Use this URL to cite or link to this record in EThOS:
Title: The use of linguistic context by children with autism
Author: Bell, S. J.
Awarding Body: University of Cambridge
Current Institution: University of Cambridge
Date of Award: 2004
Availability of Full Text:
Full text unavailable from EThOS.
Please contact the current institution’s library for further details.
Children with autism show an enhanced ability to disembed perceptual components from within meaningful surrounds (Shah and Frith, 1983; Happé, 1966). This is sometimes interpreted as being indicative of a reduced tendency to integrate information into a high-level, global representation (Frith, 1989; Happé, 1999b), and demonstrations of this have been shown in visual as well as linguistic tasks (Scheuffgen, 1998; Jolliffe and Baron-Cohen, 1999, 2000; Happé, in preparation). More recently, it has been suggested that children with autism are able to build high-level contextual representations, but show a bias to form local interpretations (Scott et al., submitted; Plaisted et al., 1999). In this dissertation, three novel studies are described, which probe children’s use of contextual information using techniques more commonly associated with other psychological fields of enquiry. I start by considering embedded words—words within words such as ‘cap’ in ‘captain’. By using a modified signal detection task, in which participants signalled the presence or absence of a target word, I found that children with autism did not find these embedded words any more salient than is typical. However, a general reduction in perceptual sensitivity was observed, which could lead to an overall difficulty of language comprehension. The second experiment used a word monitoring task to examine the use of semantic and syntactic context when recognising words in continuous speech. Whilst children with autism showed a similar pattern of benefit from semantic and syntactic context to their developmentally-typical peers, they exhibited an overall attenuation in this benefit, possibly explained by impaired information flow in an interactive speech recognition network. A manipulation of speech rate on this experiment was also explored. Finally, I designed a linguistic analogue to the visual embedded figures task, in which a target occurred in a meaningful sentence, but where the global interpretation did not assist its detection. Whilst a trend for children with autism to show superior performance on this task was not significant, a comparison to a variant where the global interpretation facilitated target detection affirmed that the children with autism made reduced use of linguistic context.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available