Syncing Up for a Good Conversation: A Clinically Meaningful Methodology for Capturing Conversational Entrainment in the Speech Domain

Abstract

Purpose Conversational entrainment, the phenomenon whereby communication partners synchronize their behavior, is considered essential for productive and fulfilling conversation. Lack of entrainment could, therefore, negatively impact conversational success. Although studied in many disciplines, entrainment has received limited attention in the field of speech-language pathology, where its implications may have direct clinical relevance. Method A novel computational methodology, informed by expert clinical assessment of conversation, was developed to investigate conversational entrainment across multiple speech dimensions in a corpus of experimentally elicited conversations involving healthy participants. The predictive relationship between the methodology output and an objective measure of conversational success, communicative efficiency, was then examined. Results Using a real versus sham validation procedure, we find evidence of sustained entrainment in rhythmic, articulatory, and phonatory dimensions of speech. We further validate the methodology, showing that models built on speech signal entrainment measures consistently outperform models built on nonentrained speech signal measures in predicting communicative efficiency of the conversations. Conclusions A multidimensional, clinically meaningful methodology for capturing conversational entrainment, validated in healthy populations, has implications for disciplines such as speech-language pathology where conversational entrainment represents a critical knowledge gap in the field, as well as a potential target for remediation.

Classifying Conversational Entrainment of Speech Behavior: An Expanded Framework and Review

“…Priva and Sanker (2019) introduced a different approach, linear combination, which accounts for the acoustic-prosodic features of an interlocutor using both their own baseline features and the baseline features of their conversation partner. Other studies still have advanced more complex measures such as cross-recurrence quantification analysis (CRQA), a non-linear technique, that quantifies how often and for how long the speech of conversational partners visit similar states (e.g., Borrie et al, 2019;Duran & Fusaroli, 2017, Fusaroli & Tylen, 2016. In another line of research, studies rely on perceptual measures, utilizing naïve listeners who judge the similarity of the speech between conversation partners.…”

“…There is substantial evidence that entrainment is associated with a myriad of cognitive, social, and emotional benefits in conversation. Within the realm of speech, research has shown that higher levels of entrainment are associated with stronger relationships (e.g., Lee et al, 2010;Pardo et al, 2012), increased conversational quality (e.g., Michalsky et al, 2018;Levitan et al, 2012), and better performance on collaborative goal-directed, verbal communication tasks (e.g., Borrie et al, 2019). Thus, decreased social connection, feelings of social isolation, and reduced communicative effectiveness-trends unfortunately all too common in many populations with communication disorders (e.g., Nang et al, 2018;Coelho et al, 2002;Dean et al, 2014)-may, in part, be a consequence of reduced entrainment in conversations.…”

Do Conversational Partners Entrain on Articulatory Precision?

“…We utilize the corpus described in detail by Borrie and colleagues [6], consisting of 57 experimentally elicited conversations involving 114 college-aged participants (mean age = 22.41 years), all participants were native speakers of American English with no self-reported history of speech, language, hearing, or cognitive impairment. Of the dyads, 43 were female-female, 13 were female-male, and one was malemale.…”

“…The dyads were given 10 minutes to work on the task and were encouraged to find as many of the differences as possible. Additional details of the dialogue task, instructions and recording equipment are specified in [6]. All audio files were normalized using a standard loudness normalization procedure based on a reference level and down-sampled to 16 kHz.…”

“…Entrainment has been observed in body language, lexical content, acoustic-prosodic features such as pitch and intensity, and acoustic articulatory features (e.g. spectral features) [4]- [6]. Alignment on these features has been found to be related to higher rapport, trust, efficiency, and agreement [7]- [10].…”