A Global Network for Lambda Grammars and Abstract Categorial Grammars

  • Muskens, Reinhard (Principal Investigator)
  • Francez, Nissim (Co-Investigator)
  • de Groote, Philippe (Co-Investigator)
  • Kanazawa, Makoto (Co-Investigator)

Project: Research project

Project Details


Philippe de Groote's Abstract Categorial Grammars (ACGs) and Reinhard Muskens' Lambda Grammars (LGs) are two new approaches to the logical architecture of natural language. The two approaches were proposed simultaneously but independently and can be considered to be a single new theory of type logical grammar that distinguishes itself from existing accounts in that that its type system is undirected. The theory is very simple from a mathematical point of view, unifies treatments of syntax and semantics, immediately solves some empirical problems of standard categorial grammars, invites the definition of a hierarchy of languages not unlike the Chomsky hierarchy for phrase structure grammars, and has a natural relation to modern theories of hyperintensionality. Work on the question how kids can learn language within a semantically situated context is under way.

The ACG/LG framework has generated considerable international interest and a lot of informal exchange that should now be consolidated into a more permanent and formal structure for facilitating research. In this proposal we apply for the funding of a small network with nodes in Haifa (Technion), Nancy (INRIA, Lorraine), Tokyo (National Institute of Informatics) and Tilburg (Logic and Linguistic Analysis, Tilburg University). The aim of the network will be to further research in ACG/LGs and to disseminate the results of this research. We therefore also apply for the funding of scholarly exchange, of four short workshops, of two publications, and of the work needed to write an EU proposal for a larger network dedicated to the initial training of researchers in logical grammar.
Effective start/end date1/01/0731/12/09


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.