Subject and Object Positions in Swedish

The issue of how to analyze so-called verb second languages has received a lot of attention in the recent theoretical literature. In most approaches, the analysis relies heavily on structure for the explanation of the word order phenomena. Starting with den Besten (1983), the finite verb is assumed to head a functional projection, whose specifier position provides the landing site for the sentence initial phrase. The main ideas behind this theoretical approach has been taken up not just within Chomskyan approaches, but can also be said to underlie aspects of Sells' (2001) analysis within LFG. These hierarchically based analyses contrast sharply with the flat analysis in the field approach which is commonly used in standard reference grammars of the Scandinavian languages. Its most well-known formulation can be found in Diderichsen (1946), but there are also more formal implementations of similar ideas by for instance Ahrenberg (1992). In this paper, we will focus on the part of the clause that follows the finite verb, the so called midfield. We will argue that the flexibility in midfield word order which we find in Swedish as well as in the other Scandinavian languages is best captured by a flat structure and that the approach to c-structure formulated within Lexical Functional Grammar (Bresnan 2001) is best equipped to capture this. The actual constituent ordering is taken care of by Optimality Theoretic constraints. We will show that the order between the elements results not just from syntactic factors, but that information structure, morphology and prosody play an important role.