No abstract

Always use the definitive version when citing.

A well-known fact about Basque is that word order is sensitive to sentence polarity. In negative sentences, the relative order of the auxiliary and main verb is V-Aux, while in affirmative (neutral) sentences the order is Neg-Aux-V. The main goal of this paper is to argue for a new approach to this phenomenon within a grammar that assumes antisymmetry (Kayne 1994), and only XP movement. The proposal has two main components. First, drawing on Cinque's (1999) universal hierarchy of functional heads, it argues that the negative order-Aux-V-is the merged order. The affirmative V-Aux order is produced by raising of the VP to a higher, polarity-related position called PolP.Second, PolP is also argued to host the negative morpheme, ez, in negative sentences, and the emphatic marker, ba, in emphatic affirmatives. Evidence from scope interaction between negation and evidential and speech act particles suggests that the surface position of ez cannot be its merged position; rather ez is merged in a lower position and raises to spec, PolP.Part 1 of this paper discusses the surface ordering of clausal functional heads within Cinque's (1999) framework, and argues for VP raising in affirmative sentences.Part 2 argues for raising of ez to PolP. Part 3 discusses the behavior of verbal dependents drawing on Koopman and Szabolcsi's (2000) analysis of verb movement in Hungarian, German and Dutch. Part 4 addresses some remaining empirical problems.

This paper focuses on cross-speaker variation in the syntax and semantics of English be like quotatives as in (1). (1) She was like, "Ok, fine." (a) 'She thought/felt like, "Ok, fine."' (b) 'She said "Ok, fine."' Recent corpus-based work on be like has suggested that as it has continued to spread, it has undergone semantic change (e.g. Tagliamonte & Hudson 1999): be like predicates, originally used exclusively to describe non-speech states of individuals as in (1a), have taken on an additional guise as descriptors of saying events as in (1b). This paper investigates these claims further using experimental data, and proposes a syntactic and semantic account of be like quotative predicates. Of particular relevance to the conference theme, our results best fit a semantic rather than lexical ambiguity account of variation in the eventsemantics of copular be. Contrary to the direction of much current theory (Chomsky 2004), our results therefore indicate more rather than less syntax/semantics in grammar. Experiment. We examined 121 native speakers of American English aged 18-73 (M=31.3, SD=11.6) in a self-paced online magnitude estimation procedure during the summer of 2009. The experiment compares scores for be like and say sentences in six environments, as illustrated in (1)-(6). The baseline context was designed with no stativity/eventivity bias, as in (1). The contexts in (2)-(5) bias event readings (Dowty 1979). A final environment with for adverbials, as in (6), biased atelic readings-stative, non-speech be like in the be like case. (2) She was being like/saying, "They're coming tomorrow at 11:00" (progressives) (3) Just be like/say, "They won't ever do it." (imperatives) (4) Tim forced him to say/be like, "Fine, I'll do it next week." (force…to) (5) What she needs to do is say/be like, "John already quit." (do pseudoclefts) (6) For an hour, Mark was like/said, "Let's go to McDonald's." (for adverbials) Two lexicalisations were created for each environment, each assigned either to a be like or say condition yielding two test sets. Subjects were randomly assigned to test sets, and a unique random order of the 12 test sentences and 18 fillers was created for each subject. Results. The results support two main findings. First, the data show continued diffusion of be like in both event-and state-biased contexts. Mixed effect linear models revealed a significant age*verb interaction for all environments except progressives; this is suggestive at p=.08. Second, as shown in Figure 1, the trajectories of change for be like in the for condition (purple) and event-biased conditions have similar slopes: linear regression analyses revealed no significant interaction between between age and condition (p≥.32), a finding in keeping with corpus results (Buchstaller & D'Arcy 2009, Tagliamonte & D'Arcy 2007). Discussion. The above results showing the diffusion of be like in both state-and eventbiased contexts, recall the much studied availability of copula be in active contexts as in (8) (Partee 1973, Dowty 1979, Parsons 1990, Rot...

This paper proposes an account of some properties of the manner quotative constructionsbe like[Quote] in English andhebben (zo)iets van[Quote] in Dutch. We make two main claims about these constructions. First, in the spirit of Rothstein’s (1999) proposal for adjectival predicates of copulabe, we propose that eventive direct speech interpretations of these quotatives are derived via a coercion mechanism akin to those that make count readings out of mass nouns in the nominal domain. Second, adapting a proposal forbe likeoriginally made by Kayne (2007), we propose that some exceptional syntactic properties ofbe likeas a quote introducer in English are explained by the presence of a silentsomethingquantifier, which takes alike-headed PP as its complement. We compare Englishbe likequotatives with innovative (zo)iets vanquotative constructions in Dutch, which contain an overtsomethingquantifier and behave similarly. Keywords: quotative; English; Dutch; copula; event; coercion; have/be alternations