Cognitive Aspects of Verbal Interaction

2010-09 till 2012-10
Research Areas: 

CAVI focuses on the investigation of verbal interaction in humans and artificial systems that are inherently impaired with respect to their verbal, cognitive and social skills. The project aims to evaluate the characteristics of successful verbal interaction or communication and investigates the cognitive and social prerequisites of successful interaction.

Methods and Research Questions: 

Verbal interaction concerns a socially highly relevant aspect of cognition, which does not constitute an isolated but rather an interactive process: While human interlocutors can easily judge whether an interaction is successful, little is known about the cognitive and social prerequisites of successful verbal interaction. Interacting partners often differ in their verbal, cognitive and social skills. The basic question is - given these differences - how do these affect verbal interaction: Do interlocutors change their verbal interaction behavior depending on reduced verbal, cognitive or other interactive skills of their dialog partner? And if they do so, in which direction does this process occur? Is there a hierarchical order of higher and lower skills that are more likely to be impaired when verbal interaction fails? Do interlocutors with subtle language disorders or in interaction with technical systems show a certain interactive behavior? And are there certain characteristics that indicate even subtle communication impairments? On the basis of behavioral data, possibly extended by an investigation of the neuronal correlates of verbal interaction, we expect to make a relevant contribution to the growing discipline of neuropragmatics in interaction.


We investigated verbal interaction in healthy subjects to identify characteristics of successful communication (Thiele et al., in prep.). In the disguise of a dialog game, we examined the potential influence of cognitive (e.g. working memory) and social factors (e.g. interlocutor’s native language) on the use of lexical terms (with high vs. low naming agreement among speakers) in participants interacting with a native and a non-native dialog partner. Results showed that participants changed their verbal behavior depending on whether they interacted with a non-native or a native speaker. Additionally we found that cognitive abilities did not account for differences in verbal interaction that may contribute to successful communication, whereas social-strategic factors did.
Our ongoing research aims to specify subtle or minimal impairments of verbal interaction as well as to identify its underlying mechanisms (e.g. cognitive prerequisites; social skills). A relevant question is: Can subtle impairments of interactive verbal behavior (e.g. in patients with minimal traumatic brain injury) provide a model for Human-Robot-Interaction (technical systems seen as “inherently impaired” with respect to their cognitive and social skills).