Search

In this paper we address the major word order patterns in Old English (OE) embedded clauses and examine to what extent it is possible to capture the motivation behind the variation observed in purely syntactic terms. As a representative of the syntacticocentric approach, we take the account by Biberauer and Roberts (2005) and critically evaluate it against the larger empirical background, while also pointing out certain theoretical challenges it faces. We then attempt to characterize the various information-structural inferences (focus, topicality, givenness, pragmatic presupposition) that arise for different word order patterns, and that must be taken into consideration in any successful analysis. We leave it open whether the IS/pragmatic effects we describe arise because notions like [+Focus] are represented in the syntax, or because languages exploit the contrasts that are independently generated by the syntax for discourse-pragmatic purposes. [ABSTRACT FROM AUTHOR]

This paper examines the syntactic properties of the Serbian modal verb trebati ‘need’, which appears in the environment ‘NP – trebati ‘need’ – finite da-clause’. I show that trebati is a raising verb and that the preverbal NP is a (raised) subject. Trebati (φ-)agrees with the preverbal NP only optionally, which is surprising since other Serbian verbs agree with their subjects obligatorily. Furthermore, the subject is free to remain in the embedded clause, suggesting that the raising operation is not triggered by the need to satisfy unvalued features on matrix T (contra e.g., Chomsky 1981, 2008). I instead propose that A-movement (of this kind) is ‘free’; more precisely, it is fully optional, it can occur at any stage of the derivation (or not), and it is constrained only by the requirement that the output be well-formed. I show that this analysis accounts for the full range of data with trebati, but that it can also be applied to English-style raising constructions. [ABSTRACT FROM AUTHOR]