In certain experimental contexts, neurotypical young adult listeners predict the syntactic and semantic features of upcoming words. Evidently, these predictions benefit comprehenders both in the moment (by facilitating processing) and long-term (by facilitating ongoing language learning). However, prediction effects are less robust in other neurotypical populations (e.g., healthy older adults).
Given this, how important is prediction for sentence comprehension? This question is important for models of sentence processing as well as for language rehabilitation in disorders such as aphasia. If prediction is critical for sentence comprehension, then we should consider including it in rehabilitation protocols; however, if people can get by without it, it may be better to focus clinical efforts elsewhere.
I will explore this question through a series of experiments examining linguistic prediction in people with aphasia caused by stroke and neurodegenerative disease (primary progressive aphasia). I will situate the results within a neurobiological model of sentence comprehension as well as within the broader context of aphasia rehabilitation.