Proceedings of the Annual Meeting of the Cognitive Science Society

One of the hallmarks of human natural language (NL) inter-action is the ability for people to balance a variety of so-cial and communicative goals when choosing how to realizetheir speech actions. These goals can include pragmatic criteriasuch as correctness, informativeness, and brevity (i.e., Griceanconversational maxims) or social factors such as politeness.However, there currently does not exist a general algorithmicmethod to explicitly modulate language generated by artificialagents based on an arbitrary number of pragmatic and socialcriteria. We propose a novel method to accomplish this task,in which rankings of candidate utterances by different prag-matic or social criteria are fused by use of a voting algorithm.We then give a proof-of-concept demonstration of the applica-tion of this method in the context of directive generation forhuman-robot interaction.