Bilingual Children’s Language Development Differs by Disorder and Language

A new longitudinal study published in Frontiers in Education reveals that the development of complex syntax in Spanish–English bilingual children follows distinct trajectories depending on whether the child has a developmental language disorder (DLD) and which language is being used. The research, conducted by a multi-university team, tracked 54 children from prekindergarten through third grade to understand how grammatical complexity and subordination evolve in both English and Spanish. The findings indicate that while both groups of children showed overall growth in syntactic complexity, the heritage language, Spanish, exhibited greater vulnerability to developmental delays in children with DLD compared to English.

The study was designed to address a significant gap in the literature regarding the longitudinal development of complex syntax in bilingual children. Previous research has often relied on cross-sectional data or focused on a single language, leaving unclear how the two linguistic systems interact over time in children with and without language disorders. By examining narrative retellings in both languages at three distinct time points, the researchers aimed to provide a more nuanced understanding of typical and atypical language development in this population, which represents one of the largest and fastest-growing groups of children in the United States.

The participants were drawn from a larger longitudinal study involving Spanish–English bilingual children in the greater Phoenix metropolitan area. The final analytic sample consisted of 39 children with typically developing language (TDL) and 15 children with DLD. All participants were Latino and attended English-only schools, with Spanish serving as their primary home language. The children completed narrative retellings of wordless picture books in both English and Spanish during prekindergarten, first grade, and third grade. These narratives were transcribed and coded using the Systematic Analysis of Language Samples (SALT) software to measure various aspects of syntactic development.

The researchers employed several key metrics to assess syntactic complexity and grammatical accuracy. The Subordination Index (SI) was used to measure finite clause density, representing the average number of clauses per utterance. Ungrammaticality (UG) was calculated to capture morphosyntactic errors, such as incorrect verb tense or agreement. Additionally, the Percentage of Complex Sentences (PCS) was determined to measure how frequently children produced utterances containing at least one subordinate clause. Subordinate clauses were further categorized into nominal, adverbial, and relative types to examine the development of specific syntactic structures.

In English, the results showed broadly similar growth patterns for both the TDL and DLD groups. There were no significant group differences in the Subordination Index, overall subordinate clause use, or the production of specific clause types. Both groups demonstrated significant increases in syntactic complexity from prekindergarten to third grade. This suggests that sustained exposure to English through formal schooling may support the development of complex syntax in bilingual children, even those with DLD, allowing them to produce a range of subordinate structures at rates comparable to their typically developing peers.

In contrast, the patterns in Spanish, the heritage language, were more pronounced and revealed significant group differences. Children with TDL showed significant gains in overall subordination and the use of nominal clauses across the grades. However, children with DLD showed limited growth after first grade, with their syntactic complexity plateauing. By third grade, the TDL group had significantly higher Subordination Index values and produced a significantly higher percentage of complex sentences than the DLD group. This indicates that Spanish complex syntax may be particularly vulnerable in children with DLD, possibly due to reduced opportunities for complex language use in the heritage language as they progress through English-only schooling.

The analysis of subordinate clause types further highlighted these cross-language differences. In English, both groups showed robust growth in nominal and adverbial clauses, with no significant differences between the TDL and DLD groups. Relative clauses, which are generally more difficult to acquire, increased modestly across grades but did not differ significantly between the groups. In Spanish, however, nominal clauses distinguished the two groups in prekindergarten and again in third grade. The TDL children showed steady growth in nominal clause use, while the DLD children plateaued after first grade. Adverbial and relative clauses in Spanish showed no significant developmental change or group differences, although descriptive analysis suggested that TDL children used a wider variety of connectors and constructions.

The study also examined grammatical accuracy using the ungrammaticality measure. In English, the TDL group demonstrated a substantial decrease in ungrammaticality by third grade, while the DLD group showed a descriptive reduction that was not statistically significant. In Spanish, the DLD group maintained consistently high levels of ungrammaticality across all grades, with no significant improvement over time. This persistent difficulty with grammatical accuracy in Spanish suggests that morphosyntactic vulnerabilities associated with DLD are more pronounced in the heritage language, where children may have less exposure to complex and grammatical language use.

The authors suggest that the observed patterns may be influenced by the differential exposure to the two languages. Formal education occurred in English, providing children with sustained opportunities to encounter and produce complex syntactic structures. In contrast, Spanish received little or no academic support, potentially limiting the development of complex syntax in the heritage language. However, because language exposure and use were not measured longitudinally in the study, the specific contribution of differential input to these developmental patterns cannot be determined directly. The findings highlight the importance of examining complex syntax in both languages when assessing bilingual children with DLD.

