Language Evolution and Information Theory – Plotkin & Nowak (2000)
Links Evolutionary Game Theory to Shannon’s Information Theory. Shows that with a noisy channel their is a limit to how much adding signals can improve fitness. The limit can be achieved by using combinations of signals to form words so that now a group of signals designate a state rather than individual signals. This allows the communication channel to reduce error exponentially in the number of signals and surpass the fitness limit of a non combinatorial signal system. Provides a framework for modeling different levels of linguistic structure in a probabilistic manner and to examine the dynamics of such systems.