Sandra Muckel1, 2 & Thomas Pechmann2
1 Saarland University & 2 University of Leipzig
The constraints that prosody imposes on sentence structure building are subject to recent parsing research. Most studies deal with prosodic disambiguation effects on the processing of ambiguous syntactic structures. Besides, the role of prosody in predictive parsing is discussed. There is some evidence that prosody may help to identify the base position (the gap or trace) of a fronted constituent (Nagel, Shapiro, & Nawy, 1994). It has been suggested that prosodic information gives indirect rather than direct hints to the location of gaps (Straub, Wilson, & Badecker, 2001).
In a series of cross-modal lexical priming experiments, we explored prosodic effects on gap-filling in German verb-final sentences. Crucially, verb-specific subcategorization information that might license the gap is not available in a preverbal position. We claim that German intonation structure gives a hint to the location of a preverbal object gap. The neutral sentence accent typically falls on the constituent immediately preceding the sentence-final verb (Féry, 1993). This regularity might help to predict the sentence-final verb position. As the verb is adjacent to the direct object's base position, the parser might exploit the predictability of the verb position for the identification of the preverbal gap. To test for this possibility, we presented (1) sentences with neutral accentuation (NA) and (2) sentences with contrastive accentuation of the sentence-final verb (VA).
| (1) | Der Krug(i) ist [einem jungen Richter des Berliner (c) GeRICHTS (t(i)) zerbrochen] | NA |
| (2) | Der Krug(i) ist einem jungen Richter des Berliner (c) Gerichts (t(i)) [zerBROchen] | VA |
| the+NOM jug is a+DAT young judge of the Berlin court broken | ||
| 'The jug broke on a young judge of the Berlin court.' |
In Experiment 1, (1) yielded shorter lexical decision times for identical targets in the gap position (t(i)) as compared to a control position (c). No difference was found for unrelated targets. In Experiment 2, the reactivation effect found for identical targets in (1) was replicated. By contrast, no similar effect for identical targets was found in (2). In Experiment 3, there was no decision time difference for identical or unrelated targets with respect to probe positions in (2). We conclude that prosodic information is used to license a direct object gap in German neutrally focused verb-final sentences.
References
Féry, Caroline (1993). German intonational patterns. Tübingen: Niemeyer.
Nagel, H. Nicholas, Shapiro, Lewis P., & Nawy, Rebecca (1994). Prosody and the processing of filler-gap sentences. Journal of Psycholinguistic Research, 23, 473 - 485.
Straub, Kathleen, Wilson, Colin, McCollum, Courtney, & Badecker, William (2001). Prosodic structure and wh-questions. Journal of Psycholinguistic Research, 30, 379 - 394.