Understanding humor in human social interaction is a prerequisite to the creation of engaging interactions between humans and digital assistants, embodied conversational agents and social robots. As these HCI methods become more prevalent and pervasive, more advanced conversational discourse abilities will be required. However, many models of dialogue and human communication used in computer science remain based upon out-dated understanding of the manner in which humans communicate with one another. This paper addresses these issues introducing a view of communication based on Relevance theory and the Analogical Peacock Hypothesis in which humor and humorous interactions are viewed as ostensive challenges inviting the receiver of a communication to engage in greater levels of cognitive processing and effort to resolve the challenge set by a humorous display. The increased effort is rewarded with positive socio-cognitive effects—a humorous payoff and knowledge of the mind-reading abilities of the humor producer.
McKeown, G. (2017). Humor as an ostensive challenge that displays mind-reading ability. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10291 LNCS, pp. 627–639). Springer Verlag. https://doi.org/10.1007/978-3-319-58697-7_47