Clausal syntax of German. By Judith Berman. (Studies in constraint-based lexicalism.) Stanford, CA: CSLI Publications, 2003. Pp. 200. ISBN 1575863626. $25.

Reviewed by Ivan Ortega-Santos, College Park, MD

Based on her 2000 University of Stuttgart dissertation, Judith Berman’s book focuses on the status of subjects, complement clauses, and the debate on the existence of traces in German, a language with relatively free word order and a relatively rich morphology. In particular, B discusses verb-second constructions, so-called ‘subjectless’ clauses, expletives and agreement, weak-crossover, long-distance dependencies, the distribution of subordinate clauses, and the cooccurrence of correlative pronouns and embedded clauses. The framework used is lexical-functional grammar (LFG). B’s goal is not only to shed light on such topics but also to present their relevance for the theory of LFG in general.

After presenting the LFG framework, B studies verb-second phenomena and the free distribution of subjects and objects. Such discussion, together with the fact that the subject may be included in VP-topicalizations, leads her to conclude that syntactic functions in this language are not structurally encoded but rather identified by morphology (Ch. 3, 23–44). As to the status of subjects, B argues that German is consistent with the subject condition (the requirement that every sentence have a subject (Baker 1983)) in spite of the fact that certain kinds of finite clauses can or must occur without a lexically realized subject. Such apparent counterexamples would be explained by the satisfaction of the subject condition by the verbal agreement morphology (Ch.4, 45–74).

With regard to the debate on the (non)existence of traces, B defends the view that local word-order alternations do not involve an antecedent-gap configuration, but nonlocal dependencies do. The fact that in German free word order is restricted to the local clause suggests that morphology identifies the syntactic functions only locally. Under this view, in the case of nonlocal dependencies an empty category in the local domain of the predicate is necessary to guarantee the right predicate-argument relation (Chs. 5 and 6, 75–121), an analysis in the spirit of Bresnan 2001. In addition, B argues that sentential arguments bear the same grammatical function as the corresponding nominal or prepositional arguments, in contrast to the traditional LFG analysis. Other proposals are that in German, there is a thematic as well as a nonthematic es, and that finite clauses in sentence-initial position are obligatorily left-dislocated.

This work not only is remarkable as the first-large scale treatment of German syntax in LFG, but it also discusses different hot topics within that theory (e.g. the subject condition or the status of traces).