Summer School: Techniques in Interaction and Conversation Analysis

tba.

tba.

The multimodal turn in conversation analysis has sparked a renewed interest in the interactive functions of gaze. In this practical workshop, we will explore a technology and method that can continuously and precisely capture and analyze the gaze behavior of participants during social interactions: mobile eye-tracking.

We will begin with a brief joint reflection on the various methods of video recording social interactions, and the extent to which these methods capture the participants‘ perspectives in a manner consistent with CA’s interest in reconstructing participant behavior. I will then provide a brief introduction to the technical aspects of eye tracking in real-life situations using mobile glasses, such as the Tobii Pro Glasses 3, after which we will record a short interaction together. Finally, we will export this data together to see what raw eye-tracking data looks like, and how video recordings can be synchronized to create a split-screen video ready for transcription and analysis.

The second part of the workshop will see participants working together in small groups to focus on a particular gaze-related research question. Eye-tracking data ready for analysis will be provided for this purpose. The workshop will conclude with a short presentation of the group work, after which there will be an opportunity to discuss the potential applications of eye-tracking in one’s own research projects.

For preparation, participants are advised to read Zima, E., Auer, P. & Rühlemann, C. (2025). Why multimodal interaction research on gaze needs mobile eyetracking. In: Stukenbrock, A. & Zima, E. (eds). Mobile Eyetracking: new avenues for the study of gaze in social interaction. Amsterdan: John Benjamins, 31-73.

In this workshop, we will examine the significance of prosody for social interaction from the perspectives of conversation analysis and interactional linguistics. Starting by delimiting the field conceptually, we will first take a general look at how interactants use prosody as a resource to fulfill interactive functions—namely, to form and delimit units, to organize turn-taking, to constitute social actions and to contextualize affect, communicative genres, and more. In this theoretical overview, I will focus on the intonational shaping of particle-based (responsive) actions, which we will then explore further in the second practical part of the workshop using data from natural conversations. For this purpose, we will use data from the Database for Spoken German (DGD/FOLK, dgd.ids-mannheim.de) and the program PRAAT (praat.org). Please register with the DGD well in advance of the workshop. If you would like to discuss your own data and research questions (you’re welcome to!), it’s best to contact me beforehand (alexandra.gross@uni-bayreuth.de).

This workshop offers a hands-on introduction to the technical workflows involved in processing video material after acquisition, guiding participants from raw footage to analysis-ready data.

We will begin with preprocessing and conversion, covering the essential steps needed to prepare raw video files for further editing and analysis. From there, we move into DaVinci Resolve, exploring its capabilities for image transformation, cropping, and compositing — techniques that allow for precise control over framing, alignment, and the combination of multiple video sources into a coherent final product.

A key focus of the session will be automatic anonymization, demonstrating how DaVinci Resolve can be used to protect the identity of individuals appearing in footage, an increasingly important consideration when working with sensitive or personal video data.

The workshop will then turn to data preparation and corpus creation, where we will introduce tools and workflows for automatic transcription, showing how spoken content can be efficiently converted into text to support the construction of structured, searchable corpora.

Finally, we will explore automated systems for movement visualization, comparing approaches based on point-based motion tracking and full body pose reconstruction, using tools such as VITPose among others. This section will highlight how these methods can be applied to analyze and visualize human motion directly from video data.

By the end of the workshop, participants will have a practical understanding of the full pipeline — from raw video acquisition to anonymized, transcribed, and motion-annotated datasets ready for research or analysis.

Many researchers in Conversation Analysis and Interactional Analysis work with ELAN. If the research aims are qualitative in nature, ELAN is easy to work with. If the goal is a quantitative analysis based on numerous ELAN files and numerous variables of interest, exporting and processing of ELAN annotations requires coding capacities.

This workshop is designed to provide a practical introduction to doing quantitative analysis of ELAN data in R. We will demonstrate and practice the following methods:

• how to export into R multiple ELAN files at once
• how to structure the ELAN data in a data frame in R
• how to analyze and visualize the data in R

A basic knowledge of R and Regex is required to make the most of this workshop and to be able to apply the proposed methods to your own ELAN data. Participants may feel free to bring their own ELAN data.