Restrukturierung und Lokalitaet. Universelle Beschraenkungen fuer Wortstellungsvarianten

Restructuring phenomena are usually spoken of in the context of processes of movement which, in languages such as German and Spanish, can only cross the sentence only in certain infinite constructions. Long NP-Movement, scrambling and clitic climbing out of infinitives are typical restructuring phenomena. The first challenge that arises with regard to these processes from a universal grammatical perspective is to subsume restructuring in the Romance and Germanic languages under a single explanation. This attempt is made in the present book. Furthermore, it is argued that languages such as Japanese, Korean, Hindi and Persian with finite languages with finite clauses exhibit "restructuring", and that the same process is used that is responsible for restructuring with infinitives. The basic idea is that restructuring is licensed by incorporation. Incorporation is seen as a feature-checking process responsible for the restructuring possibilities in the different languages. Based on the assumption that parameterised properties of functional categories are decisive for language-specific differences, a theory is developed which describes language variations with regard to the restructuring option in a unified way and in accordance with the basic ideas of the Minimalist syntax program. The work is set in the theoretical framework of the Principles and Parameters Model (Chomsky 1986a, 1986b, 1991, 1993, 1994, 1995). Chapters 1-3 (Part One) explain the theoretical foundations from which I start my analysis. The sentence structure of German is discussed in Chapter 1. Chapter 2 contains a discussion of the ECP and barriers. The meaning of intermediate adjunction in movement is addressed in Chapter 3 in the context of different types of movement. I formulate the "Constraint on Adjunction Movement," which states that intermediate adjunction is generally excluded in both XP and X° movement. Transparency effects of control infinitives have in the past often been derived from special assumptions concerning their categorial status. At the beginning of Part Two (chapters 4-6), chapter 4 discusses the different restructuring analyses that have been proposed in the past. In chapter 5, I propose my own analysis focusing on restructuring phenomena related to control infinitives in German, Polish and Spanish. First, the observation is made that the contexts in which restructuring can occur in these languages are identical. Long scrambling and clitic climbing are possible in Polish and German exclusively when the infinitive complements occupy the position of the direct object, to give just one example. These and other similarities are explained on the basis of an incorporation analysis. In chapter 6, this restructuring analysis is applied to languages that show corresponding transparency effects with finite clauses. Type 1 languages (Korean, Japanese) only allow long object scrambling, whereas in type 2 languages (Hindi, Persian) objects, subjects and adjuncts can be scrambled from finite clauses. In languages such as German and Polish (type 3 languages), on the other hand, long scrambling from finite clauses is excluded entirely. The parameterised properties of C° are held responsible for these differences. In the last part of the book (Part Three), some problems arising from the discussion in the previous chapters are discussed. In connection with remnant movement, the question arises whether the ECP allows deriving locality constraints for movement also in the context of unbound traces. I argue that this is the case if one assumes that the barrier character of moving XPs can be determined "dynamically", i.e. in principle at each step of the derivation. Independent evidence for this assumption is provided by answering the question when Principle A and Principle B/C of the binding theory apply (Chapter 7). Furthermore, it is unclear why control infinitives in German, which behave like CPs, do not allow sentence-internal wh-movement (chapter 8). The absence of wh-infinitives in German and other languages or constructions is attributed to independent parameterised properties of the functional C°-head.