The impact of phonological complexity on past tense inflection in children with Grammatical-SLI

TLDR

English‑speaking children with Specific Language Impairment variably produce inflected and bare stem forms in past‑tense contexts, and while the deficit is debated as morphosyntactic or morphophonological, the Computational Grammatical Complexity hypothesis predicts that deficits in syntax, morphology, and phonology jointly impair regular past‑tense formation, a claim previously tested for syntactic and morphological complexity. This study extends that work by examining how phonological complexity influences regular past‑tense inflection in children with Grammatical‑SLI. Using a past‑tense elicitation task that varied the phonological complexity of the inflected verb end, we found that G‑SLI children were less likely to add the suffix when the inflected form ended in a consonant cluster. In contrast, typically developing controls showed no such effect, underscoring the independent contributions of language components to impaired and normal performance.

Abstract

English-speaking children with Specific Language Impairment (SLI) variably produce inflected and bare stem forms in obligatory past tense contexts. Researchers have not reached consensus as to whether the underlying deficit is morphosyntactic or morphophonological in nature. The Computational Grammatical Complexity (CGC) Hypothesis takes a different tack: it hypothesizes that for children with a particular form of SLI, Grammatical-SLI, the deficit is in representing linguistic structural complexity in at least three components of the computational grammatical system – syntax, morphology and phonology. Deficits in all these components are predicted to impact on regular past tense formation. The impact of syntactic and morphological complexity on G-SLI children's realization of tense has been tested previously. Here we complete the picture by considering phonological effects on their production of regular past tense inflection. Using a past tense elicitation task where we manipulate the phonological complexity of the inflected verb end, we show that, as predicted, verb-end phonological complexity impacts on suffixation: G-SLI children are less likely to suffix stems when the inflected form ends in a consonant cluster. Typically developing controls show no such effect. The results of this study highlight the need to consider the independent contributions of language components to impaired and normal performance.