![]() |
||||
![]() |
CNL Lunch Henk J. Haarmann with Randi C. Martin (Rice U.) Semantic Short
Term Memory Thursday May 8th, 12:30pm, 3416 Marie Mount Hall We examined the role of semantic short-term memory in modulating the scope of conceptual planning during on-line sentence production. Experimental sentences could either have the form (1) complex-NP verb simple-NP or (2) simple-NP verb complex-NP and were matched in length and content. A complex NP expressed two nouns, whereas a simple NP expressed a single noun. A previous picture-description study by Smith and Wheeldon (1999) found that voice onset latencies were longer for sentences that began with a complex noun phrase (complexity effect), indicating incremental planning of the two noun phrases within the sentence. A neuropsychological study by R.C. Martin, Miller, and Vu (in press) further found that a semantic but not phonological short-term memory deficit in persons with aphasia increased the size of the complexity effect, suggesting that semantic short-term memory capacity modulates the scope of conceptual planning during incremental sentence production. To further test this hypothesis, we examined how individual differences in the semantic and phonological short-term memory capacity of normal adults affect the size of the complexity effect. Semantic short-term memory capacity was assessed with the conceptual span test (Haarmann, Davelaar, & Usher, 2003), while phonological short-term memory capacity was assessed with a non-word span task. We and our colleagues previously found that conceptual span but not non-word span predicts on-line meaning integration during sentence comprehension (Haarmann et al., 2003). In a currently ongoing experiment (N= 48 University of Maryland students tested and analyzed so far), we found that conceptual span but not non-word span predicts the size of the complexity effect in on-line sentence production, providing additional evidence that semantic short-term memory capacity modulates the scope of conceptual planning during incremental sentence production. Non-word span did not correlate with conceptual span, but did predict the size of a lexical repetition effect in the same experiment, providing further evidence that semantic and phonological short-term memory are separable systems that make their own contributions to sentence processing. |
|||