Edinburgh Napier University

  Iterated cooperation games (e.g. Prisoner’s Dilemma) are used to analyze the emergence and evolution of cooperation among selfish individuals. Uncertainty of outcomes of games is an important factor that influences the level of cooperation. Communication of intentions also has a major impact on the outcome of situations that may lead to cooperation. Here we present an agent-based simulation that implements the uncertainty of outcomes together with the communication of intentions between agents. This simulation is used to analyze the relationship between uncertainty and the complexity of the language that the agents use to communicate about their intentions. The complexity of the language is measured in terms of variability of its usage among agents. The results show that more outcome uncertainty implies lower complexity of the agent language.