You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster A13, Wednesday, November 8, 10:30 – 11:45 am, Harborview and Loch Raven Ballrooms

The brain differentiates between known and unknown word compositions but not between transparent and opaque meaning composition: ERP-evidence from the processing of German nominal compounds and pseudo-compounds

Carsten Eulitz1, Eva Smolka1;1University of Konstanz, Department of Linguistics

This study investigated whether lexical representations of German compounds incorporate the constituent structure, and whether this is affected by the semantic transparency of the whole-word compound. To this end, we compared compound triplets that held the same head such as /auge/ ("eye"): (a) semantically transparent compounds such as /Hundeauge/ ("dog's eye"), (b) semantically opaque compounds such as /Hühnerauge/ (“corn”; literal: “hen's eye”), and (c) pseudo-compounds that were neologisms consisting of legal noun constituents such as /Hosenauge/ ("trousers' eye"). Pseudo-compounds were possible but non-existent word formations. We measured event-related potentials while participants read the compound stimuli and made nonword decisions to (d) nonword compounds with scrambled constituents such as /Hosenegaug/ (nonwords had a rate of 25%). Among the 35 recorded subjects, 24 participants showed a good nonword detection with d primes larger than 2. ERP-results of good performers showed an interaction of condition and electrode position. This effect was driven by a more negative amplitude of the pseudo-compound-condition compared to both semantically transparent and opaque compound conditions. This relative negativity started at 400 ms post-stimulus onset and had a central maximum that can be interpreted as an N400 effect, indicating that the brain differentiates between known (semantically transparent and opaque compounds) and unknown (pseudo-compounds) word combinations. By contrast, there were no reliable ERP effects differentiating between the semantically transparent and opaque compound conditions. This latter finding replicates our previous behavioral findings on German compounds where we observed equivalent priming effects regardless of the semantic transparency of modifiers or heads. The present ERP findings indicate that the lexical representation of German compounds refers to their constituent structure, and this occurs regardless of the semantic transparency of the whole-word compound. Different from other Indo-European languages, constituent structure thus needs to be incorporated in the lexical representation and processing of German. Altogether, these findings stress the importance of cross-language comparisons to understand the representation and processing of morphologically complex words.

Topic Area: Grammar: Morphology

Back to Poster Schedule