Natural Leaders: Some Interlocutors Elicit Greater Convergence Across Conversations and Across Characteristics

Abstract

Are there individual tendencies in convergence, such that some speakers consistently converge more than others? Similarly, are there natural "leaders," speakers with whom others converge more? Are such tendencies consistent across different linguistic characteristics? We use the Switchboard Corpus to perform a large-scale convergence study of speakers in multiple conversations with different interlocutors, across six linguistic characteristics. Because each speaker participated in several conversations, it is possible to look for individual differences in speakers' likelihood of converging and interlocutors' likelihood of eliciting convergence. We only find evidence for individual differences by interlocutor, not by speaker: There are natural leaders of convergence, who elicit more convergence than others across characteristics and across conversations. The lack of similar evidence for speakers who converge more than others suggests that social factors have a stronger effect in mediating convergence than putative individual tendencies in producing convergence, or that such tendencies are characteristic-specific.

Phonetic Dissimilarity and L2 Category Formation in L2 Accommodation

“…For example, if the pre-distance is larger than the post-distance, DID would be a positive value that indicates convergence. However, several studies have challenged using DID to measure the degree of convergence (Cohen Priva & Sanker, 2019, 2020; Kim et al, 2011; MacLeod, 2021). Cohen Priva and Sanker (2019) discussed potential issues with DID and proposed a linear combination method that takes into consideration the participants’ and interlocutors’ baselines.…”

“…DID is also likely to introduce the starting distance bias (MacLeod, 2021) to measurements because when the starting distance between speakers and their interlocutors is small, “noise from random variability is more likely to overshadow actual convergent shifts” due to the little room left for accommodation (Cohen Priva et al, 2017, p. 4). The linear combination approach, as proposed by Cohen Priva and his colleagues (Cohen Priva et al, 2017; Cohen Priva & Sanker, 2019, 2020), is a method that uses factors such as the speaker’s baseline production and interlocutor’s production in a statistical model to eliminate the bias caused by DID.…”

Underpinnings of explicit phonetic imitation: perception, production, and variability

“…The list of social or personality-related factors proposed to condition imitation expands even further when considering the better-studied domain of phonetic convergence (see Wade et al, 2020 for a review). While intriguing, findings can be inconsistent and often fail to replicate across studies, making it difficult to draw firm conclusions about the individual characteristics governing imitation (Cohen Priva & Sanker, 2020;Wade, 2022).…”

Bonding Through Language Imitation: Influence of Social Ostracism on Structural Priming

“…However, these two theoretical accounts seem to focus primarily on linguistic factors. In recent years, an emerging line of research suggests that structural priming is not an independent cognitive process but conditioned by various socio-cognitive factors (Cohen Priva & Sanker, 2020; Heyselaar et al, 2017; Nitschke et al, 2014; Suffill et al, 2021). For example, Weatherholtz et al (2014) found that participants who rated the speaker's accent as more standard and who favored a compromising conflict style showed more syntactic coordination in dialogue.…”