relationships, attraction, semantic similarity in dyadic interactions, and individual differences. William Ickes is a Distinguished Professor of Psychology at UTA, author of Everyday Mind Reading and Strangers in a Strange Lab, and a recognized researcher in empathic accuracy.
Source: Tony Malkovic – UT Arlington
Image Source: Image credited to UTA.
Original Research: Ta, V. P., Babcock, M. J., & Ickes, W. (2016). Developing Latent Semantic Similarity in Initial, Unstructured Interactions: The Words May Be All You Need. Journal of Language and Social Psychology. Published online March 14, 2016. doi:10.1177/0261927X16638386
UT Arlington. “Strangers Reach Mutual Understanding Through Talking and Asking Questions, Not From Non-Verbal Cues.” NeuroscienceNews. June 8, 2016.
Abstract
Developing Latent Semantic Similarity in Initial, Unstructured Interactions: The Words May Be All You Need
This study examined how strangers develop latent semantic similarity (LSS)—the tendency to use words in the same way—during initial interactions. Building on prior work, researchers analyzed multiple samples of unstructured same-sex and mixed-sex interactions. Across datasets, only behaviors that introduce words into the conversation (speaking and asking questions) uniquely predicted the development of dyad-level LSS. These results support the idea that verbal exchange alone can produce semantic alignment, suggesting that non-face-to-face, computer-mediated conversations that rely solely on words may achieve similar development of shared meaning as face-to-face encounters.