Every human utterance, no matter how simple or straightforward, is constrained by its context and intended purpose. Accounts of pragmatic knowledge and processing are necessary to fully explain how infants learn language in the context of social interactions, and how adults use the system to communicate in actual conversations. It’s exciting, then, that the field of experimental pragmatics has been growing rapidly for the last 15 years. However, the new experimental data stand alongside fundamental but unresolved theoretical questions. What does pragmatic competence consist of? What does it mean for a sentence to “mean” something? Where do we draw the line between linguistic meaning and the rest of human cognition? In the face of such questions, the prospects for progress in experimental pragmatics may seem bleak. But I argue that the tools of psycholinguistic research are actually well suited for dealing with these issues. While traditional linguistic theory depends on defining an “idealized speaker-hearer” as the object of study, psycholinguistic models are designed to account for the behavior of actual humans, which reflects a mix of competence and performance effects and contributions from multiple cognitive systems.

In the first part of the talk, I discuss the interpretation of pragmatic processing experiments. Studies on pragmatic processing in comprehension are often intended to adjudicate between different competence theories. Similar to early work on syntactic processing, these efforts are limited by the lack of a realistic linking hypothesis to constrain predictions about behavior. (Sentence processing experiments are no better suited to distinguishing Grice from Relevance Theory than they are to distinguishing “copies" from “traces".) I outline a framework for understanding pragmatic behavior in terms of competence and multiple layers of performance mechanisms. I discuss how this framework might elucidate some apparently conflicting results on adult’s processing of scalar implicature in comprehension.

In the second part of the talk, I discuss how this framework helps to make sense of apparently contradictory claims about children's pragmatic development. Children are impressively sensitive in some pragmatic tasks, and yet often have difficulty understanding non-literal meaning. I discuss two examples from my own work that illustrate the complex pattern of children's strengths and weaknesses: 3-year-olds' interpretations of indirect requests and of belief reports. I argue that core components of pragmatic knowledge seem to be in place as early as we’ve tested, but mechanisms for deploying this knowledge in the course of comprehension undergo significant development throughout childhood.