Processes of alignment are not restricted to support language and dialog functions and the execution of speech production, but similar structured mechanisms as well support the cooperative execution of more general actions. Conceptually this can be regarded as the logical extension of alignment from the field of verbal-communicative actions into the space of non-verbal cooperative actions and as an opportunity to investigate processes bridging the areas of language processing and production on the one side and action processing and execution on the other side. Besides this extension of alignment into the action domain, we forward the hypothesis that the formation of alignment is adaptive: Repetitions of cooperative actions facilitate the appearance of and raise the level of the achievable alignment. This process of adaptation is an important component of the acquisition of team expertise. Adaptation and adaptive alignment greatly facilitates the need for explicit communication and paves the way for smooth and thus effective common acting. Operationally this project focuses on three levels of common action processing and execution: (i) an extension of joint attention in perception to a joint attention at the level of constituents of actions (e.g. focusing on relevant motor degrees of freedoms and action trackers), (ii) the level of relations between the constituents in space and time (e.g. time synchronisation of common actions, correct sequencing), and (iii) on the highest level the recognition and cooperation in task solving plans. Especially we will develop a multi-level architecture based on the idea to organise behaviour as flexible bindings of subordinate representations to (i) provide interfaces for interactive priming of action modalities, to (ii) enable top-down and bottom-up interaction between levels of action representation and execution, as well as (iii) differentiated “priming-by-observation” and “priming-by-repeated interaction” mechanisms. Alignment in these contexts appears as an increasing capability of smooth cooperative interaction of the robot with the human user.
