On the Relative Order of Adverbs in the I-domain: A study of English and Swedish

Sammanfattning: This dissertation is a theoretical and empirical investigation of the ordering of adverbs in the I-domain. Within the generative-minimalist framework, adverbs have traditionally been analysed as adjuncts, adjoined to the syntactic structure in semantically appropriate positions (e.g. Jackendoff 1972). This view has been questioned and argued against by scholars such as Alexiadou (1997) and Cinque (e.g. 1997, 1999). According to the seminal Cinque (1999), sentential adverbs are merged to the structure as a result of feature matching with hierarchically ordered functional heads. As a consequence of the hierarchical order of the functional heads, the adverbs associated with them are also hierarchically ordered. One aim of this dissertation is to investigate whether Cinque's proposal regarding the relative order of adverbs is corroborated by authentic English and Swedish data and by the intuition of native speakers. Secondly, even though Cinque's theory makes no predictions about adverbs that have no features in common with the projections in the I-domain, a further aim of this study is to examine to what extent other types of adverbs (e.g. connective adverbs) can be argued to be hierarchically ordered as well. In other words, this study is partly an attempt at falsifying (or corroborating) Cinque, and partly an independent investigation of the order of English and Swedish adverbs in the I-domain. The following five research questions are addressed: (a) Can English and Swedish adverbs from various semantic adverb classes be ordered relative to each other, and if they can, are the two languages? relative adverb orders congruent?; (b) If the answer to research question (a) turns out to be positive, in the sense that adverbs from English and/or Swedish can actually be ordered relative to each other, what is the relation between these orders and the order proposed by Cinque?; (c) Is it reasonable to assume a universal hierarchy of functional projections in the I-domain?; (d) Is it reasonable to suppose that IP-adverbs occupy the specifier-positions of the above-mentioned functional projections, or is the old adjunction hypothesis more theoretically and empirically adequate, or should some third solution be chosen?; (e) If research question (c) can be answered in the affirmative, is it then possible to create an intuitively appealing, empirically defensible, and theoretically adequate and elegant theory that capitalises on Cinque's empirical findings and theoretical solutions, Thráinsson's (1996) Real Minimalist Principle, and Starke's (2001, 2004) views on X-bar theory, Merge/Move, and the functional sequence?