This paper investigates the position of'heavy' nominal objects in Old Saxon and other Germanic languages. A new empirical study of Old Saxon is carried out and regression analysis performed, with information status, grammatical weight and case all serving as predictors. On the theoretical side it is argued that an analysis in terms of movement to specifier positions in a low left periphery goes some of the way towards capturing the observed facts, but still suffers from certain problems.
1.
Introduction
1The factors conditioning the alternation between object-verb (OV) and verb-object (VO) orders in Old English (OE) have been the subject of a lot of attention in recent years (see Pintzuk 2002Pintzuk , 2005 and the references given there; Biberauer & Roberts 2005 Wallenberg 2009, to appear; Taylor & Pintzuk 2010, to appear). Initial investigation of the data revealed that information structure seemed to play a role in conditioning the alternation, leading naturally to the simple hypothesis that objects are postverbal iff they are new information (cf. e.g. Roberts 1997:412). Taylor and Pintzuk (2010, to appear), however, show that this hypothesis is false for OE, and that the mapping from surface constituent order to information structure must therefore be more complex than previously assumed; furthermore, they show that grammatical weight -how 'long' or 'heavy' the object is -plays a role. In this paper I investigate Benjamins, 2014. -(Linguistik aktuell ; 213). -S. 313-340. -ISBN 978-90-272-5596-9 https://dx.doi.org/10.1075/la.213.12wal 314 whether Old Saxon (OS), the closest attested contemporary relative of OE, patterns with OE in these respects -a question that is interesting not only for its own sake but also for the light it may shed on the diachrony of the West Germanic languages. The paper is split into two key segments. Firstly, after laying out the theoretical background in Section 2 and methodology in Section 3, I present new data drawn from the OS Helimzd, and discuss their relevance for analyses of the OV/VO alternation, in Section 4. The results indicate that with respect to the information structure and weight of objects OS can indeed be said to pattern with OE. Section 5 broadens the focu s to processes of'rightward movement'/Heavy NP Shift (HNPS) in Germanic as a whole, and deals with the question of how to analyse these. The starting point is the analysis presented by Wallenberg (2009, to appear), in which HNPS is viewed as a diachronically stable phenomenon involving movement of the object to SpecFocusP at the clausal left periphery followed by remnant movement ofTP to a higher SpecTopicP. While the analysis is attractive in unifying 'rightward movement' cross-linguistically with leftward movement, within a uniformly head-initial approach to phrase structure d la Kayne (1994) and restrictive clausal cartography following Rizzi (1997), I present i a number of empirical problems, which lead me to propose a slightly different alternative; my account maintains the assumptions of head-initia...
