The construct of rapid rate auditory processing as it applies to specifically language impaired (SLI) children's on-line sublexical and lexical processing has not been adequately framed within current models of language processing. Accordingly, at the sublexical level, we test the hypothesis that SLI children should have particular difficulty perceiving verb inflections regarded as low in """"""""phonetic substance"""""""" because of their brief-duration characteristics. At the lexical level, we test the hypothesis that rapid rate processing should have a specific and local influence on the lexical mapping (i.e. acoustic-phonetic processing) phase of word recognition. We also examine the effects of variations in sentence-level phonetic context on SLI children's on-line word recognition to evaluate any (a) potential independent contribution of lexical retrieval difficulties and/or (b) interaction of lexical mapping and retrieval difficulties. In years 1-3 subjects, all with normal-range nonverbal IQ, will include 24 SLI, 24 age-matched. Normal children (LN-A), and 24 comprehension- matched normal children (LN-C). In experiments 1,2, and 4 subjects will [participate in a word recognition reaction time (RT) task in which they will listen for familiar nouns occurring in fixed positions in sentences controlled for vocabulary, length, and complexity. In experiment 1 (Morpheme Processing), nouns will follow verbs (low-phonetic or high phonetic) that have been properly inflected (INF) or improperly uninflected (STEM). In experiment 2 (Isolated Word Recognition), subjects will listen to word list for the repeated occurrence of a target word (target starting/ending with a stop or nonstop). In Experiment 4 (Sentence Processing), subjects will listen for words in sentences which vary in phonetic processing demands. Word recognition RT and number of errors (misses, false alarms) are the dependent variables in these experiments. In Experiment 3 (gating task) subjects will be presented successive 50ms chunks of monosyllabic words beginning at word onset until word offset and asked to identify the words. In years 4-5, subjects will participate in two experiments in which they will listen to prosodically and acoustically altered sentences to determine whether such alterations influence their on-line inflectional and lexical processing. Between and within subjects ANOVAs will be performed to test the primary hypotheses that SLI subjects (a) have greater trouble processing low- phonetic inflections than higher-phonetic ones (Exper 1) (b) do not have greater difficulty recognizing stop words than nonstop words in isolation (Exp 2,3) or in sentences (Exp 4, depending on the phonetic processing demands). Correlational (and when appropriate regression analyses with other variables [Exper 1] will be run to determine the relationship between rapid rate processing and morpheme/lexical processing. Similar analyses will be used to test the effects of prosodic/acoustic alterations on SLI and at least LN-C children's processing of inflections and words in Experiments 5 and 6.
