Processing of Partially Free Word Order

One of the unsolved problems in computational linguistics is the adequate and efficient treatment of free word order. Because of an anglocentric viewpoint, most of current software technology for generation and analysis of human language is not suited for handling word order regularities in most of the world's languages. The adequate treatment of word order phenomena is a challenging problem because of the complex interaction of phonetics, syntax, semantics, and pragmatics.

The problem of word order has been investigated in several projects, and linguistic analyses and efficient processing models have been developed.

What is still needed are empirical data about human processing of free word order. In a joint project funded by the Deutsche Forschungsgemeinschaft with psycholinguists from the department of psychology, psycholinguistic experiments were performed to validate the theoretical predictions.