Abstract
How can we communicate with people who don’t share a common language? To explore this question, this paper describes a computational model that simulates a experimental study. In the experiment, we employed a coordination task in which participants used a novel communication media that restricted usual communication methods. The model was developed to build behaivor charactersits observed in the experiment from repeated interactions of ACT-R agents. We focus on the cognitive mechanisms implemented in ACT-R architecture such as instance-based learning, imitative learning, and implicit memory process provide the scaffold for building a common communication system. From a simulation study manipulating these mechanisms, we discussed factors establishing a common communication system.