Entrainment profiles: Comparison by gender, role, and feature set

Abstract

We examine prosodic entrainment in cooperative game dialogs for new feature sets describing register, pitch accent shape, and rhythmic aspects of utterances. For these as well as for established features we present entrainment profiles to detect within-and across-dialog entrainment by the speakers' gender and role in the game. It turned out, that feature sets undergo entrainment in different quantitative and qualitative ways, which can partly be attributed to their different functions. Furthermore, interactions between speaker gender and role (describer vs. follower) suggest gender-dependent strategies in cooperative solution-oriented interactions: female describers entrain most, male describers least. Our data suggests a slight advantage of the latter strategy on task success.

Prosodic Entrainment in Conversations of Verbal Children and Teens on the Autism Spectrum

“…Entrainment can be realized as either a continuous increase in proximity over the course of the conversation such that conversation partners are more similar in a given linguistic feature later in the conversation when compared to the beginning of the conversation, a process referred to as convergence, or it can be an ongoing process of the coordination of linguistic features during which conversation partners vary linguistic features in a parallel manner—often referred to as synchrony ( Levitan and Hirschberg, 2011 ). Convergence can happen either linearly over the course of an entire conversation (i.e., Gregory et al, 1997 ; Collins, 1998 ; Kousidis et al, 2009 ; Levitan and Hirschberg, 2011 ), or at the level of the conversational turn (i.e., Levitan and Hirschberg, 2011 ; Reichel et al, 2018 ; Weise and Levitan, 2018 ). Hence, accommodation has been shown to increase linearly over the course of an entire conversation as well as to vary from turn to turn.…”

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

“…In their seminal work, Edlund et al, (2009) classified entrainment into two broad categories: synchrony and convergence. Levitan and Hirschberg (2011; see also Levitan, 2014) subsequently added an additional measure, proximity, and this terminology has been used across several studies (e.g., Reichel et al, 2018;Weise et al, 2019;Xia et al, 2014). While we argue that convergence can be viewed as a specific subtype of either other category (see entrainment dynamicity below), we support the use of the other terms, proximity and synchrony, and note that the operational definitions of these terms provided here are similar to those found in Levitan (2014).…”

“…For example, researchers often compute absolute difference scores between interlocutors and use them to compare differences between real data and sham data. Sometimes, these studies compare the differences between the features of two inconversation interlocutors to differences between non-conversation interlocutors (i.e., speakers who did not converse with one another; e.g., Levitan et al, 2012;Reichel et al, 2018). Others examined differences between interlocutors on adjacent turns compared to nonadjacent turns (e.g., Levitan & Hirschberg, 2011;Lubold & Pon-Barry, 2014).…”

“…There are a couple important ideas to note about entrainment level. First, global entrainment may include measures from individual turns (e.g., Reichel et al, 2018) or even smaller units such as phrases (e.g., Kim et al, 2011;Pardo, 2006) and even phonemes (e.g., Lelong & Bailly, 2011). Key is that in all of these instances the comparisons were not necessarily between units in adjacent utterances.…”

“…Our review of the literature provides evidence that all of these types have been used to capture entrainment (see also Table 1). The use 2014t-test comparing differences between feature values of conversation partners on adjacent and non-adjacent turns Willi et al (2018) Linear discriminant analysis using turn-level acoustic feature values Reichel et al (2018) Linear mixed models comparing the absolute difference between feature values of conversation partners in adjacent and non-adjacent turns…”