An economic theory of Language

In this paper, we argue that the process by which abstract words acquire meaning in language is the result of competition between inter-acting agents in a community. We assume that individuals ’ restricted cognitive resources and limits on communication ability require the use of a reduced number of abstract concepts (words) to represent the rich reality of the world. We develop a model where agents ’ choices of the meanings of concepts is driven by two objectives. On the one hand, agents want to keep representation error minimal, i.e. use concepts that best serve their decision making process. Preferred concepts are the ones that describe the world most accurately given the agent’s idiosyn-cratic preferences. On the other hand, agents also seek to share their representations to enable communication. As preferences in the com-munity are heterogeneous, agents essentially play a non-cooperative game, the Nash equilibria of which are set(s) of meanings (concepts) shared across agents that we interpret as language(s). We show that there always exists at least one Language equilibrium but it is usually not unique. Analysis of the Language game’s adjustment dynamics reveals path-dependence and suggests that certain Nash equilibria are not accessible dynamically. As the importance of communication in-creases the average coherence of the equilibrium Language increases monotonically. The model sheds light on interesting theoretical ques-tions such as the origins of languages, the language-culture relationship or the Whorf hypothesis on linguistic relativism