Models of Language and Multiword Expressions.

Area of Science:

• Linguistics
• Cognitive Science
• Computational Linguistics

Background:

• The dominant linguistic model separates language into words (lexicon) and rules (grammar).
• This words-and-rules framework has been influential but shows limitations in explaining real-world language use.

Purpose of the Study:

• To review evidence highlighting the significance of multiword expressions in language.
• To question the traditional lexicon-grammar dichotomy based on the role of multiword units.
• To explore alternative theoretical frameworks for understanding language representation and processing.

Main Methods:

• Review of evidence from language acquisition studies.
• Analysis of sentence processing research.
• Examination of computational modeling approaches.
• Synthesis of diverse theoretical perspectives on multiword expressions.

Main Results:

• Multiword expressions (idioms, collocations) are central to linguistic knowledge organization.
• These expressions blur the distinction between lexicon and grammar.
• Existing models struggle to fully incorporate the role of multiword units.

Conclusions:

• The traditional lexicon-grammar distinction may be insufficient for a complete understanding of language.
• Multiword expressions necessitate a re-evaluation of foundational linguistic theories.
• Further research is needed to develop comprehensive models accounting for multiword expression processing and acquisition.