Longitudinal Changes in Pronoun Reversal in Children with ASD and TD Children

Background: Pronoun reversal (e.g., using “you” for “I”) is a striking characteristic of ASD language (Fay, 1979). However, the actual incidence of reversals is still unknown. Previous research has reported 15% reversals in 5-7-year-old children with ASD (Tager-Flusberg, 1994), and around 85% reversals in a two-year-old child with Asperger’s syndrome (Evans & Demuth, in press). It is also unclear what proportion of ASD and TD children reverse.

Objectives: To assess these issues we examined spontaneous pronoun use in children with ASD and TD controls in a longitudinal study of language development. 

Methods: Children (n=10 ASD, 18 TD) were visited every four months for two years. At visit 1, ASD children had a mean age of 33 months, had begun intensive ABA therapy, had mean Mullen Visual Reception T-scores of 45.6, and produced 159.7 words on the CDI-1. TD controls had a mean age of 20 months, mean Mullen VR T-scores of 53, and produced 118.8 words on the CDI-1.

At each visit, children engaged in 30-minute semi-structured play sessions with their parents. The children’s spontaneous speech was coded for the occurrence of 1^st and 2^nd person pronouns, and for whether each pronoun was correct or reversed. Data from visits 1, 3, and 5 are reported here. 

Results: Pronoun use increased with age for both groups, the ASD group producing an average of 12.8 (SD=14.15) at visit 1, 36.5 (SD=36.12) at visit 3, and 46.55 (SD=37.97), and the TD group producing an average of 3.22 (SD=6.84) at visit 1, 37.61 (SD=32.02) at visit 3, and 55.88 (SD=24.31) at visit 5 (group differences ns). Both groups produced many more pronouns referring to self than other (ranges 63%-77%) at all visits (group differences ns). Percent of pronoun uses that were reversals also varied across visits; the ASD group produced 14.12% reversals at visit 1, 3.14% reversals at visit 3, and 8.56% reversals at visit 5, and the TD group produced 2.5% reversals at visit 1, 2.25% reversals at visit 3, and 1.2% reversals at visit 5. The ASD group produced a marginally higher percentage of pronoun reversals only at visit 5 (t(9) = 2.17, p =.057). Developmental changes were seen in both groups concerning which pronouns were reversed more: Children with ASD used “I” for “you” more at visit 1, but “you” for “I” more at visits 3 and 5 (X² = 5.47, p = .065); TD children’ used “you” for “I” more at visit 1, but “I” for “you” more at visits 3 and 5 (X² = 6.83, p = .033). The groups differed in their patterns of reversals at visits 1 (X² = 5.6, p = .018) and 3 (X² = 5.61, p = .018).

Conclusions: Both TD and ASD children increased pronoun use with age, and both groups produced pronoun reversals. Although the incidence of reversals was lower than previously reported, the ASD children produced a higher proportion of pronouns as reversed, and more frequently used “you” to refer to themselves. Pronoun reversals may be both contextually- and developmentally-sensitive.