Does linguistic synchrony between therapist and client predict better self-disclosure?

Interpersonal synchrony — the responsive convergence of communicative behavior between interlocutors — manifests across modalities: body movement, vocal pitch, and linguistic style. In therapeutic settings, synchrony is associated with building affiliation, cooperation, and rapport, and is specifically identified as critical in therapist-client relationships.

This study operationalizes linguistic synchrony through the normalized Conversational Linguistic Distance (nCLiD) and evaluates it alongside two measures of client self-disclosure quality: descriptive intimacy (disclosure of private facts), evaluative intimacy (disclosure of personal opinions and feelings), and engagement (active participation beyond presence). The key finding: higher synchrony is significantly associated with higher intimacy and higher engagement, supporting the hypothesis that a therapist's linguistic synchrony encourages greater self-disclosures.

When LLMs are compared to trained therapists and non-expert online peer supporters in a CBT setting, the LLM is outperformed by both groups. This matters because peer supporters have no formal training — they are volunteers with basic conversational skills. If LLMs cannot match even untrained human synchrony, the deficiency is not in clinical technique but in the fundamental conversational responsiveness that makes dialogue feel reciprocal.

Since Can we measure empathy and rapport through word embedding distances?, synchrony appears to be a converging metric from multiple measurement approaches (WMD and nCLiD). The practical implication: synchrony could serve as an automatic quality metric for therapeutic AI, complementing traditional outcome measures. But the LLM underperformance suggests current models lack the adaptive linguistic mirroring that emerges naturally in human dialogue.

Source: Psychology Therapy Practice