Predictive Dynamics in Language Comprehension across the Adult Lifespan

Objective

Language is fundamental for maintaining social bonds and well-being in older adulthood. However, over one-third of adults over 60 face increasing difficulties in understanding spoken language, resulting in social isolation and reduced autonomy. Predictive processing is a key neural mechanism supporting fluent language comprehension, yet its neural basis across the lifespan is poorly understood. PREDYNA-AGE aims to elucidate age-related changes in the brain’s predictive processing networks, addressing the unresolved question: Do neural dynamics in aging reflect compensation, decline, or adaptive reorganization? Filling gaps of prior research, PREDYNA-AGE maps predictive networks across the adult lifespan using ecological listening tasks with varying intelligibility, magnetoencephalography (MEG), graph theory, and deep computational language models. By including middle-aged participants and employing naturalistic auditory materials, the project systematically charts when and how predictive architectures reorganize to support real-world communication. Direct alignment between neural signals and hierarchical AI-based models will clarify the fidelity and flexibility of linguistic prediction from acoustic to semantic levels—identifying the brain mechanisms that sustain resilience and adaptability in neurocognitive aging. Planned secondments at leading European labs in MEG connectivity and computational linguistics will ensure methodological rigor, interdisciplinary expertise, and knowledge transfer. Commitment to open and reproducible science will maximize societal value and accessibility. The outcomes of PREDYNA-AGE promise mechanistic insight crucial for developing personalized interventions while addressing EU priorities in healthy aging, digital health, and AI-driven precision medicine. Moreover, the fellowship will enhance my expertise in innovative brain-language research and accelerate my path toward scientific independence in the European Research Area.