Article Omission in Dutch Children with SLI: A Processing Approach
- Resource Type
- Authors
- Sergey Avrutin; Lizet van Ewijk
- Source
- Entropy, Vol 12, Iss 4, Pp 798-817 (2010)
Entropy
Volume 12
Issue 4
Pages 798-817
- Subject
- Dependency (UML)
SLI
Kullback-Leibler divergence
Computer science
General Physics and Astronomy
lcsh:Astrophysics
Specific language impairment
computer.software_genre
Channel capacity
Typically developing
lcsh:QB460-466
medicine
Set (psychology)
lcsh:Science
Divergence (linguistics)
business.industry
medicine.disease
lcsh:QC1-999
articles
processing
lcsh:Q
Artificial intelligence
business
computer
Natural language processing
lcsh:Physics
Cognitive psychology
- Language
- English
- ISSN
- 1099-4300
Children with Specific Language Impairment (SLI) show difficulties with grammatical morphology. Based on the data from 12 Dutch children with SLI, an information-theoretical model is proposed in which the noun-article set dependency is modeled as a channel. We propose that reduced capacity of this channel is responsible for article omission. The Kullback-Leibler divergence between input and output distribution of article production provides an index of the channel capacity, which is shown to correlate with the percentage of article omission and to lag behind in SLI development as compared to typically developing children.