The study has several limitations that should be considered when interpreting the results. The use of narrative retellings may not elicit the full range of subordinate clauses, particularly adverbial and relative clauses, which are more common in expository and conversational contexts. The DLD group was relatively small, which limits statistical power to detect subtle differences, especially in interaction effects. Additionally, the study did not quantify language exposure or dominance over time, making it difficult to disentangle the role of input from developmental constraints. Future research should incorporate detailed longitudinal measures of language exposure and proficiency to better understand how shifts in input affect the trajectories of complex syntax in heritage languages.

This research provides valuable longitudinal evidence of distinct syntactic trajectories across the two languages in bilingual children with and without DLD. It underscores the need for a nuanced approach to assessing language development in bilingual populations, recognizing that the heritage language may present unique challenges for children with language disorders. The findings have implications for clinical practice, suggesting that interventions may need to target specific syntactic structures in the heritage language to support the development of complex syntax in bilingual children with DLD.

The study contributes to the growing body of literature on bilingual language development and the impact of developmental language disorders. By examining both languages over time, the researchers were able to capture the dynamic nature of syntactic development in bilingual children. The results suggest that while English schooling may support the development of complex syntax in both groups, the heritage language requires further attention to ensure that children with DLD do not fall behind in their syntactic development. This highlights the importance of providing opportunities for complex language use in both languages to support the overall language development of bilingual children.

The study utilized a linear mixed-model framework to account for the repeated measures collected from each child across the three time points. By treating participants as random effects, the statistical approach allowed the researchers to model individual developmental trajectories while retaining data from children who may have had incomplete records. This method is particularly suited for longitudinal designs with small sample sizes, as it maximizes statistical power by leveraging the dependency among observations from the same individual. The models included fixed effects for grade level and group status, enabling the researchers to isolate the specific contributions of age-related development and language disorder status to syntactic growth in each language.

Coding reliability for the subordinate clause analyses was established through independent evaluation by two trained raters for every transcript. The interrater agreement was strong across all years and languages, with pooled Cohen’s kappa values ranging from .911 to .998. For specific clause types, agreement remained substantial to near-perfect, with nominal clauses ranging from .72 to 1.00, adverbial clauses from .92 to 1.00, and relative clauses from .75 to 1.00. This high level of consistency ensures that the observed developmental patterns and group differences are not artifacts of coding variability, providing a robust foundation for the subsequent statistical comparisons of syntactic complexity and clause-type diversity.

The classification of children into the DLD and TDL groups relied on a combination of standardized language assessments and educational documentation. To be classified as having DLD, a child had to meet at least two of three specific criteria: scoring below designated thresholds on the Word Structure and Recalling Sentences subtests of the Clinical Evaluation of Language Fundamentals–Preschool, Spanish edition, or scoring below a cutoff on the Spanish Screener for Language Impairment in Children. Additionally, having an Individualized Education Plan documenting speech-language services served as a qualifying factor. Thirteen of the fifteen children in the DLD group had an IEP, while the remaining two were retained in the group based on their standardized test scores, ensuring a rigorous and multi-faceted approach to group identification.

The narrative tasks used to elicit language samples were carefully structured to ensure consistency across languages and time points. Children retold stories from two wordless picture books, Frog on His Own and A Boy, a Dog, a Frog, and a Friend. The versions of these books were counterbalanced across languages and years to prevent any single story from being presented consecutively in the same language. This procedural control helps mitigate potential practice effects or story-specific influences on the children’s syntactic production. The narratives were segmented into Terminable Units, a standard unit of analysis in discourse research, which allowed for the precise calculation of measures such as the Subordination Index and the Percentage of Complex Sentences.

While the primary analysis focused on finite clauses, the researchers also accounted for non-finite structures such as infinitival and gerundive clauses. These structures were coded as subordinate because they represent independent propositions that contribute to overall syntactic complexity. For instance, an infinitive clause expressing purpose or a gerundive clause functioning as a noun was included in the respective functional categories. This inclusive approach to defining subordination provides a more comprehensive picture of a child’s ability to embed clauses, capturing a wider range of linguistic strategies than measures that rely solely on finite clause density. It allows for a more nuanced assessment of how children with DLD manage the morphosyntactic demands of complex sentence construction in both English and Spanish.

Subject of Research: Longitudinal development of complex syntax and subordination in Spanish–English bilingual children with and without developmental language disorder

Article Title: Trajectories of syntactic complexity and subordination across Spanish and English in bilingual children with and without developmental language disorder

Article References: Martinez-Nieto, L., Ames, K., Restrepo, M. A., Mesa, C., Pierce-Rivera, M., Moen, T., & Agbo, L. (2026). Trajectories of syntactic complexity and subordination across Spanish and English in bilingual children with and without developmental language disorder. Frontiers in Education, 11, Article 1940578. https://doi.org/10.3389/feduc.2026.1940578

Image Credits: AI Generated

DOI: 10.3389/feduc.2026.1940578

Keywords: bilingual development, developmental language disorder, syntactic complexity, subordination, heritage language, longitudinal study, Spanish–English bilingualism, grammatical accuracy, narrative retell, clause types, Trajectories, syntactic