index by author
index by subject
| |
home index by author index by subject |
Friday stream 1-4 Session 16.10 - 17.25 Length 25 minutes
Malin Ericson
Department of Linguistics Stockholm University, Sweden malin@ling.su.se
Abstract
Using a computerised phoneme monitoring technique, dyslexic subjects and controls are asked to respond to the presence of three phonemes - /j/, /ng/ and /g/ - in a number of spoken sentences in Swedish. Correctness of response and RT are registered. As it is possible that orthography may indirectly influence subjects' judgements - despite the fact that the subjects not are shown the text - the sentences vary with regard to the relationship between phonemes and potential orthographic representations. For example can /j/ be represented as <j,> <g> and a number of digraphs, while the grapheme <g> can represent /g/, /k/, /j/ and /nothing/. Some of the sentences also include graphemes, which, from the reading training point of view, normally not are considered as representations of the phoneme in question. It is assumed that dyslexics and controls will differ in performance, an assumption based on two related hypotheses: 1) Orthographical knowledge interferes with phonological perception even though the stimuli are presented in an auditory mode only. 2) Interference from orthography is weaker for dyslexics than for normal readers. Specifically, normal readers as compared to dyslexics will have longer RT's when detecting instances of a phoneme not represented by its primary grapheme.
 |
 |