Repairs and reformulations can be observed in dialogues, when communicators suppose some sort of misalignment. Their main function is to re-establish alignment. This can be the case either between goals and linguistic output of one interlocutor, where a self-repair leads to an intrapersonally aligned state, or between linguistic output of interlocutor A and assumed common ground interlocutor B, where an other-repair (by B) leads to interpersonal alignment. Accordingly, repairs and reformulations can be considered as a central topic sui generis of alignment in communication. Furthermore, as instances triggered by misalignment they can be utilized methodologically to gain general insights into the conditions and mechanisms of alignment. The project starts with a reanalysis of corpus data within the theoretical framework of alignment.

This will lead to the design of experiments, in which repairs and reformulations are elicited in order to test hypotheses on conditions of alignment. Main linguistic aspects will be syntactic and semantic/pragmatic factors, a special focus will be set on the time course of alignment processes studied by reaction time and eye-movement measurements. The experimental results will be integrated into theoretical models, which in turn will be evaluated by computer based simulations.