Independence of syntactic and phonological deficits in dyslexia: A study using the attraction error paradigm

Abstract

This paper addresses the question of whether dyslexic children suffer from syntactic deficits that are independent of limitations with phonological processing. We looked at subject‐verb agreement errors after sentence subjects containing a second noun (the attractor) known to be able to attract incorrect agreement (e.g., “the owner(s) of the house(s) is/are away”). In the general population, attraction errors are not straightforwardly dependent on the presence or absence of morphophonological plural markers but on their syntactic configuration. The same would be expected for dyslexic children if their syntactic problems are not phonological in nature. We also looked at the possible effect of system overload on syntactic processing by comparing auditory and written presentation of stimuli and stimuli with high and low frequency attractors. Dyslexic children produced more agreement errors than age‐matched controls, but their errors were distributed in the expected manner and did not align with the presence of morphophonological number markers in the subject overall. Furthermore, there was no effect of either presentation mode or attractor frequency on the number of agreement errors. Our results confirm the existence of syntactic difficulties in dyslexia and suggest that they are not due to a phonological deficit or to verbal working memory limitations.

Do children with developmental dyslexia have syntactic awareness problems once phonological processing and memory are controlled?

“…The Phonological Deficit Hypothesis claims that a phonological processing deficit could strain memory while performing oral syntactic awareness tasks (Smith et al, 1989;Shankweiler et al, 1995). Only a handful of studies have addressed this hypothesis empirically, and most have done so by examining verbal working memory through task demands with mixed results to date (e.g., Shankweiler et al, 1995;Robertson and Joanisse, 2010;Antón-Méndez et al, 2019;Robertson and Gallant, 2019).…”

“…. Syntactic awareness problems in children with dyslexia A large set of studies have now demonstrated that children with dyslexia show poor performance on syntactic awareness tasks in comparison to age-matched controls (e.g., Bentin et al, 1990;Abu-Rabia et al, 2003;Leikin and Assayag-Bouskila, 2004;Rispens et al, 2004;Rispens and Been, 2007;Casalis et al, 2013;Chung et al, 2013;Yeung et al, 2014;Delage and Durrleman, 2018;Antón-Méndez et al, 2019;Robertson and Gallant, 2019; but see Smith et al, 1989;Shankweiler et al, 1995;Robertson and Joanisse, 2010). For instance, poor performance on spoken sentence comprehension tests has emerged in studies of Hebrewspeaking and French-speaking children with dyslexia (Leikin and Assayag-Bouskila, 2004;Casalis et al, 2013, respectively).…”

“…The pattern might differ in dyslexic children learning to read in alphabetic languages, such as English, given strong connection between phonological processing and reading in these languages (e.g., Melby-Lervåg et al, 2012;Snowling and Hulme, 2012). Antón-Méndez et al (2019) examined morphophonological processing in the context of a syntactic awareness task in Spanishspeaking children with dyslexia. Children had to complete a sentence that adhered to subject-verb agreement (e.g., the key to the cabinets___).…”

“…However, if task demands minimize memory loads, no syntax problems should be observed (Smith et al, 1989;Shankweiler et al, 1995). The existing literature which examines this hypothesis comes from studies that examined verbal working memory through task demands (Smith et al, 1989;Shankweiler et al, 1995;Robertson and Joanisse, 2010;Antón-Méndez et al, 2019). Smith et al (1989) implemented a syntactic awareness task that had minimal working memory loads.…”

“…The potential effects of working memory on syntactic awareness in children with dyslexia were examined differently in a study by Antón-Méndez et al (2019). Working memory was manipulated by examining different processing loads in a subjectverb agreement sentence completion task.…”