PUB - Publikationen an der Universität Bielefeld

This explorative study focuses on _grammatical taboos_ in German, morphosyntactic constructions which are subject to stigmatisation, as they regularly occur in standard languages. They are subjected to systematic experimental testing in a questionnaire study with gradient rating scales on two salient and two non-salient grammatical taboo phenomena of German. The study is divided into three subexperiments with different judgement types, an aesthetic judgement, a norm-oriented judgement and the sort of possibility judgement that comes closest to linguists’ understanding of grammar. Included in the investigated material are also examples of ordinary gradient grammaticality: unmarked, marked and ungrammatical sentences. The empirical characteristics of grammatical taboos are compared to those ordinary cases with the finding that they are rated at the level of markedness, but differ from ordinary markedness in that they produce a different pattern of between-subject variance. In addition, we find that grammatical taboos have a particular disadvantage under the aesthetic judgement type. The paper also introduces the concept of _empirical grammaticality_ as a necessary theoretical cornerstone for empirical linguistics. Methodically, the study applies a mix of parametric and non-parametric methods of statistical analysis.