Spectral processing deficits appear to underlie developmental language disorders

More than 50 years of research has suggested something is awry in the auditory functioning of children with language disorders, yet there remains no consensus on exactly what the problem is. To address this gap in knowledge, we have proposed a model termed the Auditory Bases of Phonological Acquisition. This model is based on the language feature known as dual patterning, which describes two levels of structure: one consisting of meaningful words and how they can be combined and the other consisting of word-internal, meaningless units and how they can be combined. We hypothesized that acquisition of this second level—phonological structure—is most affected by deficits in auditory functions, especially those involving the spectral domain. To test this hypothesis, we assessed children with and without language disorders between 7 and 10 years old on their abilities to detect spectral and temporal modulation and to perform relevant language tasks. The largest group differences were found at younger ages for spectral modulation detection (SMD). SMD thresholds were also found to correlate most strongly with language measures, especially those based on phonological structure. Outcomes suggest that keen sensitivity to spectral structure in the speech signal is necessary for acquiring precise phonological representations.