Premium
Sentence repetition in children with specific language impairment: an investigation of underlying mechanisms
Author(s) -
Riches Nick G.
Publication year - 2012
Publication title -
international journal of language and communication disorders
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.101
H-Index - 67
eISSN - 1460-6984
pISSN - 1368-2822
DOI - 10.1111/j.1460-6984.2012.00158.x
Subject(s) - specific language impairment , psychology , short term memory , repetition priming , sentence , working memory , phonology , cognitive psychology , repetition (rhetorical device) , cognition , language production , priming (agriculture) , developmental psychology , task (project management) , audiology , linguistics , lexical decision task , neuroscience , medicine , philosophy , botany , germination , management , biology , economics
Background: Sentence repetition (SR) is a reliable clinical marker of specific language impairment (SLI). However, little is known about cognitive processes underpinning SR, or areas of breakdown in children with SLI. Aims: The study investigated which cognitive mechanisms were most closely involved in SR performance: syntactic knowledge, phonological short‐term memory (STM) and working memory (WM). Methods & Procedures: Twenty‐three children with SLI (mean age = 6;7), 18 age‐matched (mean age = 6;5) and 21 language‐matched children (mean age = 4;8) repeated 180 sentences of varying length and complexity. Total words omitted, added or substituted were counted. Assessments of syntactic knowledge using a structural priming task, and assessments of WM, and phonological STM, were conducted. Twenty sentences were presented in a delayed repetition condition to investigate the role of phonological STM. Outcomes & Results: The children with SLI made more SR errors than controls and found delayed repetition especially difficult. Their SR errors were qualitatively similar to errors on other production tasks. In the SLI group, all assessments were good predictors, though the priming task was the strongest. In the control groups WM tasks were the best predictors, but the phonological STM task was a poor predictor. Conclusions & Implications: The data support a multifaceted view of SR with a role for syntactic knowledge, WM and STM. ‘Redintegration’, whereby long‐term memory representations are used to maintain information in short‐term memory, is likely to be a key process. Specific errors in the SLI group are likely to reflect difficulties with underlying syntactic representations. Children with SLI may be more dependent on phonological STM than typically developing children.