Jenkins, H. 1 , Smith, F. 2 , Riches, N. 2 , Petkov, C. 1 & Wilson, B. 1
1 Institute of Neuroscience, Newcastle University, Newcastle-upon-Tyne, UK
2 Education, Communication and Language Sciences, Newcastle University, Newcastle-upon-Tyne, UK
Statistical learning processes are important for language and reading development, and impairments in sequence processing abilities have been identified in language disorders such as dyslexia. Previous research suggests that these deficits may be particularly pronounced in more complex sequence processing tasks. Here, we tested adult dyslexic (n=10) and control (n=18) participants using two different artificial grammars and a suite of language and cognitive tests. The two grammars generated sequences of nonsense words containing either relatively simple adjacent dependencies or more complex nonadjacent dependencies ('xAB' and 'AxB' respectively, where 'A' elements predict 'B' elements and 'x' elements are uninformative). Participants were initially exposed to legal sequences, before making grammaticality judgements about sequences that were either consistent with or violated the dependencies. The control and dyslexic groups performed equally well, and detected violations of both grammars above chance levels as well as generalising the dependencies to novel vocabularies. However, dyslexic participants responded significantly slower, suggesting some differences in sequence processing, independent of sequence complexity. These results raise questions about whether some of the impairments observed in dyslexia are due to specific sequence processing deficits or may be the result of other difficulties, for example in phonological processing of speech sounds.