| 研究概要 |
It is well established that long-term phonological knowledge contributes to short-term retention of verbal sequences. Recent studies indicate that lexical prosody also affects phonological short-term memory performance. Prosodic features of languages are founded on temporal structures of verbal items (e.g., word length) and should be implemented with such temporal structures. We examined whether temporal factors influence usage of long-term phonotactic knowledge in two sets of immediate serial recall experiments. The first set compared recall performance for two types of nonword lists, sharing the same bi-mora frequency as lists but consisting of a mixture of high- and low- frequency bi-morae (a mora is a Japanese subsyllabic unit). In the first type of list, temporal gaps were inserted at low-frequency bi-mora positions, creating temporal grouping. In the second type, temporal gaps were inserted at high-frequency bi-mora positions, for which we expected a disconnection of the two morae, which otherwise possessed a strong association. The latter manipulation reduced the effect of bi-mora frequency on recall performance. The second set of experiments manipulated position-specific and position-free bi-mora frequency - both affected recall performance. The results indicated the presence of two types of long-term phonotactic knowledge implementation, one being more and the other being less sensitive to temporal structure.
|
