Exhaustivity and Anti-Exhaustivity in the RSA Framework: Testing the Effect of Prior Beliefs

Area of Science:

• Linguistics
• Cognitive Science
• Philosophy of Language

Background:

• Listener interpretation of utterances depends on prior beliefs.
• Exhaustivity interpretations (e.g., 'only Mary came') stem from either pragmatic or semantic processes.
• Existing models, including the Rational Speech Act (RSA) framework, attempt to explain these inferences.

Purpose of the Study:

• To investigate the role of prior beliefs in exhaustivity and anti-exhaustivity interpretations.
• To adjudicate between purely pragmatic and encapsulated semantic explanations for exhaustivity.
• To test predictions of different RSA models against new empirical data.

Main Methods:

• Developed and compared theoretical models of exhaustivity within the RSA framework.
• Designed and conducted a new study examining the effects of prior beliefs on language production and comprehension.
• Evaluated model fit against experimental data.

Main Results:

• Theoretical analysis showed RSA models can predict both exhaustivity and anti-exhaustivity.
• Experimental results indicated that models incorporating an encapsulated semantic mechanism best fit human behavior.
• Prior beliefs significantly impact both the production and comprehension of exhaustive interpretations.

Conclusions:

• Semantics plays a more significant role in exhaustivity effects than previously assumed.
• The Rational Speech Act (RSA) framework, considering informativity and cost, is essential for pragmatic effects.
• Findings suggest a nuanced division of labor between semantics and pragmatics in language understanding.