Ransom, K.Voorspoels, W.Perfors, A.Navarro, D.2018-03-262018-03-262017Proceedings of the 39th Annual Meeting of the Cognitive Science Society: Computational Foundations of Cognition, 2017, pp.992-9979780991196760http://hdl.handle.net/2440/111253When the interests of interlocutors are not aligned, either party may wish to avoid truthful disclosure. A sender wishing to conceal the truth from a receiver may lie by providing false information, mislead by actively encouraging the receiver to reach a false conclusion, or simply be uninformative by providing little or no relevant information. Lying entails moral and other hazards, such as detection and its consequences, and is thus often avoided. We focus here on the latter two strategies, arguably more pernicious and prevalent, but not without their own drawbacks. We argue and show in two studies that when choosing between these options, senders consider the level of suspicion likely to be exercised on the part of the receiver and how much truth must be revealed in order to mislead. Extending Bayesian models of cooperative communication to include higher level inference regarding the helpfulness of the sender leads to insight into the strategies employed in non-cooperative contexts.enThe copyright for individual articles and figures published in the Proceedings are held by the authors, not the Society.Deception; inductive inference; communication; pragmaticsConference paper0030084130366792Ransom, K. [0000-0001-5423-6455]Navarro, D. [0000-0001-7648-6578]