Volume 41, Issue 2

This paper discusses the forms and functions of the German response particle jein – a blend of ja and nein – in spoken conversation and in written texts. It is shown that in German spoken conversation, jein is realized as a sequentially expanding second part of an adjacency pair which refers back to the preceding turn as the first part (usually, but not necessarily a question) and indicates that more conversational work needs to be done before the conditional relevance of the first pair part is successfully resolved. In written newspaper articles, by contrast, jein is a rather stylistic and rhetoric means to establish the impression of argumentative balance, to condense complex issues and to reflect them in a linguistically innovative way. Since the written jein also allows for iconic plays with its graphematic gestalt to distinguish between a more or less ‘disagreeing’ jein , jein can be analyzed as a formally and functionally fully-fledged communicative means, regardless of whether it is used in spoken conversation or in written texts: Even though its spoken and written forms and functions are related to each other, the written jein is not simply a secondary reproduction of the spoken jein .

In German, singular count nouns usually are accompanied by a determiner (cf. Ich habe ein Auto , ‘I have a car’ vs. * Ich habe Auto , ‘I have car’). In recent years, however, the singular count noun Vertrag (‘contract’) is frequently used without a determiner, as in Er hat Vertrag bis 2013 (‘He has contract until 2013’). Although there is some doubt, intuitively, whether or not this construction is grammatically acceptable to native speakers, it definitely exists in today’s German. In this contribution, we examine the in-between status of the Vertrag+haben -construction in German in more detail. We try to find an answer to the following question: How can we explain that Vertrag haben , on the one hand, is less acceptable than the regular Hunger haben (‘be hungry’ [‘have hunger’]) but, on the other hand, is much more acceptable than * Auto haben (‘have car’)? In a diachronic corpus study on (Er) hat Vertrag , we show, first, that this construction is used in German newspaper texts since 1990 with increasing frequency. Then, in a short survey of the general distribution of determinerless count nouns in German, we show that Vertrag haben does not fall under one of the established classes of determinerless constructions in German. Against this background, we take a contrastive look at Swedish, which allows for a great variety of determinerless constructions (e. g., Han har bil ‘He has car’). More specifically, we discuss morpho-syntactic, semantic and information structural properties of Swedish determinerless constructions. Some of these characteristics turn out to be similar for Vertrag haben . The contrastive view leads us to the more general assumption that Vertrag haben is marked because of lacking morpho-syntactic licensing, in contrast to the relevant Swedish constructions. As an account of the problem of how Vertrag haben could achieve the status of a German construction to begin with, we propose that it is the semantic noun class of Vertrag , which we define as abstract, but countable noun, which contributes to the licensing of the construction.

Compounding is one of of the most productive and flexible word formation processes in German, yet it represents a challenge for German learners by virtue of its heterogeneous semantics and morphological complexity. This paper examines, on the basis of corpus data, how advanced German learners employ compounding in practice, how learner usage of compounding differs from native usage, and what types of compounds can be distinguished in these respects. The analysis of the data shows that much like their native counterparts, learners acquire very many semantically distinct compounding constructions at all levels of abstraction, which can be described as a hierarchical constructional network. Although advanced learners produce remarkably few formal errors in compound formation it will be shown that compounding occurs significantly more rarely and is used substantially less productively in learner data as compared to native controls, and that learners have added difficulties in the formation of especially deverbal compounds with internal argument structure.

Philosophers such as Bertrand Russell, Ludwig Wittgenstein, Alfred Tarski and Paul Grice have taken the position that wenn-dann (Engl. if-then ) constitutes a truth-value function. They identify the subjunction wenn-dann ( if-then ) with the conditional of classical bivalent logic, or at least regard it as equivalent. Our article objects to the assumption that the truth value of a conditional sentence depends exclusively on the truth values of its sub-sentences. While it is far from clear whether the inference rules of classical logic, such as contraposition, are universally valid, application of classical logic may nevertheless have some advantages. In the context of natural and ordinary language varieties the inference rules appear both elegant and easy to use. Our article takes a pragmatic approach which is based on classical logic, but without equalizing wenn-dann ( if-then ) with the classical conditional. It is essential to draw a linguistic distinction between real conditional sentences, which are based on a ground-consequence-relation, and concessive if -clauses. In English it may sometimes be difficult to distinguish between real conditionals and concessive if -clauses: there is no grammatical need to mark concessive if -clauses with expressions like even or still . German, by contrast, has a stronger iconic potential: it clearly marks most concessive if -clauses with selbst wenn , auch wenn , or with concessive terms like trotzdem and dennoch . Our article shows that it is always possible to apply contraposition to real conditional sentences, while it is impossible to apply it to concessive if -clauses.