Meaning and Use in the Expression of Estimative Probability

Area of Science:

• Cognitive Science
• Computational Linguistics
• Psychology

Background:

• Words of estimative probability (WEPs) express uncertainty but are theorized to have crisp thresholds.
• Experimental data suggest WEP usage exhibits gradience and focality, challenging current semantic theories.
• Understanding WEPs is crucial for modeling human communication under uncertainty.

Purpose of the Study:

• To computationally model the production of WEPs.
• To compare the explanatory power of different semantic models for WEP usage.
• To investigate the relationship between autistic traits and WEP production rationality.

Main Methods:

• Development and comparison of computational models for WEP production.
• Analysis of novel WEP production data.
• Validation of models using the Autism Spectrum Quotient (AQ) test to assess autistic traits.

Main Results:

• A threshold-based semantics model explains WEP production data as well as a model incorporating gradience and focality.
• Cognitive limitations and goal-directed speech assumptions improve model fit.
• Communicative difficulties associated with autistic traits correlate with a reduced rationality parameter in the WEP production model.

Conclusions:

• Both simple and complex semantic models can account for WEP usage patterns.
• Individual differences in communicative abilities, such as those related to autistic traits, impact pragmatic language choices.
• Computational modeling provides a framework for understanding WEP semantics and their link to cognitive and communicative factors.