This monograph addresses divergent views in the linguistic literature on whether German displays the that-trace effect and other subject/object asymmetries commonly found for long extractions in English and other languages. Using newly developed rating methodologies, the author exposes consistent and robust subject/object asymmetries in German – a surprisingly unequivocal result given that the existence of these effects is controversial. This finding raises important questions: how can one account for the discrepancy between the clear experimental evidence on the one hand, and the lack of consensus in the linguistic literature on the other? And secondly, it raises again the old question of why subject extractions are dispreferred. This work also provides intriguing new insights into the long-standing question on how to analyse German constructions such as Wer glaubst du hat recht? – the ‘parenthesis versus extraction debate'. In this work decisive evidence points in favour of the parenthetical analysis.

Wolfgang Sternefeld, University of Tübingen:
This volume is the first in-depth empirical study of long extraction in German, using experimental rating studies. The results invite the linguist to take a fresh look at the nature of universal grammar and the issue of parameter setting, suggesting that parameters are set by threshold levels in an underlying system of continuous deviations from universal principles.