Retour vers liste

Détail de la contribution

Auteur: Eric HAEBERLI

Co-Auteur(s): Tabea IHSANE, University of Geneva, Suisse

On the Interaction between Syntax and Morphology: New Evidence from the Loss of Verb Movement in the History of English

Abstract/Résumé: Ever since Emonds (1978), the variation found across languages with respect to verb movement has been extensively discussed in the generative literature. Much of this discussion has been driven by the insight that verb movement seems to be related somehow to richness in inflectional morphology (agreement). This correlation, often referred to as the Rich Agreement Hypothesis (RAH), has been expressed in different ways. According to the strong version of RAH (cf. e.g. Rohrbacher 1999, Koeneman and Zeijlstra 2012), the relation between syntax and morphology is biconditional: A language has verb movement iff it has rich agreement. A weaker version of the RAH (cf. e.g. Bobaljik 2006, Bobaljik and Thráinsson 1998), suggests that rich agreement morphology entails the occurrence of verb movement but nothing can be said about languages with impoverished agreement morphology. Finally, certain authors argue that the RAH cannot be maintained at all, not even in its weaker form (e.g. Alexiadou and Fanselow 2002, Anderson 2002, Jonas 2002). This paper re-evaluates the debate on the relation between verb movement and agreement morphology on the basis of new diachronic evidence from the history of English. English has often been used as an illustration of a language that lost verb movement, and the changes in the verbal agreement paradigm around the same time have been taken as support in favour of the RAH (e.g. Roberts 1985, Rohrbacher 1999, Vikner 1995, 1997). However, the empirical basis that has been used to discuss the loss of verb movement in the history of English has remained incomplete. The focus has been to a large extent on changes in the syntax of negation (rise of do-support), and hardly any attention has been paid to the development of the placement of verbs with respect to adverbs. Taking a closer look at adverb placement in a number of parsed corpora, we will show that the changes in this domain of the syntax do not proceed in parallel with those in the domain of negation. Given the adverb data, we conclude that the decline of verb movement must have started considerably before what has been suggested on the basis of the negation data (end of 15th century instead of end of 16th century). We analyze this contrast between adverbs and negation as evidence for the hypothesis that verb movement was lost sequentially rather than in one step in the history of English (cf. also Han 2000, Han and Kroch 2000). We will then closely examine the status of inflectional morphology in the relevant period of our data (15th/16th centuries) in order to evaluate the validity of the different versions of the RAH.