The repair of misunderstanding and non-understanding

The difference between computers and humans in understanding language is not so much that computers are inept while people are perfect, but rather that people are flexible and have the ability to recover from the misunderstandings and non-understandings that are actually quite frequent in human linguistic communication. We  developed models of human conversational repair, basing the models upon research in psycholinguistics and sociolinguistics. To address non-understanding, Peter Heeman and Philip Edmonds developed two plan-based models of collaboration in identifying the correct referent of a description: one covers situations where both conversants know of the referent, and the other covers situations, such as direction-giving, where the recipient does not. In the models, conversants use the mechanisms of refashioning, suggestion, and elaboration, to collaboratively refine a referring expression until it is successful. To address misunderstanding, Susan McRoy developed an abductive model that combines intentional and social accounts of discourse. Reflecting the inherent symmetry of conversation, all these models can act as both speaker and hearer, and can play both the role of the conversant who is not understood or is misunderstood and that of the conversant who fails to understand.

References

Edmonds, Philip.  “Collaboration on reference to objects that are not mutually known”, Proceedings of the 15th International Conference on Computational Linguistics (COLING-94), Kyoto, August 1994, 1118–1122.  [PDF]

Heeman, Peter and Hirst, Graeme. “Collaborating on referring expressions.” Computational Linguistics, 21(3), September 1995, 351-382.  [PDF]

Hirst, Graeme; McRoy, Susan; Heeman, Peter; Edmonds, Philip; and Horton, Diane. “Repairing conversational misunderstandings and non-understandings.” Speech Communication, 15(3-4), December 1994, 213–229.   [PDF]

McRoy, Susan and Hirst, Graeme. “The repair of speech act misunderstandings by abductive inference.” Computational Linguistics, 21(4), December 1995, 435-478.  [PDF]

Graeme Hirst

Professor of Computational Linguistics

University of Toronto, Department of Computer Science

Research