This project examines repair organization in asymmetric interaction. Interactional repairs enable change and facilitate learning of language and other cognitive skills. We study conversational problems stemming from deficits in cognitive, linguistic, motor, and sensory-perceptual levels of human performance, unpack the interactional repair activities handling the problems, and the distribution of labor in repair work between the participants.
The data are videotaped conversational interactions involving participants with cognitive (dementia, autism), linguistic (adult aphasia and developmental language disorder), motor (dysarthria), and sensory-perceptual (hearing impairment) deficits. Different atypical data sets are compared with each other and with conversational interactions of adults and children without impairments.
The results provide new theoretical insight to the fundamentals of human conversational interaction in occasions of communication breakdowns and on the effects of perceptual, motor, linguistic and cognitive deficits on the management of intersubjective understanding between the interlocutors. The results will be used to develop ecologically valid assessment and intervention methods that have potential to generalize into the everyday life of the people with communication impairments.
Funding: Academy of Finland, decision no. 333858, 28.5.2020, decision no. 365567, 28.11.2024 COMPAIR. Comparing repair analytic tool for the assessment of communicative functioning.
In January 2025, a two-year Proof of Concept (POC) project was launched as part of the COMPAIR research project. This POC aims to explore the potential of artificial intelligence (AI) in analysing conversational interaction data collected during the COMPAIR project.
The main objective of the POC project is to train AI models to detect and analyse repair sequences in conversational interaction data. The data comprises video-recorded interactions involving participants with cognitive, linguistic, motor, and sensory-perceptual deficits. The current focus of the POC project is on identifying and evaluating AI tools that can reliably detect repair sequences in the data collected.
From left: Kerttu Rautio, Hodan Ali, Iida Kinanen, and Johanna Puska work on the POC project in summer 2025.
Kati Pajo, Seija Pekkala, Satu Saalasti, Inkeri Salmenlinna, Leena Tuomiranta, PhD students Melina Meritähti and Minea Tikkanen, Research Assistant Heidi Liljeblad
Scott Barnes (Macquarie University), Suzanne Beeke (University College London), Steven Bloch (University College London), Andrea Bruun (University College London), Katie Ekberg (University of Queensland), Barbara A. Fox (University of Colorado), Carolyn Rickard (University of Colorado), and Marja-Leena Sorjonen.