Evolutionary Models of Language: A Methodological and Philosophical Study

Abstract

My dissertation applies models borrowed from game theory to the evolution of language and the emergence of meaning. To justify the use of such models, I begin in Chapter 1: Quantitative Methods in Philosophy by examining the prevalence of modeling techniques and other quantitative methods in philosophy. I then argue that reliance on quantitative methods is beneficial in some philosophical fields. Chapter 2: The Constructive Empiricist's Dilemma turns to computer models. In this chapter, I argue that only scientific realism can justify the use of computational models that assess the reliability of inference methods. This is surprising, given that modelers’ urge to idealize and simplify is often claimed to be at odds with scientific realism. In Chapter 3: Cheap Talk in Structured Populations, I present a model for the evolution of signaling and show both analytically and with computer simulations that signaling evolves with significantly lower signal cost if the population is spatially structured. Drawing on similar models of signaling, I examine in Chapter 4: Ambiguous Signals, Partial Beliefs, and Propositional Content the part that propositions are supposed to play in explanations of rational behavior. Simply put, my claim is that propositions cannot account for the behavior of rational agents.