Language Processing

The focus in this module is on the nature of the mental representations and processes underpinning online language use in experienced language users, although their precursors in the acquisition process are also included in the study. Our primary interests concern reading and spelling.

In the domain of reading, our attention has mainly gone to the basic process that supports all higher-order reading processes, i.e., word recognition, although more recently we have also addressed issues of syntactic parsing in sentence processing. Three topics have dominated our research on word recognition:

  • the role of a word's morphological structure, which is linked with the question whether morphological representations are activated by a postlexical process or by a 'blind' process at
    a prelexical level, whose task it is to identify all ortographic strings that are potential morphemes,
  • the role of prelexical phonological recoding as an automatic, i.e., uncontrollable, process, achieving lexical access through a word's phonological representation,
  • the organization of a bilingual's mental lexicon and the nature of its access processes (language-nonselective or language-selective).

An early precursor of reading achievement is the child's phonological awareness, a topic that has also resulted in a number of studies. In our spelling research we have focused on a peculiar property of the Dutch language, where descriptively quite simple morphographic spelling rules remain a source of spelling errors, even among highly educated people. Studies on the nature of the error patterns during the process of rule learning and at the stage of adequate rule knowledge have enabled us to identify a source of noise that interferes with rule application under conditions of limited working memory resources: the relative occurrence frequency of two homophonic inflectional variants.

Past Projects

The project aims at processing the morphological part of the Goeman - Taeldeman corpus. In June 2005, the first volume of the MAND was published. Topics: pluralisation, diminutivisation and gender.
By analysing the availability of various resources needed for Dutch Language and Speech Technology, priority was assigned to the development of those missing resources that are known to be vital for various types of applications and at the same time can be developed in the current state of the art. The priority lists were submitted to the Dutch Language Union Platform for Language and Speech Technology.