Conclusions and Limitations
Previous literature suggests that children with LLI are less likely to report responses indicating the integration of auditory and visual speech signals in illusory contexts. Despite this, the current study demonstrates that this population can use visual speech cues to support auditory
perception given a relatively more naturalistic paradigm. It is suggested that in the context of informational masking, visual cues may assist the segregation and preferential processing of the target auditory speech signal. It should be noted, though, that the children with LLI also showed atypical development of speechreading skills, indicating that although beneficial, access to and/or use of visual cues is not equivalent across groups. This raises the possibility that improving speechreading skills could assist children with LLI in their speech perception and language comprehension abilities, either by strengthening amodal phonological representations or by encouraging more robust support for auditory processing. Given that the youngest children in the current study were also able to benefit from visual speech cues, this study emphasizes the importance of better understanding visual speech processing in relation to the development of speech perception and language comprehension over the preschool and early school years. Over this period, children’s vocabularies are developing rapidly, and the typical listening conditions change dramatically as children enter the classroom where they are expected to listen to a single speaker in a background of other voices. The primary limitation of the current data set is the cross-sectional nature of the sample, meaning that the exacerbation of the speechreading deficit over age in the LLI group may be an effect of individual differences (arising from different selection biases across age, for example) rather than developmental change. This is an important consideration in the case of language disorders as profiles vary substantially over developmental time in the LLI population (e.g., ContiRamsden & Botting, 1999; Conti-Ramsden, Botting, Simkin, & Knox, 2001), such that children recruited at different ages could show different profiles to children followed to the same age over time longitudinally. In terms of the tasks adopted here, the requirement to respond to a single target word embedded in a predictable carrier sentence during the SpiN task may have resulted in falsely low SRTs for some of the older TD children who may have been able to adequately speechread targets at low SNRs. The excellent performance of a few members of the TD group point to this task limitation. Future work charting the benefit and limitations of visual speech cues for TD children would be a valuable endeavor to promote better understanding of when and how visual cues support both everyday speech perception in childhood and ongoing changes in perceptual abilities.