Warm (for Winter): Inferring comparison classes for gradable adjectives
The meaning of an utterance can change depending on the context. Yet, what counts as context is often only implicit in everyday conversation. The utterance “it’s warm outside” signals that the temperature outside is relatively high, but the temperature could be high relative to a number of different comparison classes: other days of the year, other weeks, other seasons, etc. Theories of context-sensitive language use agree that the comparison class is a crucial feature of meaning understanding, but little is known about how a listener decides upon a comparison class. We extend a Bayesian model of pragmatic reasoning to be able to reason flexibly about the comparison class intended by the speaker and test the qualitative predictions of this model using a large-scale free-production experiment. We then quantitatively synthesize the model and data using Bayesian data analysis, which further reveals that usage frequency and a preference for basic-level categories are two main contributors to comparison class inference. The methods and results we present open the door to studying richer aspects of context-sensitive language understanding.