The construction of meaning.

Area of Science:

• Cognitive science
• Psycholinguistics
• Computational linguistics

Background:

• Current semantic models focus on stored information in long-term memory.
• The nature of word meaning representation remains a key debate in cognitive science.

Purpose of the Study:

• To propose a novel model (CI-2) for generating word meanings in working memory.
• To integrate semantic and syntactic information in meaning construction.
• To challenge the traditional view of a static mental lexicon.

Main Methods:

• Utilizing statistical models (Latent Semantic Analysis, Topic Model) to represent long-term memory traces.
• Developing the CI-2 dual-memory model distinguishing gist and explicit levels.
• Incorporating dependency grammar for syntactic constraints.

Main Results:

• The CI-2 model demonstrates how word meanings are dynamically generated.
• Meaning construction involves contextually relevant feature sampling from memory traces.
• Both semantic (topical) and syntactic (dependency) relevance are crucial.

Conclusions:

• Word meanings are not static entities but are dynamically constructed.
• The CI-2 model provides a framework for understanding meaning generation in working memory.
• This approach integrates memory, semantics, and syntax in language processing.