Typicality effects in contingency-shaped generalized equivalence classes.

Area of Science:

• Behavioral Psychology
• Cognitive Science
• Linguistics

Background:

• Lexical classes involve arbitrary and perceptual relations.
• Understanding how such classes form is crucial for cognitive and linguistic theories.
• Previous models often lack empirical validation through behavioral experiments.

Purpose of the Study:

• To model the formation of lexical classes using behavioral methodologies.
• To investigate the role of arbitrary and perceptual features in class formation.
• To compare experimentally derived stimulus classes with natural language categories.

Main Methods:

• Two experiments employed match-to-sample procedures with nonsense syllables and abstract stimuli.
• A one-to-many mapping trained participants on critical features defining class membership.
• Probe tests assessed symmetry, transitivity/equivalence, and generalization with novel stimuli.

Main Results:

• Stimuli successfully formed generalized equivalence classes, demonstrating open-ended learning.
• Novel stimuli were incorporated into classes based on feature inclusion.
• Typicality effects emerged: stimuli with more critical features were learned more efficiently and judged as better exemplars.

Conclusions:

• Match-to-sample procedures can establish contingency-shaped stimulus classes.
• These experimentally derived classes exhibit key behavioral similarities to natural lexical categories.
• The findings support a behavioral framework for understanding language acquisition and categorization.