Within this part of the iTalk-project, we propose to analyse parent-child-interaction in order to develop and implement models that simulate interactional aspects of imitative action learning. While previous research has focused on finding structural similarities of action presentations, close examination of video-recordings from both experimental and everyday parent-child-interaction brings to light a high variability in the tutors’ multimodal (i.e. verbal and bodily) conduct. Undertaking fine-grained analysis of video recordings from task-oriented parent-child-interaction, we attempt to investigate for which modalities variability can be found vs. which parts of a task are performed with consistency across subjects. Combining qualitative and quantitative approaches, we will ask at which levels of analysis variability occurs, and how this might link to conveying semantic structure of actions. Taking up an interactional perspective, we will further examine the recipient’s impact on the variability in the tutors’ actions, including phenomena such as shifting focus of attention and its interactional consequences. – With this background, we will consider the following aspects of conduct in imitative learning: multimodal action structuring, saliency, the sequential interplay of tutor’s and recipient’s behaviour as well as the interactional organisation of recipiency.
For futher information conntact Dr. Karola Pitsch.