Phonological Delay or Phonological Impairment in Autism: An Intergroup Comparison

Background:  Studies (Bartak et al., 1975; Cantwell et al., 1978) investigating language, cognition, and behavior in children with Autistic Spectrum Disorder (ASD) and children with Specific Language Impairment (SLI) suggested clear differences between these two groups. In fact, Kjelgaard et al. (2001) showed that among the children with autism there was significant heterogeneity in language skills. Some children with autism had normal language skills and others had language skills significantly below age expectations. The actual debate opposes the view that children with ASD have the same language profile as children with SLI against the view that these children present a delayed acquisition rather than a deviancy as some studies suggested it (Tager-Flusberg, 1981; Tager-Flusberg et al., 1990).

Objectives:  The objective is to determine whether one aspect of atypical language development in ASD (i.e. phonology) is only characterized by a simple delay in development or if it also shows some similarity with structural language impairment found in children with SLI. For this objective, we recruited a group of English-speaking children that were acquiring French as a second language. This group has, indeed, a delayed acquisition of French compared to typical French speaking children.

Methods:  We compared productions of three groups of children at a word repetition task recorded at one year gap (T1 and T2): 28 children with SLI aged 6;5 to 12;11 (mean 9;6), 17 children with ASD aged 6;4 to 12;9 (mean 8;9) and 26 children who have English as first language and French as second language aged 6;4 to 12;7 (mean 9;6). We use the standardized word repetition task of the BILO-3C (Khomsi, 2007). This test includes words of various lengths and different level of phonological complexity.

Results:  Overall performances on the word repetition test show that children with SLI have significantly lower scores compared to children with ASD and L2 children (respectively, mean SD: -7.1; -5.1; -2.6). Correlations between word phonological structures and word productions show that children with pathology have difficulties with structural complexity whereas L2 children are mainly influenced by word length. The analysis of vowel production shows that vowels are not problematic for any of these children. However, consonants, and especially clusters, are a source of difficulty for children with ASD and with SLI. Children with ASD show phonological difficulties not linked with their prosodic impairment. Patterns of errors are similar in children with pathology with a preference in using substitution and elision of phonemes. L2 children show a clear preference for elision. They also tend to omit entire syllables instead of isolated phonemes, possibly due to the different stress system found in these two languages.

Conclusions:  To conclude, we can say that L2 children are not really affected by structural complexity but rather by a mismatch between French and English prosody. Children with pathology are influenced by complexity rather than prosody. Differences with children with ASD and SLI are mainly seen in the amount of errors, but not in patterns of errors.