Grounding in Multi-modal Task-Oriented Collaboration

[Dillenbourg et al.1996] Dillenbourg, P., Traum, D., and Schneider, D. (1996). Grounding in multi-modal task-oriented collaboration. In Proceedings of the European Conference on AI in Education, pages 415–425, Lisbon, Portugal.

This study concentrate on the study the social grounding process, i.e., the mechanisms by which common ground is updated and maintained between two collaborating human agents. They refer to the definition of Rochelle of collaborative learning where the agents attempt to maintain a shared conception of the problem. They oppose this definition with the fact that in their findings a pair seems to build different shared spaces instead of one. They describe different results they had using an experimental setting called “bootnap” where two players had to coordinate and collaborate to solve a problem usign an MOO environment. They describe two main observations: a) the pair seems to build different shared spaces, through different grounding mechanisms; b) grounding is performed across different modalities.

In concluding they observe the fact that grounding seems to migrate through different modalities. A possible comment may be at the definition level if we really can compare two words like conception (which has a virtual connotation) with space (whch has a more concrete connotation).

Leave a Reply