Cochlear implants allow many individuals with profound hearing loss to understand spoken language, even though the impoverished signals provided by these devices poorly preserve acoustic attributes long believed to support recovery of phonetic structure. Consequently, questions may be raised regarding whether traditional psycholinguistic theories rely too heavily on phonetic segments to explain linguistic processing while ignoring potential roles of other forms of acoustic structure. This study tested that possibility. Adults and children (8 years old) performed two tasks: one involving explicit segmentation, phonemic awareness, and one involving a linguistic task thought to operate more efficiently with well-defined phonetic segments, short-term memory. Stimuli were unprocessed (UP) signals, amplitude envelopes (AE) analogous to implant signals, and unprocessed signals in noise (NOI) that provided a degraded signal for comparison. Adults’ results for short-term recall were similar for UP and NOI, but worse for AE stimuli. The phonemic awareness task revealed the opposite pattern across AE and NOI. Children's results for short-term recall showed similar decrements in performance for AE and NOI compared to UP, even though only NOI stimuli showed diminished results for segmentation. Conclusions were that perhaps traditional accounts are too focused on phonetic segments, something implant designers and clinicians need to consider.

This study examined differences between adults and children and between normal and poor readers in the use of phonemic coding strategies for storing words in working memory. In the first experiment, adults, 11-year-olds, and 8-year-olds (categorized as normal or poor readers) recalled eight-item strings of rhyming and nonrhyming words. A developmental decrease in errors was observed for adults, 11-year-olds, and normal-reading 8-year-olds that reflected an improvement in the phonemic coding of items in working memory, but no difference was found between normal- and poor-reading 8-year-olds in the use of phonemic coding strategies. A second experiment with shorter lists and more children supported the latter finding. The results were interpreted as demonstrating that the ability to access syllable-internal phonemic structure is a necessary precursor to the development of phonemic coding strategies for working memory, but that the use of that structure for storing words in working memory is a skill that develops independently and later than the ability to access phonemic structure.