Trace identification in German: (When) Is it real?

Sandra Muckel¹ and Thomas Pechmann²
muckel@cns.mpg.de, pechmann@rz.uni-leipzig.de
1 Max Planck Institute of Cognitive Neuroscience, Leipzig, Germany
2 University of Leipzig, Germany

In two cross-modal lexical priming experiments (Muckel & Pechmann, submitted), we tested for antecedent reactivation effects in ergative sentences which are assumed to be base-generated with the constituent marked for Dative preceding that marked for Nominative (Grewendorf, 1989). Targets for lexical decision were presented (1) in a preverbal trace position (t(i)) and (2) in a preverbal non-trace position (*t(j)) as well as in a control position (c).

(1)

Der Krug(i) ist einem jungen Richter des Berliner (c) Gerichts (t(i)) zerbrochen.
the+NOM jug is a+DAT young judge of the Berlin court broken
'The jug broke on a young judge of the Berlin court.'

(2)

Dem Richter(j) ist (t(j)) ein knallbunter Krug aus empfindlichem (c) Steingut (*t(j)) zerbrochen.
the+DAT judge is a+NOM colourful jug of fragile earthenware broken
'A colourful jug made of fragile earthenware broke on the judge.'

(1) yielded shorter lexical decision times for identical targets in the critical position as compared to the control position. No difference was found for unrelated targets. In (2) we observed no similar effect. The results are interpreted as evidence for the Active Filler Hypothesis (Clifton & Frazier, 1989).

In two further cross-modal lexical decision experiments, we varied the prosodic structure of the sentences with preverbal trace positions presenting them (1) as in our earlier experiments with neutral focus (NF) and (3) with verb focus (VF). The respective sentences did not include a prosodic boundary in the critical position (t(i)) (contrary to predictions by Feldhaus et al. (1996)).

(1)

Der Krug(i) ist [einem jungen Richter des Berliner (c) GeRICHTS (t(i)) zerbrochen]NF.

(3)

Der Krug(i) ist einem jungen Richter des Berliner (c) Gerichts (t(i)) [zerBROchen]VF.

In (1) the reactivation effect was replicated. By contrast, this effect disappeared in (3). We conclude that the parser uses prosodic information to identify a trace position. Entailments for the Active Filler Hypothesis will be discussed.

References

Clifton, C. & Frazier, L. (1989). Comprehending sentences with long-distance dependencies. In G. N. Carlson & M. K. Tanenhaus (eds.). Linguistic structure in language processing. Dordrecht: Kluwer.

Feldhaus, A., Batliner, A., Geissler, S., Kiessling, A., Kiss, T., Kompe, R., & Noeth, E. (1996). Prosody, Empty Categories and parsing: A success story. Proceedings of the International Conference on Spoken Language Processing, 4. Philadelphia, PA.

Grewendorf, G. (1989). Ergativity in German. Dordrecht: Foris. Muckel, S. & Pechmann, T. (submitted). Predictive antecedent reactivation: Evidence from the processing of German verb-final sentences.