Autistic Children in China Show Surprising Strength in Reading Chinese Characters

Young children with autism spectrum disorder in China may struggle with many early language skills, yet they appear to keep pace with their typically developing peers when it comes to recognizing Chinese characters and rapidly naming numbers. That is the central finding of a new study published in the Journal of Autism and Developmental Disorders, which followed 51 autistic kindergarteners and 50 typically developing children aged four to six in a coastal Chinese city. The research offers one of the first detailed portraits of how emergent literacy unfolds in a logographic language, a writing system whose demands differ sharply from the alphabetic scripts that dominate most autism research.

Emergent literacy refers to the foundational knowledge and skills that precede formal reading and writing instruction, including phonological awareness, morphological awareness, orthographic awareness, and oral vocabulary. Researchers often divide these precursors into code-related skills, such as alphabet knowledge and print concepts, and meaning-related skills, such as vocabulary and listening comprehension. Because these early abilities reliably predict later reading fluency and comprehension, identifying where autistic children diverge from their peers has become a pressing question for educators and clinicians alike.

Most previous studies of emergent literacy in autism have been conducted in English-speaking or other alphabetic-language contexts. Chinese presents a fundamentally different challenge for young readers. Unlike alphabetic letters, which map onto phonemes, each Chinese character represents a tonal syllable and a unit of meaning. Characters are visually complex, built from stroke patterns, and many are semantic-phonetic compounds containing a radical that hints at meaning and another that suggests pronunciation. Learning to read Chinese therefore places heavy demands on visual memory and on the ability to perceive and process morphological structure.

The research team, led by Zhuo Chen of Shaanxi Normal University together with colleagues at Texas A&M University, recruited children from four inclusive kindergartens serving working-class families. The autistic and typically developing groups were closely matched in age and nonverbal reasoning as measured by Raven’s Colored Progressive Matrices. Autism diagnoses were confirmed with the Autism Diagnostic Observation Schedule, and all participating autistic children could produce short sentences, understand oral instructions, and use spoken language for basic communication. None of the kindergartens provided formal literacy instruction, so the skills measured reflected spontaneous early development rather than classroom teaching.

Each child completed a battery of assessments covering working memory, inhibitory control, rapid automatized naming, phonological awareness through a syllable deletion task, morphological awareness through a compounding construction task, orthographic awareness through tasks distinguishing real characters from pseudo-characters and legal stroke patterns from illegal ones, character recognition, and receptive and expressive vocabulary. The researchers then applied independent-sample t-tests with Bonferroni correction and hierarchical regression models to compare groups and identify predictors of character recognition and vocabulary knowledge.

The results revealed a strikingly uneven profile. Children with autism performed significantly worse than their peers on phonological awareness, morphological awareness, orthographic awareness, and both receptive and expressive vocabulary. They also showed weaker working memory and inhibitory control. Yet on character recognition and rapid automatized naming, the two groups were statistically indistinguishable. The authors suggest that a detail-focused cognitive processing style, long documented in autism, may help these children extract the visual patterns of Chinese characters with unusual precision, laying a foundation for successful decoding despite broader language difficulties.

Regression analyses showed that autism severity, nonverbal reasoning, inhibitory control, rapid naming, phonological awareness, and orthographic awareness together predicted character recognition among the autistic children, with the full model explaining roughly 87 percent of the variance. For typically developing children, working memory and morphological awareness emerged as additional unique predictors. Notably, orthographic awareness, the understanding of legal stroke patterns and positional rules governing how characters are assembled, contributed significantly to character reading in both groups, underscoring the importance of visual-orthographic discrimination in Chinese literacy acquisition.

The predictors of oral vocabulary told a different story. For autistic children, autism severity, nonverbal reasoning, rapid naming, and morphological awareness significantly predicted receptive vocabulary, while expressive vocabulary depended heavily on receptive vocabulary itself. For typically developing children, morphological awareness predicted both receptive and expressive vocabulary. The authors interpret this asymmetry cautiously: morphological awareness appears strongly tied to word meaning across both groups, but expressive language development in autism likely involves additional factors beyond emergent literacy skills, and the field’s findings on receptive-expressive gaps in autism have been mixed.

Why do autistic Chinese children lag specifically in phonological awareness? The study’s syllable deletion task required children to manipulate sound units mentally, a process demanding complex cognitive and linguistic control. Earlier work suggests autistic children fare better on simpler phoneme isolation tasks than on blending and deletion, and difficulties with social interaction and language comprehension may compound the challenge in Chinese, where syllable awareness develops largely through language exposure. Similarly, delays in orthographic awareness may reflect differences in visual processing and limited exposure to written Chinese during the preschool years.

The findings carry practical weight for educators. Rather than treating autistic children solely through the lens of deficit, the authors recommend building on their apparent strength in character recognition to expand vocabulary knowledge, using naturalistic teaching, peer activities, and digital tools that make language visually concrete. Because the study was cross-sectional and involved a relatively small sample of verbally fluent children in one region, the authors caution that causal conclusions cannot yet be drawn, and they call for longitudinal research including minimally verbal autistic children. Still, the work marks an important step toward literacy interventions tailored to autistic learners in non-alphabetic languages, where the road to reading is charted by strokes and radicals rather than letters and sounds.

Subject of Research: Emergent literacy development and its contribution to character recognition and oral vocabulary in Chinese-speaking kindergarteners with autism spectrum disorder.

Article Title: The Contribution of Emergent Literacy Skills to the Development of Character Recognition and Oral Vocabulary Knowledge Among Children With Autism Spectrum Disorder: The Case of Kindergarteners in China

Article References: Chen, Z., Kuo, L.-J., Dixon, L. Q., Han, C.-L., & Ma, Y. (2026). The Contribution of Emergent Literacy Skills to the Development of Character Recognition and Oral Vocabulary Knowledge Among Children With Autism Spectrum Disorder: The Case of Kindergarteners in China. Journal of Autism and Developmental Disorders. https://doi.org/10.1007/s10803-026-07514-x

Image Credits: AI Generated

DOI: 10.1007/s10803-026-07514-x

Keywords: autism spectrum disorder, emergent literacy, character recognition, oral vocabulary, phonological awareness, morphological awareness, orthographic awareness, Chinese language, kindergarten, rapid automatized naming, inhibitory control, working memory