Infinite Mixture Chaining: An Efficiency-Based Framework for the Dynamic Construction of Word Meaning

Area of Science:

• Cognitive Science
• Computational Linguistics
• Psycholinguistics

Background:

• Word meanings evolve through semantic chaining, linking new meanings to existing ones.
• Existing models (prototype, exemplar theories) offer limited or mixed explanations for meaning extension.
• Previous approaches often present a dichotomous view of semantic chaining.

Purpose of the Study:

• To propose a unified probabilistic framework for semantic chaining: infinite mixture chaining.
• To model word meaning extension as a trade-off between representational accuracy and memory complexity.
• To provide a flexible model for the dynamic construction of word meaning.

Main Methods:

• Developed a unified probabilistic framework called infinite mixture chaining.
• Integrated cognitive efficiency as a core principle.
• Utilized computational modeling to analyze lexicon evolution.

Main Results:

• The proposed framework unifies existing chaining models.
• Demonstrated effectiveness in reconstructing historical lexicon development across languages and word classes.
• Showed correlation with human judgments of semantic change.

Conclusions:

• Infinite mixture chaining offers an efficiency-based view of word meaning extension.
• The model supports dynamic construction of word meaning through semantic clustering.
• This framework advances our understanding of semantic change and lexicon evolution.