Optimality, markedness, and word order in German

Abstract

The main goal of this article is to develop an approach to free word order structures in German that reconciles the findings of two different lines of research: competition-based models that center around the interaction of factors like definiteness, animacy, Case, and focus on the one hand, and optimality theoretic syntax with its violable and ranked constraints on the other. The analysis relies on the existence of a syntactic scrambling operation. The major claims are: (i) Scrambling is triggered by a subhierarchy of violable and ranked linearization constraints. (ii) Optimality under at least one linearization constraint results in grammaticality, optimality under the whole subhierarchy results in an unmarked structure (unmarked structures do not correspond to D-structures, as is often assumed). (iii) The distinction between subhierarchies and matrix hierarchies in optimality theory parallels the traditional distinction between weak and strong rules. It accounts for the difference between weak pronoun fronting to a Wackernagel position, which results in a fixed order, and scrambling to VP, which does not. (iv) Language-specific variation with respect to scrambling options is due to constraint reranking. * For helpful comments and discussion, I would like to thank

Phonology competes with syntax: experimental evidence for the interaction of word order and accent placement in the realization of Information Structure

“…We will illustrate this with respect to Müller's (1999) approach. Müller assumes a distinction between grammaticality (manifested in binary judgments) and markedness (associated with preferences).…”

“…A diverse number of proposals have been put forward in the OT literature regarding which representations are adequate as inputs. Proposals include predicate argument structures (Legendre et al, 1995), sets of lexemes (Grimshaw, 1997), LFG-style f-structures (Bresnan, 1999), or syntactic derivations (Müller, 1999). For the purpose of this paper, minimal assumptions concerning the input are sufficient.…”

“…Some constraints seem to induce conventional binary reactions, and a model of gradience has to take this into account. A number of two-level models have been proposed in the literature that accommodate a concept of gradience while preserving a binary notion of grammaticality for certain constraints (Keller, 1998;Müller, 1999). Such a two-level approach is compatible with the analysis of the syntax/phonology interaction in the present paper.…”

“…The purpose of these constraints is to facilitate a systematic discussion of the data and to exemplify how an OT-based approach can capture basic aspects of the experimental results. It is outside the scope of the present paper to provide a full-fledged linguistic analysis of the phenomena under consideration, and we will therefore restrict ourselves to a fairly descriptive formulation of the constraints (for more linguistically sophisticated OT accounts of Information Structure and word order, see Choi, 1996;Müller, 1999;Samek-Lodovici, 1996).…”

Weak function word shift

“…The examples in (62) as well as (63-a) use quite complex NPs as subjects. That lighter NPs precede heavier NPs is a general tendency which is well-known from word order regularities in the middle field of the German clause (see Vogel and Steinbach 1998;Müller 1999, and references cited there), as well as from English Heavy NP shift.…”