Deficits in Explicit Language Problem Solving Rather Than in Implicit Learning in Specific Language Impairment: Evidence From Learning an Artificial Morphological Rule

Purpose The purpose of this study was to delineate differences between children with specific language impairment (SLI), typical age–matched (TAM) children, and typical younger (TY) children in learning and mastering an undisclosed artificial morphological rule (AMR) through exposure and usage. Method Twenty-six participants (eight 10-year-old children with SLI, 8 TAM children, and ten 8-year-old TY children) were trained to master an AMR across multiple training sessions. The AMR required a phonological transformation of verbs depending on a semantic distinction: whether the preceding noun was animate or inanimate. All participants practiced the application of the AMR to repeated and new (generalization) items, via judgment and production tasks. Results The children with SLI derived significantly less benefit from practice than their peers in learning most aspects of the AMR, even exhibiting smaller gains compared to the TY group in some aspects. Children with SLI benefited less than TAM and even TY children from training to judge and produce repeated items of the AMR. Nevertheless, despite a significant disadvantage in baseline performance, the rate at which they mastered the task-specific phonological regularities was as robust as that of their peers. On the other hand, like 8-year-olds, only half of the SLI group succeeded in uncovering the nature of the AMR and, consequently, in generalizing it to new items. Conclusions Children with SLI were able to learn language aspects that rely on implicit, procedural learning, but experienced difficulties in learning aspects that relied on the explicit uncovering of the semantic principle of the AMR. The results suggest that some of the difficulties experienced by children with SLI when learning a complex language regularity cannot be accounted for by a broad, language-related, procedural memory disability. Rather, a deficit—perhaps a developmental delay in the ability to recruit and solve language problems and establish explicit knowledge regarding a language task—can better explain their difficulties in language learning.

Simple morphological spelling rules are not always used: Individual differences in children and adults

ABSTRACTThe English spelling system has a variety of rules and exceptions, but both theoretical and empirical accounts have generally concluded that by about age 9 or 10, children master the morphological rule that regular plural nouns (e.g., socks) and third-person singular present verbs (e.g., lacks) are spelled with the inflectional ending –s. In three experiments, however, we found that when forced to rely exclusively on morphological cues, only a minority of primary school children, secondary school children, and even adults performed significantly above chance at choosing the appropriate spelling for novel words presented as inflected or uninflected nouns and verbs. Further, significantly above-chance performance was more common in adults who had attended school until age 18, compared to age 16. We conclude that many spellers, especially those who do not go on to tertiary education, never learn some simple morphological spelling rules, and instead rely on a store of individual word-specific spellings.

Implicit and explicit instruction

This study examined the influence of implicit and explicit instruction for the acquisition of two types of Dutch spelling rules: a morphological and a phonological rule. A sample of 193 first grade, low- and high skilled spellers was assigned to an implicit-instruction, explicit-instruction, or control-group condition. The results showed that for both rules, students in the explicit condition made more progress than students in the control condition. For the morphological rule, students in the explicit condition had higher posttest scores on pseudo-words than students in the implicit condition. The effects of the three conditions were the same for low- and high-skilled spellers. Both low- and high-skilled spellers in the implicit and explicit condition did not fully generalize their knowledge of both rules to new and pseudo-words.